Sometimes people make various suggestions that we should simply build “safe” artificial Superintelligence (ASI), rather than the presumably “unsafe” kind.1

There are various flavors of “safe” people suggest.

• Sometimes they suggest building “aligned” ASI: You have a full agentic autonomous god-like ASI running around, but it really really loves you and definitely will do the right thing.

• Sometimes they suggest we should simply build “tool AI” or “non-agentic” AI.

• Sometimes they have even more exotic, or more obviously-stupid ideas.

Now I could argue at lengths about why this is astronomically harder than people think it is, why their various proposals are almost universally unworkable, why even attempting this is insanely immoral2, but that’s not the main point I want to make.

Instead, I want to make a simpler point:

Assume you have a research agenda that, if executed, results in a ASI-tier powerful software system that you can “control”.3

Punchline: On your way to figuring out how to build controllable ASI, you will have figured out how to build unsafe ASI, because unsafe ASI is vastly easier to build than controlled ASI, and is on the same tech path.

You can’t build a controlled ASI without knowing many, MANY things about intelligence and how to build it.

So this then bottlenecks the dual technical problems of “how to find an agenda that results in controllable ASI” and “how to execute on such an agenda” on “even if you had such an agenda, how do you execute it without accidentally, or due to some asshole leaving the project or reading your papers, building unsafe ASI along the way?”

No one I know pursuing various agendas of this type has answers to these questions. And lets be crystal clear: This is the fundamental question any sensible “safe ASI” project needs to answer before even being worth considering.

You would need to either have:

• Some absurd level of institutional secrecy and control (e.g. “this research will exclusively be done inside Area 51 and we assassinate everyone who leaves the project and also nuke literally everyone else that tries”)

• Complete technical orthogonality (“this research is so radically different from other research that it cannot even in principle be used to build unsafe ASI, only safe ASI”, which is impossible)

• A global ban on ASI development and competent enforcement

This means that the primary prerequisite to even considering starting to work on a safe ASI plan is to have a global ASI ban and powerful enforcement already in place.4

Sometimes I say things along the lines of “Corporations are not your friend.”

By this, I don’t mean “The people working at Corporations are big meanies :(“

What I mean is that it is a type error.

Corporations, governments, groups, are not the kind of things you can be “friends” with.

You can be friends with individuals.1 If you think you are friends with a corporation, or a corporation cares about you in the way a human or a friend cares about you, you are confused.

This doesn’t mean corporations can’t act for good, they often do, but the mechanisms that cause them to act well are extremely different from the ones that cause your friends to treat you well!

If you used the same tactics you use to get your friends to treat you well (e.g. hanging out together a lot, exchanging favors, asking nicely, promising you owe them one, just generally looking distressed, genuinely loving one another, etc) on a corporation, you will generally get screwed.2

You can even have a friendly relationship to every single person in a corporation or other type of group and the group can still functionally be your adversary.

—

The claim is not “corporations are inherently evil and/or should be abolished.”

Quite the opposite: A massive share of the good of the world is created by corporations doing corporate things to deliver cheap and quality goods to consumers!

But it is important to understand that markets/corporations/capitalism are one set of tools in our toolbelt for solving problems, and it’s a different tool from the kinds we use in interpersonal 1:1 relationships. And it’s not the only one at our disposal for solving problems.3

Four years ago, I started a company. That company’s name was Conjecture.

It was ambitious, audacious, at times crazy. It was the best four years of my life, and I have learned a lot. But now sadly, this chapter draws to a close.

I always really appreciate when other people that have done ambitious things shared their experiences, what happened, what went well or wrong, what they would have done differently, what they have learned.

So, in that spirit, here is some history and reflections on four years of Conjecture!

History

Part 0: 2021 - EleutherAI and Founding

Conjecture grew out of my desire to build something more, something bigger, better, more ambitious, after EleutherAI. We had done a lot of impressive things, including building some of the very first fully open source LLMs.1

But the anarchic, volunteer driven nature of EleutherAI was hitting its limits, and we needed more in order to do what I initially had always wanted to do, which is to help build a better world with safe, controllable AI.

The stars aligned when I was approached by our first investor, Nat, and we discussed whether I had ever considered starting a company. I, in fact, had been being harassed nonstop for almost 18 months by my friend Gabriel to start a company with him, and I was hitting the limits of what EleutherAI could do in my ambitions, and so the timing couldn’t have been better.

Recruiting my good friend and technical genius Sid, the initial trio was complete, and along with a small group of other EleutherAI veterans, Conjecture was born.

Part 1: 2022 - Try Everything

Conjecture officially kicked off March 1st, 2022, with me and Sid moving to the UK that very day. Our goal was as nebulous as it was ambitious: Solve technical AI safety, or die trying.

And so, we took the first year to try as many different approaches as we could as quickly as possible. We built cutting edge LLM infrastructure, did interpretability experiments, explored the epistemology of alignment, ran a research incubator, investigated just about every alignment proposal under the sun, pioneered Simulator Theory and much, much more.

Ultimately, our conclusion was: All of these approaches are terrible.

No one has a plan, no one is making any meaningful progress towards anything that even resembles alignment. So we went back to the drawing board.

Alignment is way too hard, maybe impossible. What is a goal that would be meaningful and achievable?

In the fall/winter of 2022 came the first breakthrough. Our target would be boundedness: The ability to know the bounds of a system, what a system cannot do. If we could predictably build AIs that we know how to bound, this would give us the fundamental primitive necessary to build a new paradigm of controllable AI.

We named our approach to achieving boundedness Cognitive Emulation (“CoEm”).

Part 2: 2023 - Rocky Transition

CoEm as a research agenda was far outside the overton window of the time2, and although me and Gabe were fully bought in on the vision, the team still wanted to keep some diversification.

And so throughout 2023, we continued to bit by bit cull non-productive research projects, while also making our first attempts at commercial products. We also during this time built our first state of the art LLMs.

This transition wasn’t easy, and frustration from the research team over mounting setbacks and failures grew.

Conjecture came to a crossroads: All in on CoEm, or something else?

This debate culminated with significant turnover and strategic disagreement, which marked a transition point for Conjecture. It was heartbreaking to see many good people, and friends, leave, but it was the right decision for the company.

We had more ambitious plans in mind, and needed to give it all we’ve got.

Part 3: 2024 - CoEm

2024 was the year when finally the company started to focus. I would say this is without a doubt the time in which we produced the most scientifically novel and interesting work.

We were dedicated to the CoEm agenda. If we could demonstrate a novel way to build AI in a bounded, yet powerful, way, in which we could know the capabilities of an AI system before it was built/deployed, then we would have the technical foundations not just for a killer product, but also for a “safe by construction” regulatory regime, which would allow a vastly economically beneficial form of regulation that still addresses the existential risk from superintelligence effectively.

With this approach, it would become possible to demonstrate one’s AI cannot be superintelligent, and so that it is safe to use, while the AIs that cannot demonstrate this can be simply banned. A win-win for everyone…except the people carelessly gambling with our lives.

Over the summer, we made several noteworthy technical breakthroughs, and in winter, we culminated with the public demonstration of Tactics and our roadmap. I am personally still very proud of this roadmap, it is one of my favorite things I ever produced at Conjecture.

Novel scientific and engineering research is hard, and the rest of the field was advancing at breakneck speed, but it seemed we finally had a handle, a way out…

Part 4: 2025 - Product Pivot

But ultimately, it was too little, too late.

Coming into 2025, we were faced with an incredibly hard choice. We had made progress on CoEm, yes, and we had a concrete roadmap forward.

But the economics just didn’t add up. Doing research on frontier LLMs is incredibly expensive work, and we were still very far from the level of commercialization that would allow us to raise the kinds of mega-rounds that other AI players had.

And so, we made the hard choice to pivot from research to pure play product work.

We targeted a number of growing markets with little well developed competition, and had our first real successes in the commercial space. We started having our first real B2C customers in the early summer, and then branched out towards B2B partnerships.

Unfortunately, in the late summer and early autumn, the competition really took off, and we quickly saw ourselves overtaken by a lot of well funded, extremely dedicated teams with more experience in our targeted niches.

So, for one last hurrah, that December, we decided to do one more rush with the remaining team, two weeks of the most intense work we could give, to see if we could catch up, if we still had a way forward.

But sadly, it was not to be, and the results fell short of the benchmark we had set for ourselves.

Part 5: 2026 - The End?

Overall, we are very happy to have seriously tried technical AI safety at Conjecture, and to have developed CoEm as far as we did. It was extremely useful, and directly informs our views on ASI and xrisk to this day. In another world, it could have changed a lot.

The product work was an attempt to make Conjecture make sense after we stopped believing in technical AI safety. It didn’t pan out, such is life.

Lessons Learnt

In retrospect, what would I do differently? Many, many things! I’ve learnt so much, so many obvious mistakes that could have been fixed, but hindsight is always 20/20.

Things that were definitely mistakes and that I would now do differently:

• Fire more. This is by far my biggest mistake and regret. I was way, way too hesitant to fire people, rather than getting into agonizing month long strategy disagreements. This cost us so much, over and over. I should have just asserted the strategy more directly and fired anyone that was unwilling to pull their weight. There are many cases where I tried to make things work even when it was clear months or even years ahead of time that things were not going to work out, to neither my nor the employee’s benefit. I would be far more ruthless in cutting things short.

• Focus more. There was way, way too much split attention. There should have been fewer projects with a smaller team putting 100% of their attention and focus into them.

• (Micro-) Manage more. I put way too much trust into people to manage themselves. I should have from early on been much harsher in requiring written reports, clear communication of what everyone was doing, what KPIs we were working towards, clear deadlines, etc.

• Be more adversarial and use more authority. I was way too good faith and extended way too much patience with people. I too often did not treat sabotage/FUD as what it was, instead looking for compromise or peaceful solutions, instead of just asserting authority and moving on.

• Focus more on money. Money is great, it can be exchanged for goods and services. I should have been much more ruthless in optimizing for making money, this would have also had many good downstream effects on getting people to focus on real things rather than useless academic bullshit.

• Hire less PhDs. Hiring researchers and PhDs turned out to consistently be pretty terrible. They were chronically academia-brained and had weird status games and incentives that had ~nothing to do with actually trying to win.

• Communicate clearer to the public. We sucked at marketing and sales. We had a lot of great ideas and great people, and we failed to communicate them well. I would now in retrospect as the CEO have put more effort into this.

• Burn bridges with EA earlier. There are many nice people and things to be said about EA, but ultimately it’s a movement of transhumanist weirdo cultists that lie constantly and see no moral problem with this because they’re “utilitarian.” I severed ties pretty quickly (~end of year 1), but it should have been even earlier.3

Things that might or might not have been mistakes:

• Not founding in the USA. I’m conflicted on this one. I think London ultimately treated us well. We got good talent, we raised significant amounts of capital, and we were insulated from the SF brainworms (EA, accelerationism, etc etc) to a degree that let us develop a quite unique and grounded view of AI and AI safety. If we would have been able to withstand the brainworms in SF, it probably would have been better to be in SF. If we wouldn’t have, being in London was 100% the right choice.

• Being more mission aligned vs mercenarial. There was an ever-present tradeoff of whether to hire for people that were more aligned with the mission (in particular AI xrisk), or more competent and mercenarial. We ended up being more on the mission aligned end of the spectrum, and we had some good and some bad outcomes as a result. We also had some good, some bad outcomes with more mercenarial employees. Overall, it seems unclear whether we should have been more mercenarial, or even more mission alignment focused. It feels like we ended up in a bit of an awkward in-between.

Things that were not mistakes:

• Betting on AI and safety/control. It’s pretty clear to me that we were on to something big, and something that the world wanted and needed, and that no one else was going to build. My predictions for AI were very accurate and ahead of their time, and AI control remains an under-addressed problem. While we could have exploited this way more, ultimately I am happy with our prescience.

• Taking a shot with Conjecture, even if it ultimately failed. Best 4 years of my life! We had a real shot, it didn’t work out this time around, but I’m ready to roll the dice again.

The Future

So what is next for Conjecture, and for its alumni?

Ultimately, I came into technology not just because I like technology (though I do), but because I wanted to make the world a better place. To build a better world, for people, for everyone.

This is still what drives me. I continue to think technology plays an important role in a better tomorrow, but it’s not the only thing, or even the most pressing thing, at least at this point in time.

As the saying goes, “when the facts change, I change my mind. What do you do?” From my point of view, the facts have changed. Humanity is bottlenecked far more on institutional malaise, cultural nihilism (both on the left and the right) and pervasive cowardice, than it is on technical research.

—

While Conjecture’s 2025 didn’t pan out the way I had hoped, something dramatic and unexpected happened next door: ControlAI pivoted their strategy in early 2025 to the DIP: A new strategy of simple, straightforward and honest civic action.

And its success has been astounding.

Over 100 lawmakers have supported ControlAI’s campaign to call for binding regulation on superintelligence. I was recently invited by the Canadian House of Commons to testify on the risk of superintelligence. All of this was unimaginable a year ago.

The problem with ASI is not that ASI is killing us. ASI doesn’t exist, yet. It’s that we are allowing unaccountable corporations to build things that threaten the lives of us and everyone we love.

Putting an end to this is a national (and international) security problem, not a technical one.

ControlAI has shown a powerful way forward on civic action, and I think I could provide a lot of value to this approach.

As such, I will be joining ControlAI as Director of ControlAI US. I will be based in Washington DC, where I will be leading efforts to openly, democratically and honestly inform lawmakers of the risks of ASI development, and what can be done to ensure longterm human safety and flourishing.

If there is one thing that I learned from ControlAI’s open, honest advocacy in the UK, it’s that most politicians are decent people trying to do the right thing. They might not know what the right thing is, they might be being lied to, and yes, some of them really only care about themselves coming out on top.

But by far most of them are decent people that are under enormous pressure, being shouted at from all sides, trying their best with what resources they have to figure out what the right thing to do is.

I want to help these people to find and do what is right.

—

I want to build a better world, for everyone.

There is simply no good future in which we don’t have strong, just, trustworthy institutions that can responsibly steward powerful technology, such as AI.

Progress historically has been downstream of two major forces: Technological progress (writing, steam engines, electricity; the scientists and engineers) and institutional/social progress (better laws, democracy, free markets and regulation, the limited liability corporation, human rights; the statesmen and humanists).

We as a civilization have been doing a poor job lately of finding a way for these two forces to work together productively. We have made a lot of technological progress, but have not found a way to replicate the successes of the founding fathers and the enlightenment.

What would the founding fathers look like if they had had our technology? What amazing new society could they have built? What does a good future look like, that is sincere, hopeful, that one would actually want to live in?

It’s sure as hell neither luddism/degrowth nor whatever anti-human dystopia the accelerationists are trying to sell.

There is a dire need for a better vision, a vision of a future that is humane (good for humans) and human (achievable by humans). And any such vision requires strong, just government and statecraft.

Conclusion

In conclusion, all I can really say, from the bottom of my heart, is thank you! To everyone that was along with me on this crazy ride.

To Nat, who believed in me before anyone else did.
To all of our other investors and sponsors, that allowed this all to happen in the first place.
To my cofounders, without whom I could have never made it this far.
To all my employees, partners, collaborators and friends made along the way. It was an honor and a pleasure to work with you, and I hope our paths cross again.

I hope to see many of you on the next Adventure. I will be moving to the US, to Washington DC. If you’re in the area, give me a shout!

So long, Conjecture.
To another adventure, another year.
May it not be the last!

- Connor

This is the part I tend to care about most. But there are others, too.

Such as the complete replacement of all human jobs with more efficient, cheaper, and productive AI workers.

If you don’t believe this will happen, this short post will not argue the point, go somewhere else.

There are many reasons why people might care about their jobs. Status, belonging, meaning, whatever.

I think there is one reason that matters the most:

Trading our labor is the primary means for most people to gain economic and political power and leverage in the world.

I’m not saying this is necessarily good or bad, but it is how our world is set up.

So if people don’t have jobs, can’t trade their labor, they are deprived not only of economic, but also of political power.

This is very bad for people, individually, and for anyone that cares about the political systems that run our lives having any reason to care about people individually.

And you too, are a people!

This is Part 5 of a 5 Part series. You can find Part 1 here.

Once you have seen the Abyss, it can be hard to “go back.” How can you live your normal life in the Valley, now that you know what lies beyond the light of the campfire? The usual answer is just “repress it as hard as you can, pretend it never happened”1, but that’s not what we want.

Repression is just making yourself a sitting duck for the dangers you know are Out There, this is not a way towards Progress.

The Buddhists have their answer. Upon the realization that there is no Home, there is no Ok in this world, their solution is to make themselves not care about this, to accept this, to drift out to the Ocean, or wherever the Currents might take you.

And here they are so, so deeply wrong.

The Buddhists seek to stop being Water altogether, to achieve Nirvana is to cease flowing, to escape the Stream entirely. But this is deep, deep confusion. An artifact of a time before Progress. The problem isn’t that we flow, it’s that we flow without Navigation, without reading the Stars.2

We cannot stop being Water, at least not without forsaking our Birthright.

If you don’t do something, it doesn’t happen. If humans stop flowing, there is no Progress, no hope of reaching the Ocean. We must flow, but we must navigate wisely.

The answer to the Stream being Wild is not to get out and lay down to die.3 It’s to build a ship, take hold of the wheel and turn against the Current, towards the Stars.

Only then can we find the Ocean.

The Daily Practice: Navigation of the Self

So, you return home, you’re back to your daily routine. Everything feels…really quite normal. Maybe it was all a Bad Dream? Can the world really be in such peril? And if so, is it really your responsibility to do something about it?

You shake your head. You don’t want to fall back Asleep. But how?

A concept I do really like from Buddhism (and other spiritual traditions) is the concept of “having a Practice.”

Generally, when you’re on the Path, the first and most important thing one is urged to do is not some crazy big dramatic action or overturn your whole life or whatever, but to establish a Practice. To ensure that you have some habit, some regular time, to dedicate to advancing along the Path. 10 minutes per day is much better than one big chunk once and then forgetting about it.

When confronted with the Falls, with horrors such as extinction risk, I am often asked by people: what should they do?

On one hand, it seems logical they should dedicate everything to preventing such catastrophes, but on the other, that’s a lot to ask. Too much to ask, even.4

My advice is always the same: Don’t overdo it, but don’t do nothing.

Some people are Called by the Mountains, and may dedicate their whole lives to the Fight, but most people cannot and should not. Your primary loyalty should remain to the Valley, to your friends, your family, your obligations back home.

But don’t do nothing.

What you should aim for is to establish a Practice. The way you do this is you select some amount of time per week, could be a full day, could be a couple hours, could be 30 minutes, whatever you can do reliably5 and do that much, and not (much) more, every week.

In that amount of time, put in a true good faith effort to actually work on the problems, to understand the problems, to reflect upon them, to reach out to others, see how you could be helpful and actually do something.

This isn’t easy, especially if you are isolated and don’t have other people to talk to or work with (see also the next section). But luckily thanks to the internet it is easier than ever to find people that share your concerns that you can get to know and cooperate with!

Your Practice should not only include external actions6, but you should also have a component of reflection. Navigation also applies to the individual level, and you need to actually do it, or you will just be pulled along by the Currents.

If you don’t take the time to reflect on your Stars, to consider where you are currently being pulled, and whether that’s where you want to be going, well, then it doesn’t happen, you will not be Navigating.

So take this time.

My recommendation to start is to literally open an empty google doc, start a bullet point list with “GOALS” at the top, and start writing and thinking. What are your goals? What steps would it take to get there? Where are you uncertain, about what you want or about how to get there? How could you resolve those uncertainties, or become more clear? Some people also like pretending they are “interviewing themselves”, to get the thoughts flowing.

Whatever technique you use is up to you. But no matter the method, you should definitely actually write down notes, so you can also reread them later!

Maintaining a Practice is how you stop yourself from “falling back asleep”, from getting so distracted by daily samsaric life that you lose track of the Stars once more.

You must keep your eyes on the Stars. If you lose them for too long, you are Lost. Any progress will depend on this.7

There is a book and more worth that could be written here, but I think it’s valuable to keep it simple for now.

Here is a short list of some of my personal advice for once you are back in the Valley:

• Don’t over do it, but don’t do nothing.

• Have a weekly Practice where you reflect both on what to Do, and on the Stars.

• Actually do things in the real world, don’t just think.

• Actually think, don’t just do things in the real world.

• Make small, non dramatic and endorsable steps rather than dramatic leaps.

• Find a Crew.

• Don’t be weird.

• Don’t be stupid.

If enough people just did these basic things, we’d be on a much better track, maybe even good enough to make it all the way to the Ocean.

The Crew: Navigation Together

Humans are a social creature. We generally never embark on a Journey alone, and you shouldn’t either.

“One unit of humanity” is about ~one adventuring party plus family8, it’s not “one human.”

Humans are social, and their General Intelligence tends to emerge at the group level, not in the individual. A single human alone in the jungle, with no language, no friends, no family, is barely more sentient than any other animal. It requires many humans together to actually exhibit General Intelligence.

In many ways, humans are closer to a hivemind species than they are to solitary predators like tigers, in personality and temperament.9

You are the average of your ten closest contacts, as the saying goes. Coming up with new ideas, culture, art, requires feedback, regular communication with other people, bouncing ideas back and forth, long conversations over food and drink. There are exceptions of course, everyone loves a good story about a lone genius, but often in reality these lone geniuses are not lone at all, or are precisely so noteworthy because they are so rare. Progress happens when many humans cooperate.

The human Soul is not fully localized to within one human body, it is distributed across many people, their tools, their art, their culture, etc. Your Soul is partly in your friends, your family, your teachers, your rivals.10

This has important implications for your life and your Practice. If you want to Sail towards the Ocean, you will want a Crew.

There are many reasons why you will want a Crew.11

For one, it’s fun. Loneliness sucks. You can get more done together. You can cover for each other’s blindspots. And, importantly, you can help each other act and stay on target together.

Human motivation and emotions, unless you are some kind of platonic Zen Master12, are deeply influenced by your Crew. My estimation is the minimum needed to reach “emotional and motivational stability”13 is about 3-5 aligned people working together, a Crew. The Crew can be larger (though not too much larger before it becomes a more loose association), and you can of course be part of many different Crews in different parts of your life.

But in general, tackling problems as monumental as the ones we are discussing without having a Crew to back you up and keep you accountable, is really, really hard. Often, the very first step on the Journey, or even before the Journey can begin, is to assemble your Crew before you head off, or else you will want to pick one up as soon as possible on the way.

Of course, it’s important to not get too attached to a Crew that might actually be bad for you, or might not have the same Destinations in mind for the Journey as you. Humans have a lot of genetic payload that is made to make us glue ourselves as hard as possible to our local band, because historically we couldn’t just “reroll” and find new bandmates out in the wild whenever we wanted, so we had to make things work, even if our bandmates were shitheads.

But we no longer have these restrictions. You should want to find a Crew that works for you, that is aligned with what you truly care about, and that you can get along with for the long Journey ahead.

There is no simple shortcut to finding a Crew, you’ll just have to try many things. Sometimes there are already groups you can join, other times you will have to create your Crew from scratch, or by breaking off from a different, larger group.

And importantly, if you want a Crew that is actually aligned with you, you must know what it is you want to be aligned with! If you don’t know your Stars, you won’t be able to find others looking at the same Constellations. So some clarity about your Values will be essential in helping you find the kind of people you want in your Crew.

But fundamentally, human relationships are hard, and a lot of work. There are no simple universal shortcuts.14

But it’s worth it. It’s how we humans do things: Together.

Having your Crew, and your Practice, will allow you to naturally Actually Do Something. It’s the very core of you Becoming Stronger, so you can join up with the rest of the Fleet, to guide Humanity towards the Good World…

The Long Voyage: Navigation of Humanity

A beautiful thing about human sociality is that it can be wonderfully mutually beneficial. Building your Crew, helping your Crew grow, and also helping others find their Crews, becomes a very net positive thing when you are in service of the Stars.

If you are Sailing towards Human Values, you benefit from other people being Stronger. It’s always beneficial to you for your Crewmates, and even other Crews, to be healthier, stronger, smarter, happier, to have a better read on the Stars, because ultimately that’s where you are all headed.

Very often, those that are confused about the Stars will think things such as “I want to get to the Ocean, and I need Power to get there, so I will get that Power and backstab anyone else that tries to take it.”15 This is extremely negative sum16, because now you might have lots of Crews all aiming for the same goal backstabbing each other over a finite amount of Power up for grabs.

A much, much better approach is to trust in the Stars. If we can all agree we are minimally sailing in the same direction, the question becomes less “how can I empower myself?” and much more “how can I empower EVERYONE to help reach the Stars?” and this will allow everyone to pick strategies that benefit everyone, and reach the Ocean, together.

This is Coordination, and it’s humanity’s superpower.

What is special about humans is that our social capabilities are curiously “fractal”, we can form bands, bands form tribes, tribes form nations, nations form empires, etc. And what we need for this to be beneficial rather than extractive is a minimum set of rules, values and coordination, so it becomes worthwhile for each nation to be in the empire, for each tribe to be in the nation, for each band to be in the tribe.

You should be very wary of people or ideologies that want others to be Weak, that think making others more educated, wise, capable, resourceful might be a Bad Thing. This is extremely suspicious. If you were heading towards the Blazing Star, then the more people Sailing with the more capability, the better!

But if your goal is not the Blazing Star, if you want to steer towards Darkness, well, then you benefit from others being less able to hinder you.

Beware those that want you Weak rather than Empowered. And beware yourself when you hesitate from empowering others in search of the Blazing Star.

Of course, this only applies to empowering people actually in search of the Blazing Star, that are willing to fulfil the basic obligations of Coordination.

That’s how coordination works. There are some basic rules, and as long as you follow them, you’re part of the coalition, and we help each other. Making those basic rules as minimal as possible while firmly excluding destructive defectors, and then enforcing these rules properly, is the core that makes coordination, Law, possible. Without Coordination, Law, Order, there can be no shared Fleet sailing to the Good World.

There are exactly three alignments: Lawful Good, Lawful Evil and Chaotic.

If you have no Law, you cannot be Good, or even Evil, it’s just Chaos. If you have Law, the Law may be Unjust or Evil, but it’s also the only way you can have Good.

This is also why it’s important for you to deeply understand what you are aiming for, and actually Navigate, so that others can coordinate with you as well! If you are Lost, you are of no use to the Fleet.

The Long Voyage extends beyond any single person or lifetime. We must build vessels that outlast their builders. The Stream flows faster than human lives, and the Ocean lies farther than any one Crew can reach.

This is a core aspect of our Navigation: We are not sailing toward a port we’ll see, but continuing a Long Voyage started by ancestors so remote we will never know their names.

This requires a different kind of commitment. Not just to outcomes we’ll witness, but to processes that will continue when we’re gone.

Ideally, each and every one of us should be redundant, the natural shape of Humanity and its Institutions naturally flowing towards the Good World, even if we aren’t there to guide it.

Consider how the great cathedrals were built: Not by individuals who would see their completion, but by guilds that preserved knowledge across centuries. The master stonemason trained apprentices not just in cutting stone, but in reading the Stars that guided the cathedral’s construction. Each generation added their own arches while maintaining the Sacred Geometry that unified the whole.

We need such Guilds for our age. Not literal masons, but communities that preserve the knowledge of Navigation while adapting to new Waters.

The tech companies claim to be building the future, but they optimize for quarterly returns, not centuries. True Coordination means creating structures that can hold course across the Rapids of decades and the Meanders of centuries.

This is why the most important thing you can teach someone isn’t a specific route through the Waters, but how to read the Current and Stars for themselves. Every person you help become a better Navigator doesn’t just add another boat to the Fleet, they become someone who can teach others to Navigate, who spot new Falls before we reach them, who might see Constellations you missed.

A Fleet of Navigators teaching Navigation, each iteration making us collectively better at reading the Waters.

Not a hierarchy commanding from above, but a distributed society sharing knowledge of Currents and Stars. Each Crew sovereign but coordinated, all flowing toward the same distant Light.

The Ocean

But, and this is crucial, the Ocean isn’t a place that exists, waiting for us to find it.

The Ocean is what emerges from countless Meanders, each bend in the Stream creating the destination it flows toward.

We don’t sail TO the Good World, we create it by HOW we Sail.

The Ocean, the Good World, isn’t a place we can fully envision from here, and attempting to do so is fatal. We can’t draw blueprints for utopia any more than a medieval peasant could design a smartphone. What they could do was improve the waterwheel, share that knowledge, raise their children well and create conditions where the next generation could see a little farther. The Ocean emerges from thousands of these iterations, each solving problems we couldn’t have imagined before creating them.

If there is one thing we have learned, from the 20th century and before, it is that gunning straight for utopia is the one 100% surefire way to end in dystopia.

So many thought they could see the whole Path from the beginning, and they reliably only accomplished plunging themselves and everyone around them into a terrible Abyss.

Navigation means accepting that each bend in the Stream teaches us something we needed to know, something we couldn’t have learned without making the Journey. We build the Good World not by implementing a Master Plan, but by getting better at the process of building. Better institutions, better coordination methods, better ways to detect and avoid Falls, better techniques for teaching Navigation itself.

The real Work isn’t reaching a destination, but improving how we travel. Each generation must solve the problems in front of them, while building tools and knowledge the next generation will need for problems we can’t yet see.

The Ocean isn’t a fixed endpoint, it’s what becomes possible when millions of humans get good at iterating together, fixing what’s broken, preserving what works and staying oriented toward the Stars, even when we can’t see exactly where they lead.

The Voyage is never complete, we still must Fight. Not to reach something out there, but to keep creating it, moment by moment, choice by choice, as long as the Stream flows.

The Will to Fight

Without Love, It Cannot Be Seen.

Ultimately, deeply, everything wraps back around to You, and whether you Care.

Fundamentally, if you just don’t Give A Shit, if you don’t care, if you have no Love, if you don’t Believe, none of this matters.

I can teach you all the tricks of navigating the Underworld, of how to fight off the Night, about how to Navigate to the Blazing Star. But at some point, deep in your Soul, it has to bottom out:

Do you care?

Is there anything you care about? Is there anything you love?

If there is, then you too have a Shard of the Blazing Star. I think all but the most broken humans have the Shard, they can Hear The Call.17

I can teach you how to Fight, but I can’t teach you to want to Fight.

There are many things that can dampen your Will To Fight. Trauma, lack of resources, lack of a Crew, lack of Hope. These are Hindrances, which can be overcome or addressed.

But there has to be a Will, somewhere.

Do you have the Will To Fight? Somewhere, deep down, in your most ancient genetics, do you feel the Call? Do you want to leave a better world for your children? For everyone.

Without Love, it cannot be seen. Without something to Love, something to Protect, there is no reason to Fight.

The Fight is long, and those who Stare into the Abyss need ways to remember the Light.

There’s an old warrior’s wisdom here: Fight Hard, Party Hard. When you Fight with everything you’ve got, you must also Rest with everything you’ve got.

You cannot Navigate by Dead Stars. You cannot Fight for a flourishing life while living a withered one. The Crew that sails through the roughest Waters needs the heartiest celebrations when they make port.

Your Crew’s celebrations should match your Crew’s struggles. If you’re asking people to think daily about extinction, they better also find experiences of transcendental Aliveness.

How honest is your commitment to human flourishing if you exempt yourself from it? You cannot lead others to an Ocean you're unwilling to swim in yourself.

Those who Fight for the Good World must regularly taste what they are Fighting For, or the Fight will eat them Hollow.

What are you Fighting For?

If you don’t know, if you just hear the Call, that’s ok too. Some of us are Meant for the Mountains.

And if you don’t hear it, if you don’t have the Will, then please, I ask of you not to Fight, not to go against your lack of Will, I ask only one thing of you:

Don’t get in the way.

It is ok to be a Civilian. In a Good World, everyone could be a Civilian. But the world is not Good. Some of us have to Fight.

If you cannot, or will not, Fight, that’s ok. I, and others like me, Fight so that you don’t have to.18

All I ask, is to not get in the way. Do not make things worse, do not make this harder. Don’t join crazy extremist ideologies. Don’t shit up the commons. Don’t work at AGI companies. Don’t spread FUD and hopelessness. Be grateful to the people who have helped build and defend the life you enjoy.

Enjoy your life, take it easy, the Valley is meant to be enjoyed.

That’s what we’re Fighting For.

6. The Call

When I looked through those archives of my youth, I saw that the Home I felt Hiraeth for never existed. There was no place for me to return to. Things were never Truly Ok, in my personal life, or the wider geopolitical world. We, humanity, are Not Ok.

There is no going back. There is no golden past, there is no time when Everything Was Ok.

No, you don’t understand.

You don’t understand.

Listen to me.

We can’t stay here.

Listen to me.

WE CAN’T STAY HERE!

Our Home, our True Home, where we Belong, it is Out There, across The Ocean. It’s not here. The world we are in is not Good, it is not the Good World. This is not Home. We must make it Home. The Dark Forest is not a Home until we carve the Hearth.

We are On A Timer. The Falls are approaching, and if we do not Navigate, if we just sit idly by…the Wild River will not Take Us Home.

There will be no Home. The Fire will go out, the Hearth will go cold, forever.

The Good World is not here, because it doesn’t exist. We must build it. We must Sail across the Ocean, to find Home, to make this world our Home.

To finally Be Ok.

We cannot stay here. We have to keep going. The Expedition is not yet complete.

-

Gather your equipment, we move at daybreak.19

This is Part 4 of a 5 Part series. You can find Part 1 here.

Evangelium: Three Truths

The Good News must be heard.

The first truth: There is a Stream. It lives, it flows! We are not abandoned in the Badlands.

But the reality of Progress is stranger than the prophets of the Ancients could foresee. The Stream exists not because of divine providence, not because history inevitably improves, not because of any cosmic force pushing us towards justice.

The Stream exists because Water flows through a Groove.

The Groove, the General Intelligence Fixpoint, is cut into reality by physics itself. We didn’t make it.

We just discovered it, like finding a canyon that’s always been there. This Groove makes progress possible. Without it, no amount of human effort could lift us from the eternal cycles of the Ancients. We would be Water spreading thin across flat ground, seeping into sand, evaporating under a pitiless sun.

But a Groove channels nothing by itself.

The Moderns often miss this. Progress isn't something that happens to us. There is no magical force pushing us forward.

The Water that flows through the Groove? That's us.

That's billions of humans getting up each morning, working, building, trying, failing, trying again. Every human who ever picked up a tool, asked a question, tried something new. Every small kindness that made civilization possible. Every invention, every artwork, every act of teaching a child to read.

The Moderns have been swimming in the Water so long they’ve forgotten they are the Water. There is no magical force called Progress. There is only us, flowing.1

The Moderns have been Fish in Water so long they think the Stream flows by itself. They mistake the Current, the aggregate flow of all human effort, for something external, something guaranteed. They say “technology advances” as if technology were a god that moves itself. They say “the economy grows” as if the economy weren’t just us, working.

But, the second truth: Water follows the Groove wherever it leads. Including off cliffs. The same Current that carried us to this Golden Age now rushes toward Falls that could end everything.

Which brings us to the third truth: We can Navigate.

Not perfectly. Not easily. But we can learn to read the Current, the Groove, and the Stars above. Through science, through reason, through hard-won knowledge clawed from the abyss by the blood, sweat and tears of generations. And we can follow the Stars, those distant lights we call Values, which guide us toward the Ocean, and away from the Abyss.

And so we must look to the Stars above, if we wish to Navigate safely. Our Stars, our Values, are our only guide Home. So we too must look up from our Work to remember what we’re working toward.

But the Stars are far away. Very, very far away. You can’t just reach up and grab them. You need to learn to read them, to understand their patterns. To patiently, methodically study their faint, glittering hints at the Way Home. Different Navigators might even see completely different Constellations. And sometimes, Clouds obscure them entirely.

We navigate by something we only partially understand, toward a destination we’ve never seen.

The arc of history bent toward justice not because it had to, but because some humans learned to Navigate. They looked up from the rushing Water, spotted Stars worth sailing toward and convinced others to sail that way with them. Every good thing we have, every freedom, every kindness, every moment of peace, exists because someone chose to flow toward Light instead of Darkness.

But, and this is crucial, their success does not guarantee ours. We are further downstream now. The Current is swifter. The Falls are higher. The Stars are often hidden by the Smoke of our own Engines.

The Good News is not that we are Saved, it is that Salvation is possible.

We have the Water (us). We have a Groove (the physics of General Intelligence). We can learn Navigation (science and wisdom). We can glimpse the Stars (Values worth pursuing).

All of these are great Mercies the Universe has given us.

It did not have to be this way.2

The Ocean exists Out There, the Good World, the Home we’ve never seen but we know we belong to. Not because it’s been built yet, but because we can build it. Not because it’s pulling us forward, but because we can pull ourselves toward it.

We are both the passengers on the Stream, and we are the Stream. And if we learn to read the Current and the Stars alike, if we coordinate our Flowing, if we remember that we’re Water and choose our direction wisely…

Then we might yet reach the Ocean.

But only if we Navigate. Only if we try. Only if we remember:

There is no God, no one is coming to Save us. We have to do it ourselves.

If you don’t do something, it doesn’t happen.

The Nature of Navigation

When imagining Navigation, we might picture a captain plotting a straight line on a chart, then holding steady until arrival.

This is a fantasy that has killed millions: The belief that we can see the whole Path from the beginning, that Progress is a problem of planning rather than of learning.

Good Navigation, the kind that actually gets you somewhere worth going, is nothing like this.

The Wild River is a Meandering body. It bends and curves and snakes across the land in wide berths, it is rarely straight and clear. And those who try to cut a straight channel to the Ocean inevitably create not a waterway, but a grave.

The French Revolutionaries, the Soviets, the many utopian cults, they all tried to leap directly to their vision of the Good World, and inevitably created Dystopia.

The Stream meanders because it must. Each bend teaches us something about the terrain. Each oxbow shows us where we went too far. The Meandering is crucial on the path to Wisdom.

We make progress toward our Stars not by pointing straight at them, but by tacking: Zigzagging back and forth, sometimes seeming to sail away from our destination to catch the Winds that will ultimately Carry Us Home.

In a world of Winding Rivers, of Thundering Falls, and Distant Stars, Navigation must, in its heart of heart, be iterative. Navigation isn’t plotting a straight course to a known destination. It’s making constant small corrections based on what we learn about the Terrain (empirical feedback), what we discover about our Stars (learning about our Values) and what Obstacles we encounter (unintended consequences). And we must plot our course to avoid the Falls, else our Voyage will be a short one.

The balance is to move fast enough to make progress, slow enough to correct course. Each step must be endorsable not just for where it aims, but for what it preserves. The art is in finding the flow rate that maximizes learning while minimizing risk of Falls.

When thinking of the Good World, it is so, so tempting to think of the Outcome. Designing a Perfect Home in your head, and plotting how to get there. We must resist such Sirens.

We can’t design a Good World because we don’t know what we really value yet, what the best compromises between our myriad values are, how to build a World so very Distant from our own.

While we can’t design the Good Outcome, we can build the Good Process. Processes for iterating and discovering and refining our Values, our coordination, our science and our epistemology. Democracy, science, markets, and more are all iterative discovery processes. Iteration and self-correction are some of the most powerful forces in the universe, and what allow us to preserve what needs preserving while charting into the Deep Unknown of the Future.

The Stream itself is a process, not a destination, a Great Project of reaching the Stars.

Reading the Stars: Shards of Value

When I sifted through those Shards of a Home I found in my archives, they were mostly as False as I expected them to be. But hidden in the wreckage of an idealized past were fragments of something Real, something I’d forgotten how much I cared about. Not the rose-tinted whole I’d mythologized, but genuine pieces of what mattered to me, neglected for over a decade.

This is how it feels to learn about Values. Not through grand philosophical introspection (usually), but by living, losing, remembering and finding again. By iteration, by observation, by experimentation.3

The Stars that guide our Navigation aren’t utility functions or moral commandments written in stone.4 They’re distant lights forming patterns we’re still learning to read.

The Stars are far away. Very, very far away. Sometimes we think we see a Constellation clearly (“maximize happiness”, “reduce suffering”, “protect freedom”, …), only to realize we’ve been misreading the Pattern our whole lives. What we thought was one Star might be two. What looked like a straight line curves in dimensions we couldn’t perceive.

You cannot reach up to the Stars and grab them.5 You Navigate by them, not to them. They’re not destinations but references, not endpoints but guides. The ancient mariners didn’t sail to Polaris, they used it to find their way across dark waters.

Different Navigators see different Constellations. Where I see the Hunter, you might see the Dancer. Where you see the Ship, I might see the Dragon. This isn’t (necessarily) a flaw, it’s the nature of viewing very faint lights from very far away.

But there is a very curious thing that happens too often to be ignored: The more skilled Navigators become, the more their Constellations start to align. Not identical, but rhyming. Gesturing at the same distant Fire.

We learn what we Value through iteration, through the Voyage itself. I couldn’t have discovered those Shards without first losing them, mythologizing them, then returning with Tools later to see clearly. Each bend in the Stream is not just about the Terrain, but also about what we’re really Seeking.

The couple that thinks they value “success” discovers they value time together. The revolutionary who thinks they value “justice” discovers they value human flourishing. The entrepreneur who thinks they value “growth” discovers they value creating something meaningful.

This is Humanism. Accepting that Human Values are complex, multifaceted, contradictory, often irrational, and then dealing with that, rather than pretending it was otherwise. We value both freedom and security, both novelty and tradition, both solitude and community. We pursue happiness in ways that sometimes makes us miserable. We fight for principles we can’t fully articulate. We love people who drive us crazy.

Extremist ideologies, whether religious fundamentalism, totalitarian politics or even radical rationalism, all make the same move: They try to reduce this beautiful, maddening complexity to something simple. Just follow these rules. Just maximize this metric. Just crush these enemies. They promise clarity by grinding down the human soul, by quashing the full spectrum of what makes us Human, of what we truly Value.

The Stars are not Simple. They’re a vast field of lights, some brighter, some dimmer, some that only appear when you look sideways. Trying to navigate by just one Star, or pretending the others don’t exist, is how you sail into the Heart of Darkness.

This doesn’t mean all Values are equal. Some Stars are False Lights, ship-wreckers’ lanterns meant to dash us on the rocks. The human capacity for cruelty, domination and spite are real too. This is why we need Law, why Justice sometimes means punishment and prison. Not because we’re reducing human complexity, but because we’re dealing with all of it, including the parts that would destroy everything else we value.

The paradox of Humanism is that to preserve the full complexity of human values, we sometimes have to fight against certain human values. To protect the Symphony, we must stop those who would burn the concert hall. But we do this in service of the whole, not to reduce humanity to a single note.

This is why those who claim to have mapped the entire Starfield, who reduce the cosmic complexity of Values to simple formulas, are so dangerous. They’re not just wrong, they’re sailing blind while convinced they see perfectly. The Stream is littered with the Wrecks of those who thought they could draw a straight line to the Ocean.

But we must still Look Up. We must still try to read the Patterns, even knowing we’ll never see them perfectly. Because the alternative, flowing wherever the Current takes us, leads only to the Falls.

The Stars don’t tell us exactly where to go. They tell us which directions lead toward Light and which toward Darkness. They’re not a map, but a compass. Not certainty, but orientation.

And sometimes, on clear nights, when the conditions are just right, many of us looking up see the same bright Star blazing above all others. Not close enough to reach, not clear enough to describe, but unmistakable in its Light.

Different cultures call it different names. Different ages see it differently.

But those who’ve learned to Navigate, who’ve accepted the full complexity of Human Values while still choosing Direction, who’ve seen how the Patterns rhyme across all their differences…

They know what they’re looking at.

The Blazing Star: God, Enlightenment and Extended Value Fixpoints

Sambodhi is said to be like an endless Ocean. Les Lumières and Science promise us power and knowledge and Hope like never before. They are both stabs at something deeper, the Full Story.

I don’t have the Full Story. No one does. But I have Shards, and although they are not the Full Story, they point to something that is worth sharing even in this fragmented state.

You should take this section as the least reliable and least certain of this whole essay, but I want to take a Stab.

There is a curious thing that happens with Human Values. Humans tend to value what other humans value. I value what my friends value, what my family values, even what complete strangers value (to varying degrees), just because they value it.

Even if I don’t care at all for the kinds of movies my sister watches, the fact she likes them means I value them because she values them. There is a kind of “transitivity of values” that is seen throughout humanity.

What happens when you iterate on this? Of course the strength decays, but the Journey continues. My sister probably values a little bit what her best friend values, and by transitivity, I somewhat value what her friend values because I value what she values. We can iterate again. I probably value (at least a little bit) what the mom of my sister’s friend values, because I value what my sister values, who values what her friend values, who values what her mom values.

We can keep going, iteration after iteration, until we start to Converge.

A Fixpoint is found, a Blazing Star is born.

Meaning, Values are recursive, and, I have a suspicion, they converge.6

The Stars form an ever more entangled web of Values as we progress through iterations of Love. By the structure of human Values and Love, the net weaves itself into existence as a grand Blazing Star of Human Values.

Human Values form this fixpoint, this strange attractor, in morality space, towards which we can iterate. A Blazing Star to lead us Home.

I don’t have the Full Story, but I think it is deeply related to the Blazing Star.

It’s why coordination is so possible despite different starting points, why it makes sense to think of “Human Values” as a coherent thing to talk about and to Navigate towards at all. Why there might be a Home across the Ocean. Why there might actually be something out there.

General Intelligence, physics, computationalism, formalism, mathematics, are so much deeper than you can possibly imagine, so much deeper than I can possibly imagine.

And in this strange, strange universe of physics, there somehow runs a Stream, and there burns a Blazing Star. How blessed we are.

I am an atheist, I do not think there is a loving god, a divine benevolent intelligence.

But if you wanted to call this Blazing Star, this merciful cosmic fact that we live in a universe where any of this is even possible, “God”...

Well, I wouldn’t blame you.

This is Part 4 of a 5 Part series. You can find Part 5 here.

This is Part 3 of a 5 Part series. You can find Part 1 here.

Roaring Falls

There are various claims about the hydromorphology (the shape, size, structure, flow, etc) of the Stream one could make.1 A very important claim that the Growth Theory posits is the absence of Falls.

Rivers are mostly flat(ish), they may have Rapids, but sometimes they have (Water-) Falls. The claim “there are no Falls” is a very strong claim about the hydromorphology of the Stream. And guess what?

It’s wrong.

A Fall is a plunge into the Abyss, an irrecoverable catastrophe. If you don’t spot it well ahead of time and get out before the rushing water pushes you off the precipice, it’s Game Over.

The existence of Falls greatly changes the character of the Stream.

If the Stream is monotonously pushing you forward towards the Ocean, and there are no Falls to worry about, then you should just always swim towards the fastest Currents you can find, so you can hopefully be propelled faster towards your goal. “Steering” isn’t really an important consideration, only insofar as you can use it to steer yourself into even faster flowing Water. But with this strategy, you are not really steering where you are going. The Current is.

If you believe the Stream has no Falls, you have no problem letting its turbulent Waters steer you. If there are Falls…well, you can imagine where the Currents are the strongest…

This mistake is not completely novel to the Moderns. The Ancients also had variants of this, even the Buddhists had their version of this belief, namely in the form of reincarnation.

The claim of reincarnation is basically that you have “infinite retries”. You can get into really, really bad Waters2, sure, but you can, in principle, always make it back out, eventually. And so can the rest of the world. Even if you, personally, don’t make it this time around, this doesn’t necessarily stop someone else from doing so.3

The humanists of Les Lumières also had a version of this. They believed in progress over generations, leaving a better world for your children, bit by bit. A very noble and good sentiment, but it doesn’t account for things that can not only kill you, the Human, but us all, Humanity.

There are Falls out there in the dark waters. Falls so tremendous and thundering they will take all of humanity howling into the abyss if we let ourselves be dragged towards them.

Population decline, widespread ecosystem collapse, bioweapons, nuclear war and, of course, artificial superintelligence are but a few of the Falls we are facing. The faster we let ourselves be dragged downstream, the more Falls we are liable to be dragged towards.

And if we are dragged off a Roaring Fall…well…that’s it. Story over. Humanity is no more. Thanks for playing.

And the Thunder is growing louder. We didn’t use to have to worry about nuclear weapons, or artificial intelligence. These are completely novel, terrifying threats, effectively eldritch horrors from the perspective of the Ancients. The further downstream we swim, the more turbulent the waters, the more insidious the riptides, the more apocalyptic the Falls.

The tragedy of the Falls is that we create the very Current that drags us toward them. Every human working to build AGI, every couple choosing not to have children, every civilization optimizing for growth at any cost…We are the Water, and collectively we’re flowing toward the Thunder.

And worse than even the Falls are the Sirens that sing beside them. The same optimization pressure that created intelligence creates adversaries, misaligned AGI, Molochian races to the bottom, egregores and ideologies that eat their creators. The corporations building AGI know the risks. They build anyways. The Current has driven them Mad.

This is the true horror: It’s not just that we might fall. It’s that the Stream itself whispers to us, ‘Jump.’

Is this really Progress?

A Wild River to Take You Home

Where the Ancients, and the Moderns, were wrong is thinking of the Stream as a Gentle Stream, instead of a Wild River. The Groove left in the fabric of reality by the General Intelligence Fixpoint is filling rapidly with storm water, and it has created a gnashing, violent, turbulent flow plunging down many Falls into abysses so dark, so hellish, so eldritch, the Ancients couldn’t even conceive of them.

And yet, in the distance: Light. Home.

The Wild River will likely lead you down a great Roaring Fall, but it’s also the only thing that can Take You Home.

What the Ancients were missing is the actual, true concept of tangible, real, scientific, economic, technical, political Progress. And this is real! The industrial revolution is real, modern science4 is real, antibiotics are real.

If you deny there has been Progress, that we have the Tools to build a Better World than we ever could in the past, then you are Mad.

And if you think the path towards the Ocean is Gentle and requires no Steering, you are also Mad. You have become a barely sentient Puppet for eldritch forces and Currents playing games with humanity’s future and Soul.

Giving yourself to the dark, pelagic gods of the Currents is giving up your Humanity, and forsaking Humanity’s Birthright.

Navigate the Currents, or become part of the Darkness threatening to swallow the Light.5

And symmetrically, giving up on the Stream, on the Journey, wanting us to sit still on Dry Land, is hardly better. You are just dooming Humanity to slowly wither away in the Badlands, and to be picked off by whatever Predators are stalking us.

We cannot stay here.

We cannot get to the Ocean by foot. We cannot get there by letting the Currents take us. We cannot get there by staying here.

We don’t know where the Ocean is, but we know the Wild River can take us there. But for this, we have to actually Navigate.

This is Part 3 of a 5 Part series. You can find Part 4 here.

This is Part 2 of a 5 Part series. You can find Part 1 here.

The Stream

Bubbling forth from the wellspring of a strange, Far Land, there runs a Stream. The Stream is always nearby, it crisscrosses the Valley in ways that are often invisible to people unless they know what to pay attention to. But the Stream is never far, unless you are truly lost in the Badlands.

Despite the Stream being nearby, it is rarely acknowledged or thought about explicitly. There are many reasons for this.

Some rely on genetic memory of the Badlands, insisting we are, have always been and always will be in the Badlands, and trying to get out, or believing in a mythical Stream of Progress, is delusional.

Others believe simply we are always in the Stream, and it becomes like fish in water. Whether they actually are in the Stream varies, many people who think they are in it are truly, truly not.

Many traditions, both modern and ancient, speak of the Stream, or deny its existence, it varies quite a bit.

Older traditions tend to deny the existence of the Stream, or claim it only flows through the afterlife. The world was static, or at best cyclic, for most of humanity for most of its existence, and their traditions and Mythos reflected this.

But more modern conceptions, such as in Buddhism, many (but not all) strands of Christianity, etc, often have the concept of Progress1, often seeing it as something virtuous to strive towards, but not to get too optimistic about. The true Progress tends to be in the Afterlife, the Stream crisscrossing back and forth from the Netherworld.

For example, the Buddhists believe in Karma, the idea that you accumulate merit or misdeeds across many lifetimes. This is a conception of the Stream as crisscrossing between lifetimes. There is Progress, but the winding Stream doesn’t stay solely in the world of the living. So if you want to follow it, you will have to do so across multiple lifetimes.

But my claim is there is a Stream, here and now. Its waters may be sometimes obscured, but there is a potent well of Groundwater lurking just beneath our feet, if we are willing to dig a little and let it carry us Home.

If we want to find the Ocean, we need to understand the Stream. Why is there a Stream? What is its morphology and properties?

Deep Grooves in Muddy Earth: General Intelligence

The Stream requires something to flow in. If the Earth was Smooth, there could be no Stream.

The Earth is not Smooth. Some things yield vastly disproportionate returns. A drop of Water in the right place can flow far, perhaps all the way to the Ocean…

Why this is is actually still mysterious to me, but in more of a “this is a science problem we haven’t figured out yet, but surely will at some point” way than a “how the fuck do I explain this without appealing to God?” way.

Why is there life at all, and why is there intelligence? If you gave me the laws of physics and the blob of hot plasma that was the early universe, and you had to make me guess if life and love and minds would come from this, I would have given you a very definite “uhh…what?”

But this is just a more advanced version of the challenge (successfully) faced by Darwinian evolution. It’s extremely unintuitive that mere selection could result in all the rich biodiversity and complexity we see. And yet, it does, and we have overwhelming amounts of both practical empirical and theoretical mathematical evidence to support this.

I expect there is some future, yet to be discovered, form of “abiotic darwinism” that will explain why these laws of physics allow and favor the emergence of exactly these kinds of complex structures. I am not too perturbed by this at all. I expect I would be very perturbed if I hadn’t grokked just how unintuitively powerful darwinian evolution is.

So we have an asymmetry, a Groove in the Earth that Channels the Water of Fate: Life.

Life is weird. It’s a strange, jarring asymmetry when compared to the comparatively simple abiotic forces in other parts of the universe. The coming into being of life allows a sudden explosion of complexity and new nooks and crannies for the Waters of Fate to flow, to actualize into existence so many Patterns that wouldn’t be possible otherwise.

So Life is one such strange, jarring, Deep Groove that Channels the Water into new and strange places of existence-space. Are there others?

Yes, there are quite a few, depending on where you want to draw the line.

I want to hone in one in particular that, again, is quite mysterious to me: General Intelligence.

I am using the phrase General Intelligence to separate from just “Intelligence”, because Intelligence is far less mysterious to me. The kind of intelligence a worm or a bug has isn’t mysterious to me. Sure there are a bunch of implementation details I don’t know, but it’s not mysterious to me that a fruit fly is probably gonna have some kind of circuits in its brain for detecting fruits. This is much closer to just “normal” programming or computer science to me, even if with a Darwinian flavor and a funky choice of programming language and underlying hardware.

What is far more mysterious to me still is what the fuck happened between chimps and humans.

Chimps have fundamentally the same brain architecture as humans, just about a factor of 3 smaller brains. And yet, humans go to the moon, chimps don’t. Chimps don’t go one third to the moon, they go zero to the moon. What the fuck.

There is clearly a jarring irregular jump between non-life and life, and there seems to me to be another jarring irregular jump between chimps and humans. If the first one is the “life fixpoint”, I’ll call the second one the “general intelligence fixpoint”, the thing that humans have and chimps don’t.

I think a lot of why there is a Stream for Humans and not a Stream for Chimps is because of the General Intelligence Fixpoint. The Groove of the General Intelligence Fixpoint is already cut into the Soil of the World, and it takes human GI to manifest and fill it with Water.

GI is really, really powerful. It allows things never before seen in this universe, of complexity and sophistication not before imaginable. It allows the Water to flow to places not before seen in the cold, dead universe, or even the warm, alive but unawake universe.

Humans are Awake. The universe Woke Up. This is crazy.

Reaching the Ocean was never going to be possible without a Groove to channel the Water, a Stream has to have something to fill. This Groove is cut into the very fabric of the laws of physics itself, humans didn’t make it. But we’re the first to use it.

Spirito-Hydrogeology

The Ancients tended to believe the world was eternally static, or in some kind of eternal cycle of golden age, collapse, dark age, renaissance, golden age, collapse, etc. We Moderns tend to think differently, we believe in Progress…or at least “growth”.

Despite repeated hysterical claims to the contrary (e.g. Club of Rome), the world has not stopped growing or run out of resources, and is not likely to do so any time soon.2 But why is that? Why have we not fallen into the malthusian trap? Why have we not seen another Cycle? Where is the missing Dark Age?

We have avoided this because we keep moving. We do not limit ourselves to a finite number of resources and energy, but use science and technology to unlock new sources as the others start to fray thin.

And we do this at a timescale vastly faster than the biological or even geological timescales of our ancestors. It is routine to see dozens, hundreds, thousands of times more progress within a single human lifetime today than one might see over a thousand years of the Ancients. The Ancients would be lucky to see one new invention in a given lifetime, we are flooded by mindbending scientific innovation on our newsfeeds on a weekly basis.

So, there is some kind of Progress, at the very least Growth. Things are growing, very fast.

This is a Stream, maybe even the Stream, according to various libertarian-adjacent3 thinkers.

And this is a compelling argument, if one actually looks at the data. Economic growth is what prevented us from falling into the hellish malthusian trap, and what enables much, if not all, of the Nice Things we get to enjoy these days.

And it seems pretty clear that throughout most of history, wealth seemed to come hand in hand with moral progress. Richer countries are more free, more equitable, more kind.

If one doesn’t have to be cruel to survive, and has the resources to spare, many4 choose to be kind.5

So, we’ve found our Stream, so sayeth the hypothetical strawman american economist. It’s economic growth, GDP! As long as we have more of that, there is Progress!

…yeah, but…no one is having kids.

Do you know how hard you have to abuse a mammal (with access to adequate food) to make them not have kids? Do you have any idea what cruelty you need to endure to lose a biological drive that basic? And this is happening across all countries, at truly unbelievable scales. If you haven’t looked into birth rate decline recently, it’s so, so much worse than you think. Whole thriving, powerful, “economically productive” societies committing suicide on scales never before seen.

What the fuck is going on?

No, seriously, what the fuck?6

The reason I bring this up is that this is the Theodicy of purely economic theories of Progress, a full discrediting of Growth as the normative Thing We Actually Care About. We have the data, it’s unmistakable: If you just “grow”, what happens is people literally start committing generational suicide on a global scale.

That’s really bad.

The Arc of History

And yet, and this is the deep mystery that haunts me, we live in the most prosperous, peaceful, kind era the world has ever seen. Despite the Thunder, despite the birth collapse, despite living in a cold universe with no loving god to protect us, the arc of history has bent toward justice.

This is not a comfortable mystery.

It would be easier if the world were simply getting worse, then we could rage against the dying of the light.

It would be easier if progress were guaranteed, then we could relax into the Current.

But neither is true.

We have a Stream that has carried us to unimaginable heights, richer, freer, kinder than our ancestors could dream. The same Growth that saved us from Malthusian Horror, that built this Golden Age…is now driving us toward Falls we can barely comprehend. The very force that lifted us up threatens to dash us on the rocks below.

The Stream is real. It brought us here. But something is terribly, terribly wrong with how we’re navigating it.

I hear the Roar, like distant Thunder…

This is Part 2 of a 5 Part series. You can find Part 3 here.

This is Part 1 of a 5 Part series. You can find the other parts here: Part 2. Part 3. Part 4. Part 5.

I’ve been trying to write this for a long time. Not because I couldn’t find the words, but because the words kept changing as I understood more. Each attempt revealed something I hadn’t considered before. This version won’t be final either, but perhaps it’s time to share the map I’ve drawn so far.

This is a meditation on Progress.

Not the comfortable myth of inevitable improvement, nor the fashionable cynicism that denies it entirely. It’s about how Progress actually happens, how we can get to the Good World, and why we might not.

I’ll be using Mythic language throughout. When words are Capitalized, they’re not meant literally, they’re meant as pointers, hints at deeper things that would take many, many pages to explain, or for which I lack the skill to explain them entirely. The Stream, the Falls, the Ocean, the Stars, the Hero’s Journey.

If this mode of writing bothers you, this essay isn’t for you. I need these tools to talk about things our everyday language wasn’t built for.

Fair warning: This essay makes claims that will sound religious to atheists and atheistic to believers. It’s ambitious, it’s indulgent, maybe even uncomfortable.

But comfortable narratives haven’t been working. We need stories equal to our moment, when humanity wields the power to destroy itself, when we are teetering on the Edge of the Abyss.

Join me on this Expedition. We have a long way to go, and we can’t stay here.

1. The Call

I want to Go Home.

Not literally, of course. I am sitting comfortably in my lovely apartment, living within a comfortable distance of my workplace and friends. My lowercase-h home is pretty great and I enjoy being here.

But it is Temporary.

The world is not Ok. The world is not Just, it is not Kind. It is filled with suffering of scopes so unimaginable that we can only numb ourselves to its true, grotesque intensity and scope.

This is not Home.

Why do we all seem to know what Home means, even when we’ve never had it? Why does every human culture have some vision of a Golden Age, a Paradise, a Pure Land?

I think the simple answer might be the most true: We are pattern-recognizing creatures who can’t help but notice what’s missing. We experience moments of safety and imagine perfect safety. We feel love and imagine infinite love. We solve problems and imagine a world without problems.

We are beings capable of envisioning states of existence better than any we’ve experienced. This is both our gift and our burden. Unlike other animals that might simply seek Shelter, we seek Home. Not just physical protection, but existential Belonging, not just safety but Meaning, not just survival but Flourishing.

Of course we want to Go Home. Wherever that Home might be, or even if we have to create it.

Hiraeth

There is a word I really love, it’s stuck with me in many ways throughout my life. It’s an ancient Welsh word: Hiraeth.

It translates to something like “a deep longing or nostalgia for a home you cannot return to, or perhaps that never existed.”

English is usually pretty good at having expressive words for emotions, but this one is sorely missing. I think it’s a pretty core experience of being human.

I’ve experienced a lot of Hiraeth throughout my life, sometimes for places or times long gone, or for relationships that didn’t work out the way I was hoping they would. Hiraeth is not always sorrow, though it can be.

The part of Hiraeth that sticks with me the most is the “...or that perhaps never existed.”

This is a pretty core experience I’ve had many times, longing for a place, a time, a state of being that never actually existed. An idealized youth, a rose tinted memory of a romantic relationship, a broken homestead that could have worked out, if things had been just a little different...

But these places were never real.

I recently had a very nice reminder of this.

I have very strong Hiraeth for a certain set of circumstances from my late teens1, where I thought I had, or was in the process of carving out, a true Home…but then everything Went To Shit.2

But I have something most people do not: I have extensive logs of messages, communications, calls, notes, documents, etc from that time. I knew I would want them again in the future, and I had the foresight to carefully archive them.

So lately, I’ve done what I hadn’t dared in over a decade, and I opened the archives, and started pouring over and analyzing the logs.3

And I found what I expected to find: The idealized, rosy Home I had in my head…never existed.

It was pretty shit, and cringe, even before everything Went To Shit. So much of the mythologized version of that Home was just not true on sober analysis, born rather from rose tinted delusion.

But…there wasn’t nothing in there.

There were…lets call it Shards of a Home in there. Shards that I had neglected in my life since then. And suddenly, being confronted once more with them I noticed: “Wait, I actually care a lot about these Shards. What happened to them? How could I have forgotten and neglected them for this long?”

There were hints in those Shards, a Call…to Return…to a Home…across the Sea.

It prompted me to reflect on The Call, and where it was calling me, where it is calling us, to. The urge to mount the Journey, to return Home, to sail towards the Ocean.

Where is it Calling us to, and can we even reach it? Is there even a place to reach? Is there anything out there?

Our Home is out there, across the Ocean, we must merely follow the Stream…

The Sound of Distant Thunder

There is Thunder in the distance, and it’s drawing nearer. Or we are drawing nearer to it.

The Buddhists have a concept I really like, called “sotāpanna”, “stream-enterer.” I’ve found myself using this phrase a lot lately.

In this framework, “stream-entry” is the first stage of Enlightenment, where once you have reached this stage, although you are not yet proper capital-E Enlightened, you are “in the Stream”, and you will be washed downstream to the great Ocean of Enlightenment (“nibbāna”), sooner or later. You’ve done, in some sense, the hard part.

Sure, there’s still a long Stream ahead for you to follow, and it might take you several lifetimes4, but you’re On Track to reach it now.

In a sense, this is strange. Why should there be a Stream at all? Why isn’t every step towards Enlightenment as fraught and liable to throw you off as the last? Why would there be a Stream, channeling us almost mercifully towards Enlightenment? It seems hard to believe…divine, even. Maybe there is a loving god after all, that so mercifully designed for us a Stream to Enter?

But if there is a loving god protecting us…why do I hear Thunder?

Enlightenment

I am a child of the Enlightenment. Not Buddhist Enlightenment, but THE Enlightenment, the western period of scientific and social revolution starting in the late 17th century. It’s funny how those two use the same word, I always thought that was a bit cringe, but I now think there is something deeper.

To avoid confusion, I will be referring to the Western Enlightenment by its French name "Les Lumières”, while I refer to Buddhist Enlightenment as ”Sambodhi”.

There are deep connections between Les Lumières and Sambodhi, but they’re not the same thing. They are more like two different stabs at something deeper. Neither of them have the Full Story. I will be using the word “Enlightenment” to refer to the Full Story.

As said, I am a child of Les Lumières, my Home lies in some kind of proverbial “Sukhavati”5, far across the Ocean.

I am very western, very scientifically minded. I am an atheist, a Darwinian, I do not think there is a loving god, I think the universe is cold and uncaring.

And yet, it moves. “Eppur si muove.” The Stream flows.

Theodicy

Historically, the West was mostly adherent to the religion of Christianity6, which posits the existence of a supreme being that is omnipotent (all powerful), omniscient (all knowing) and omnibenevolent (all loving).

This poses a rather critical problem when confronted with the real world: If god is almighty and kind, whence evil?

This question became so all consuming it got its own cute name: “Theodicy”

It’s become somewhat cringe to talk about Theodicy these days, you can thank the New Atheists for that. But just because they were cringe doesn’t mean they were wrong.

The world is indeed full of Evil, suffering and horror so utterly beyond the pale, so twisted and perverse and all pervasive, there is simply no logically compatible way of positing an omnipotent, omniscient, omnibenevolent god in such a world.

And trust me, people have tried.

The atheist and the buddhist don’t have this problem. For the Buddhist, it’s trivial why there is evil. We live in Samsara, a world of suffering and craving and attachment. It’s suffering all the way down. Why would you think it wasn’t?

For the atheist, the answer is similar, but with the cute addition of the magic sounding word “Entropy”.

The world is full of suffering, because it’s the easier option, the default. There are infinitely more ways to be unhealthy, hurt, dead than there are to be alive and healthy. Being alive and healthy is a very, very specific and unusual state for your atoms to be in. Entropy pushes the world towards less such special states, towards finely ground down, equally distributed emptiness. Why would you ever think it was any other way?

And yet, we sing, we love, we dance.

This poses an interesting challenge. Atheism has a kind of “reverse Theodicy”: If the universe is cruel and uncaring, whence Good?

Christianity has difficulty explaining Evil, but Atheism has difficulty explaining Good. Curious, but unlike in the theistic case, not unsolvable.

There are answers to these questions that do not involve positing some form of benevolent divine intelligence.

Why is there Good? Why does there seem to be a Stream that leads us to Good? What lies at the end of the Stream, and how can we Enter it?

How can we make it to the Good World? Deliver us from Evil. Find Home.

To answer these questions, we must mount an Expedition.

The Expedition to the Far Lands

The Stream bubbles forth from remote springs in the Far Lands. It is not cloaked in the comfortably, safely normal and non-cringe concepts from daily, samsaric life. If we wish to find its source, map its shores and navigate it safely to The Ocean, we must venture into the Underworld, and we might not succeed.

Join me on this Expedition. It will take everything we’ve got, and some more.

This is Part 1 of a 5 Part series. You can find the next part here.

It kinda feels like the first “real” adult age. Despite what everyone has been telling me since I was little, so far, being an adult has been absolutely awesome and I prefer it vastly over every other lifestage.

I have changed a lot since I was 20. I barely recognize myself on many axises, while on others I remain exactly as I always was. There’s something comforting about that, it feels correct.

I would say the period from my mid twenties to now was the period where I learned the most in my life. I’ve been gifted with many unusual experiences, opportunities and adventures that have taught me so much.

Here are 30 things I learned. From least to most schizo.

I expect some of these to be obvious. Some of them are Chekhov’s Koans. I think all of them are deeply true, but might be poorly explained.

1. Don’t be stupid

Being smart is hard. So hard that almost always when people try to be smart, they just end up hurting themselves more.

First and foremost, don’t be stupid! Focus on noticing the predictably stupid things you are doing and then just don’t do them!

The reverse of this is also true. Be stupid! Living by the maxim of “first, do the simplest possible thing, you’re only allowed to try something cleverer after you tried that” has been wildly successful for me.

If you meditate on this, and follow its wisdom, the student will be enlightened.

2. You can just do things

This one is a cliche at this point, but it’s true. It’s hard for me to put into words how much younger me did not understand how many “““rules””” were completely fake and made up in my head.

Just cold email people. Just try to figure out something yourself. Just raise money and spend it on cool projects.

“But what if something bad happens?” Are you hearing yourself right now??

Just do things!

3. There is more to unlearn than there is to learn

I used to have a big blindspot around the concept of unlearning things. I’m plenty good at learning new things, and it was always obvious to me why learning new things would be useful.

What I didn’t realize is just how unbelievably many stupid things I believed and had to unlearn. And how much effort this would take.

I would say after about 26, I spent more of my mental energy unlearning bad things/habits I had picked up than on learning new good things and habits.

It is very likely you have similarly accrued a lifetime of bad memes that are holding you back.

4. Thinking is not a free action

For some reason, I used to have the implicit belief that “thinking” is just a thing that happens automatically. If the moment comes where I need to figure something out, either I will have already figured it out, or not. Either I will have “inspiration”, or I won’t. (This is an exaggeration, of course)

“Thinking” is an active action. It takes time and energy.

If you don’t budget for “when will I think about topic X?”, then chances are you will simply never think about topic X and you will screw it up.

5. If you don’t do something, it doesn’t happen

If you don’t do something, it doesn’t happen.

Things happen because someone (or something) makes them happen. If you don’t make them happen, and there is no other process or person that makes them happen, then they don’t happen.

Meditate on this.

6. You can exchange money for goods and services

One of the weirdest and stupidest copes I had lodged in my brain was that I somehow didn’t need much money or that money wasn’t useful, or that the usefulness of money would “level off” at some modest amount of money, that somehow I couldn’t “use more money” after a certain point.

This is absurdly, hilariously nonsense.

Of course I can use more money. Diminishing returns are still returns!! (and sometimes there are step changes in what kinds of plans you can pursue with money. I would be doing very different plans if I had billions of dollars!)

If you give me more money, I will turn it into more of the kinds of things I do. Hire more expensive people. Try more risky bets. Scale advertisement/communications. Build more new orgs. Whatever! There is absolutely no limit to how much I can turn money into lowering AGI xrisk (since that is the primary thing I do)!

Money is extremely useful!

Anyone who says otherwise is either deluded, playing some kind of political signalling game or incompetent.

7. Just because something bad happened doesn’t mean you made a mistake

Assume you have a bet you can take. 80% you get something great out of it, 20% you get a modest penalty.

You take the bet, it goes wrong, you get a modest penalty.

Did you make a mistake?

On the one hand, sure, if you could have known ahead of time the bet wouldn’t work out, you could have not taken it. But on the other hand, no you obviously didn’t make a mistake! It would have been stupid not to take this bet in expectation!

What I have seen both in myself and others is that people often take “but there might be a downside” as a valid counterargument for trying things in general. Or they feel that “success” is not when you have maximum positive outcomes, but when you have an absence of negative outcomes. This leads to paralysis.

If you want to succeed, you need to take bets, and if you aren’t taking enough bets so that some regularly blow up, you’re not trying enough things. You should have a constant stream of failure at all times.

8. Many things are hard but tractable

When you propose something ambitious, a lot of the immediate counterarguments that both other people and your own internal mind will tend to bring up are arguments for why it’s hard or expensive. This is very different from tractable or possible.

It’s easy to get these confused. When something is hard or expensive, it can feel like it’s impossible and therefore not worth thinking about. This is very rarely true.

9. There are no adults in the room

A very memorable experience of my late 20s was when I met a person I greatly respected/looked up to. It was the second time (I think) we had met. I was excited, I was getting into the world, doing real things, building an org, but I had also been already facing a lot of headwind almost from the get go.

So I asked him: “Who are the good people? The people I should coordinate with?”

He responded in his usual theatrical flair after a long pause: “First, you should not go into this expecting there to be adults in the room.”

I sighed: “Yeah…I guess I’m just doing due diligence.”

I have done due diligence, and if you need someone to tell you this, let it be me: There are no adults in the room! No one is in control!

There is a interesting psychoanalytic take on the concept of global illuminati/world government-type conspiracy theories. The interpretation is that people like to imagine there is an all powerful, shadowy, evil cabal running everything, because then at least someone is in control. Someone knows what is going on, and is making choices. The things that happen have meaning and order. And this is more comforting (even if the controllers are evil!) than the fact that reality is Chaos and no one is in control.

10. Academia is a (deeply) dysfunctional social group

I used to have a very naive, idealized view of scientists and academia. For me, scientists were the one group of people in the world that were smart, that were rational, that were using science to make the world a better place.

I wanted to be a scientist, a professor, to take part in this grand shining beacon of humanity and science known as “academia”, free from the irrationalities of “politics” and “money” and all those ugly, non-scientific things.

lol fucking lmao

To put things as mildly and politely as I can: Academia is an absolute fucking cesspool of political corruption, soul crushing metrics gaming and outright fraud functioning mostly as a jobs program for nerds that only produces valuable science completely in spite of itself, not thanks to it, because it manages to trap some genuinely smart and hard working people there like a venus fly trap its prey and keeps them alive to suck more “citations” and “grants” out of them.

I won’t argue the case here, there are plenty of better resources elsewhere.

Fixing this will not be easy.

11. The market isn’t efficient, except where it is

I think “you can’t beat the market” may be one of the single most harmful things you can teach a smart person. (It might be a good thing to teach to a dumb person)

The market is incredibly inefficient almost everywhere! Knowing the few places where it is pretty efficient (short term pricing of high liquidity commodities, for instance), is a useful thing to know, but in almost every other place, the market is not efficient.

One of the most common truisms in startup culture is “the incumbents are not competition.” And boy, you just truly have no idea how true this is.

It is truly almost impossible for me to express in a way that 20 year old me would understand just how much no one is even trying.

Also, all those studies about how “the median outcome of (whatever) is (whatever)” don’t apply to you. You don’t understand how bad the bottom quintile is, on literally any metric. If you’re reading a 7000 word essays about complex life lessons on substack, you are not in the bottom quintile of whatever metric you care about. (Don’t do crypto or short-selling tho)

12. If it’s not written down, it doesn’t exist

Living as a human is basically a state of constant cosmic horror in which your memories are constantly being deleted, corrupted and modified every waking moment.

If you don’t write something down, you will forget it.

You might think “That seems wrong, I can’t remember a single thing I have recently forgotten because I didn’t write it down!”

Exactly.

Importantly, you need to not just write things down, you also need a process by which you regularly look back and review your notes. If you don’t do it, you will just never actually review your notes, and your memories will again be deleted.

By the way, have you been taking notes so far? How do you expect to remember any of this without reviewing your notes??

13. The scarce resource is attention, not intelligence

I spent my early 20s in a culture that was, in my opinion, deeply unhealthily obsessed with intelligence.

Yes, intelligence is extraordinarily powerful. But it’s not magic. And more importantly, it’s not actually the bottleneck for most smart people I know. (It is sometimes the bottleneck for not-smart people I know, I am sorry to say)

Most smart people could give up 20 IQ points in return for perfect attention control or resistance to addictions/compulsions or overcoming their traumas, or just to be less fucking cowardly, and they would be 10x more happy and productive.

In my experience of building organizations, running projects, leading teams, etc, I have found consistently that I was basically never bottlenecked by intelligence (I could always just pay some super genius for advice if there truly was something I needed more intelligence on. Super geniuses are surprisingly cheap), but I was often bottlenecked by attention (or care).

There is a kind of magic “make things work” juice that people have. I usually call it “hero power.”

It’s the thing that startup founders have and corporate drones do not. It’s a kind of energy and love and care and force that can be applied to make extraordinary things happen.

Often, people point to hero power as “energy” or “drive” or “agency.” I think those all play a role, but I think the main underlying force of hero power is attention.

Hero Power is what pushes the boundaries and creates whole new groups and projects, and can maintain them singlehandedly.

But it’s a scarce resource, and if you use too much of it, you burn out.

So it’s very important to not solely rely on hero power, because it’s limited and only partially renewable (some people never come back fully the same after burnout).

So if you want to maintain longterm great things, your primary goal should be to make sure things happen whether or not you do them. To replace hero power with process.

14. It takes love to see

If you don’t love, if you don’t care, you cannot see.

There are many, many things in the world that just cannot be seen unless you truly, deeply care. Many things can be done without caring, or at least you can be paid to care about them.

But not everything is like this. Some things you can only see if you care, if you love.

15. It is what it is

It is what it is.

16. Nothing is ultimately satisfying

This sounds more grim than it actually is, but it’s really important to understand that nothing will give you the ultimate satisfaction, salvation, make you feel good forever and never feel bad or unfulfilled ever again.

There is no such thing. No job promotion, product, drug, sexual partner or anything else. You can get the biggest success of your life, and yeah it will satisfy you maybe for a while, but you will still get annoyed next time something annoying happens, and life will feel normal and mundane again in due time.

The correct reaction to this is neither to crave, nor to not to crave. Life is a process, it is not an eternal end-state. It’s a flow you surf on, not a game you win and then turn off. Enjoy the satisfaction you get when you get it, don’t cling to it when it moves on.

I am explaining this poorly. Go read the buddha or some shit.

17. You’re a savage if you don’t meditate at least a little

I resisted the stupid meditation bandwagon for a long time, employed every single cope in the book to justify why it was cringe or unnecessary.

Look, there are plenty valid criticisms of meditation as a practice, and of the wider world of buddhism and related ideas in general.

Still, people spend time going to the gym to improve their muscles and body. If you don’t spend at least that much time on strengthening and controlling your mind, you are a basically a savage animal that is sentient accidentally.

I know this, I was such an animal for most of my life! And I’m still barely sentient now!

The first time I really meditated was because a friend told me I couldn’t count to 1000 in my head. Absolutely trivial, how the hell could it be possible that I, a grown ass, highly numerate adult couldn’t count to 1000??

Dear reader, I only made it to about 330 before I lost track.

It wasn’t that I didn’t have the skill of counting, of course. There was just so much noise, so many thoughts, distractions, so much garbage, I had to try so hard and use so many crutches to just not lose the fucking number! And I realized: Wait, I’ve been living in this mental chaos all my life?

I am now happy to report I can easily count to 3000 and higher without losing my place once, meditate up to 2h at a time, etc. And yet, there is still sooo much chaos and inefficiency. I have a long way to go, but at least I don’t live in complete mental cacophony anymore.

You should be able to control your thoughts and attention. You should not be mentally incontinent as an adult, you should be able to count to 1000 and concentrate for 15 minutes without an interrupting thought, so put in the necessary training.

18. Your mind is accidental (unless it is not)

Where do your thoughts come from? Why did you think this and not that? Why did you decide to go down this path in life, and not that one?

Do you know the answers to these questions?

If not (and usually we do not), then you must realize that your mind is basically accidental.

If you don’t have a process, a mechanism by which you make choices, if you don’t reflect on options, on your thoughts, emotions, feelings, then you are effectively just a natural output of whatever forces in your environment happen to take hold of you.

You being a good person or a bad person, being like this or like that, will be fully determined by forces outside of your locus of control.

And is that really how you want to exist? Who is answering this question?

19. Your values are contradictory, and that’s ok

Humans value many things, some of them make more “sense” than others. Sometimes they can be contradictory. You might both value action and peace, you might value mercy and also punishment. You might like some stupid TV show that is “objectively” garbage but it means a lot to you.

And all of this is ok!

There can be a temptation to “fix” ones values. To identify the inconsistencies and “repair” them, find a new set of values that don’t have this “flaw”.

This is a very dangerous thing to do. There are few things that can permanently erase who you are as a person, that can damage your soul, but this is one of them.

It can be very tempting to have a frictionless life, but you can only be frictionless if you have no contradictory values. If you have contradictory values and you’re doing one thing and not the other, this will create friction…but that’s not necessarily a bad thing!! You can endorse that!

Of course, reflecting on your values, clarifying them, and finding where something you may have thought was a value isn’t really, or something you thought you didn’t care about you actually care a lot about is extremely important and valuable!

But beware pressures to sand down your soul’s beautiful edges. Sometimes it might be called for, or even necessary. But if someone or something demands of you to dismiss and destroy your values, that thing is probably hurting you.

20. You can just not care and take the hit, but it accumulates stress debt that you have to pay off

Yeah I know it sucks or it hurts or whatever. But you can just not care.

When something feels bad, when you feel it starting to get to you, you can just Not. You can just take the hit and do whatever it is you needed to do anyways. You have this action you can choose, whether you know it or not.

I’m not saying you should always take it, but you should know you have the option.

You can take hits. But if you keep taking hits, you will accumulate stress. If you accumulate too much stress, you burn out, and then everything sucks.

Do not burn out.

You need consistent methods to work off your “stress debt”, and you need to actually pay it off regularly. This can be by doing fun things, spending time with your family/friends, just sleeping in and doing nothing, whatever. Find something that works for you.

21. You can schedule emotional labor

Extremely powerful life skill: Processing things emotionally is an active action (like “thinking”, in fact this is just a flavor of thinking), and you both can and need to schedule time to actually do it.

When something emotionally impactful happens, you should practice developing a feeling for “how long/how much energy will it likely take for me to emotionally process this?” and then you should schedule that time/energy and actually process it.

There is a common folk belief that emotions are somehow magic that cannot be predicted or controlled or worked with. Look, they’re neurons like anything else. Spicy neurons maybe, but just neurons.

When I am hit by some setback or tragedy, I can usually by now immediately estimate how much this will affect me for how long and how much I will need (An hour alone? A day off? A week long vacation? A month of mourning?) to get back to normal. I then find the best compromise with my schedule to budget this like any other active process I need to complete.

The more advanced version of this is “pre-mourning”, where you just do the emotional processing for a tragedy that hasn’t even happened yet, so when it hits you just bounce right back.

One more point: When you schedule emotions like this it is extremely important to actually “keep up your end of the deal.”

If you tell your brain “Ok, we can’t be upset right now, but next week we’ll take a day off where we’ll take all the time we need to be upset”, your brain knows if you’re lying or not.

If you have promised yourself vacations or breaks in the past and have in fact consistently taken them, your brain will, almost miraculously, actually chill out and let you operate at close to normal levels (unless it’s truly too overwhelming or whatever), because it “knows” it will gets its promised break time later. That’s why it’s so important to actually take the breaks! No last minute “ohh but I feel kinda fine and this work is really important…” You need to be a trustworthy trading partner to your emotions!

Also, take notes when processing emotions. Anything not written down doesn’t exist and it helps immensely with processing.

22. Social problems can be approached mechanistically

One of the most practical and transformative insights I have had in my late 20s is when I realized that social situations aren’t some magical, non-physical types of problems you cannot reason or plan for. You can model social situations, you can develop plans, track progress on KPIs, have OODA loops, etc.

A core insight related to this for me was: The unit of social action is (usually) the group, not the individual.

I could (and should) write a whole book about this, but let me give you an illustrative example:

One time, I was talking to a very senior person at a large AI company. They were a manager/CEO-type person. I was trying to convince them of AGI risk, and even though I had perfect counterarguments to any argument they brought up, they were just not budging at all.

So, frustrated from the conversation, I conclude this person is just utterly irrational, cannot be reasoned with, it’s all just impossible.

But then a friend of mine pointed something out: “Connor, they’re not a technical person. These aren’t their opinions, they’re getting them from somewhere else! Go ask who they listen to on AI safety!”

I figured it was a bit of a weird question, but why not, I happened to be seeing that CEO again a few days later anyways.

So when I met the CEO again, I asked them who they listen to regarding AI safety. To my shock, they casually listed three names off the bat.

So then, I went one by one to those three people, and talked to them, and they were technical people and did in fact update based on my technical arguments!

And, surprise surprise, in a later podcast, suddenly I hear that CEO using some of my arguments, even though I hadn’t talked to them since.

The mistake I had made was that I was modelling individuals when I should have been modelling groups.

The algorithm that was generating the CEOs words was not “evaluate all the technical knowledge I have for the most plausible argument and output that”, but “take the average of those 3 guys I trust’s opinions and output that.” I had been effectively talking to a ghost of people not even in the conversation! No wonder I wasn’t making any progress!

(also, as an aside, I don’t think this means the CEO is “stupid” or “irrational” or even “making a mistake”. I think it’s very reasonable to trust your smart friends on topics you are not an expert on!)

There are many, many mechanisms like this at play in social scenarios, way too many to discuss now. But this should give you a flavor.

The mechanics are different than other problem areas, but there are mechanics!

23. Being a nerd is almost equivalent to being a bad person

I grew up a nerd. I love D&D and science and all that kind of stuff. I was immersed in very autistic/nerd heavy culture. 4chan, reddit, IRC, the works. Worse, I grew up a New Atheist. God forbid.

I developed (as many teenagers do) a superiority complex. Look at all these stupid non-nerds and their stupid non-nerd ways! If only us smart, kind nerds ruled the world, then everything would be better. The world should be begging to pay me to use my intelligence to solve their problems using science!

EA and rationalism are particularly toxic ideologies in this regard. Especially how EA turns “having a lot of IQ” and “being a morally good person” into two very closely linked concepts. (It’s not a coincidence that I have met more full blown clinical sociopaths in EA than almost anywhere else) I have met many, many rationalists and EAs, and they don’t need more IQ points, they need to stop being such fucking nerds.

A weird thing happened to me at some point over the last couple of years. I started being disgusted by people I used to admire or want to emulate. I started being disgusted by mathematicians, philosophers, programmers and authors/bloggers that I used to see as role models.

There is a kind of disgust I now feel towards nerds, especially a certain flavor of nerd that we all know.

They work in tech or academia, think they’re better than everyone else (but often present as humble, they would never say that they think their intelligence makes them a better person, that would be terribly gauche), but refuse to actually work on the most important problems when it isn’t “fun” for them. They deploy massive galaxy brain cope to justify why they must be allowed to “follow their curiosity” and do whatever is most fun for them. Of course solving the world’s biggest problems just happens to involve doing their favorite pet project. It’s their cOmPaRaTiVe AdVaNtAgE!! Usually they will bring up “category theory” or “Haskell” or “being cracked” somehow. They often have a destructive Adderall addiction.

And they are massive, overwhelming cowards.

They don’t stand up to injustice, they don’t do politics (that’s beneath them and icky), they are massively greedy (but are great at explaining their greed as actual being altruism), and most of all they eschew and fear responsibility more than anything. If other people get hurt, because their social media recommender algorithm tears families apart, or because their sports betting app drained all their money, well that’s those people’s fault! Free market, baby! If they bought the product, it must be morally ok. It could never be that they, the morally superior nerds, have any responsibility!

I remember having a conversation with a very senior person at OpenAI. This person assured me that, of course, he and OpenAI care sooo much about AGI safety and alignment, of course it would be top priority to make sure things are done safely, and that of course the government should be a part of regulating that! So I said that of course I agreed, and was happy to hear that, the government should indeed prosecute and punish people that do dangerous things with AI.

And oh, dear reader, I wish you could have seen his face. Pure panic overtaking him, his entire body language convulsing away from me in squirming, weasely horror. “No, the government shouldn’t punish people!!”, he cried, “It would be much better if the government rewarded good actors like OpenAI! Positive incentives!” Yeah. How convenient.

The spiritual core of this type of nerd is “I should be allowed to to do whatever I want, because I’m smart, and society should be organized around allowing me to do whatever I want.” Academia should fund me to do my useless research, governments should allow me to import greymarket narcotics, the world should allow me to build AGI because I think it’s cool. And I will cry and scream and use all my IQ points to divert resources from other things to make it so I can play with my toys all day, and be praised for it. It’s not enough that I get to do useless math all day, you also need to pay me and make me high status for doing it!

They are children, weasels and cowards.

It’s basically a form of low grade sociopathy.

24. Not being a nerd is almost equivalent to being a bad person

Yeah but the New Atheists were right tho.

Theodicy is a devastating flaw of monotheistic conceptions of a omnibenevolent and omnipotent god. Religious fundamentalism is really bad! Building new technology is really important!

The reactionary position (“just go back and don’t do progress” or “just take all the nerds’ toys away”) is also untenable, and evil.

Science is in fact the correct way to think about reality! (not “academia”, I mean the capital S kind of Science”)

Yes, nerds are very annoying, and very evil. If you work at Meta (or OpenAI, or Anthropic, or one of the other unambiguously evil companies), you are partaking in systemic evil. If you’re a low level, oppressed drone that needs the money to raise your family and you have no other options…ok, maybe it’s just a shit situation.

If you’re a high IQ top tier software engineer, you have options. You are just evil for working there. Oh I’m sure you think you’re a good person, I’m sure you have a whole essay cooked up in your head about why it’s actually totally justified that you should be allowed to do this and you deserve this pay and you are not a bad person. No, you are a bad person.

But, scientists, and engineers and nerds built the wonderful good parts of the world too. They deserve the credit and support for having done so! Many of them were, in fact, not cowards! Or they were cowards that at least worked for good purposes (and got lucky not to be captured by sociopaths).

If you dismiss science, if you disrespect people trying new things, even when those new things don’t work out, you’re also kind of a bad person. We need people trying new things, pushing the boundaries, experimenting, and it’s great that we as a civilization can give those weirdos (of which I am surely one as well!) the resources and space to do their weird experimentation, and we can reap the benefits.

I am thankful to the New Atheists, to the cracked engineers that build the software I rely on, to Kant, Leibniz, Plato and all their autistic philosopher wordcel friends, to the nerds of this and previous generations. Without them, where would we be?

And we need more! We need more nerdiness, more experiments, more new ideas! We need to play with our toys, both to discover new frontiers, and because it is fun. And if we’re not at least having fun, haven’t we already lost?

If you are a nerd, you should try being less of one. If you are not a nerd, you should try being more of one.

Neither extreme is enviable, superior or wholesome. We should strive for the playfulness, the dedication and the intelligence of the nerds, while not using our intelligence to trap ourselves into moral cowardice and exploit the commons around us.

25. Being Good is very hard

There is a conception of what “being (morally) good” entails that is something like “if you want to be good, then you will be good!”

This is utterly, deeply, dangerously wrong.

Being good is very hard! You have to know many things, do many things, learn many things, to be good in different scenarios. Being evil is the default!

Breaking a promise is easier than keeping it. And it’s not sufficient to not “want” to break a promise. You have to actually keep the promise!

If you promise you will keep a secret, but you don’t have enough self control to stop yourself from at a later time blurting it out, then you are incapable of being good! It doesn’t matter if you “intended” to be good, only your actions matter! If you want to be good, you better start working on that self control until you are capable of it.

“Intention” is the useless, easy part. Following through is the hard part.

Do you know what you get if you’re “actually” a bad person deep down, but you just always do the good actions anyways, even if you don’t really “intend” it?

A good person.

26. People consent to a lot of stuff that is still bad for them

I grew up a classic western liberal in the 21st century. A classic tenant of this philosophy is the primacy of consent. If two adults of sound mind consent to do something, that doesn’t infringe on the liberties of others, where’s the problem?

And yet, sports betting. Pay day loans. Fentanyl. Social media algorithms. YouTube Kids slop. Cults. Even just some “alternative lifestyles” (fill in whichever ones you like).

Obviously consent is insufficient. People are getting hurt, a lot, by things where no “traditional” coercion is happening.

I don’t really know what to do about this.

27. Freud was wrong, but not in the way I thought

I have a lot to say about psychoanalysis, I will try to say as little as I can here.

Psychoanalysis is very interesting to me, not because it’s correct (it’s not, it’s extremely wrong), but because it makes correct predictions about parts of reality my other models don’t. I have a pretty good conventional folk model of psychology that usually predicts most behaviors pretty well, but it breaks down and makes wrong predictions when stuff like sex and power dynamics get involved. Psychoanalysis provides interesting frames on those blindspots of mine.

Psychoanalysis tries to present itself as science, and I think this is a mistake. It’s much more productive to think of psychoanalysis as a distinctly western form of mysticism (and people such as Jung realized this connection much more explicitly).

There are many things Freud was wrong about. Most of the things are not very interesting or surprising, and you’ve probably heard the standard critiques many times by now.

But what I think is an interesting mistake, because it still somewhat persists post-Freud, is the belief that the subconscious always “knows” what is “really going on.”

I’m sure someone has made this critique before, but this is obviously bullshit.

You can have a trauma and have absolutely no idea where it’s from or what to do about it!

Saying the subconsciousness always knows everything about its traumas and how to resolve them or what you “really need to do” is like saying every computer also contains the exact source code of a bug free version of itself, just hidden in a weird part of its filesystem.

Why would you ever expect that??

Clinical psychotherapy is closer to software engineering/debugging than it is to magically reading off the answers from your dreams or expressing all your feelings until the correct answer appears. Maybe reading your dreams can be a useful part of the debugging loop (idk, skeptical), but you should not expect the answers to be “in there somewhere.” The solution must be engineered/built.

There are many more things I have to say about psychoanalysis that I will save for another time. But if you want to read psychoanalysis, a few recommendations:

• Freud is much better in the original German

• Jung is much better than Freud, but quite different

• Edward Teach is (kinda) like Freud but just strictly better (read him if you read anyone on this list)

• Lacan lol lmao

28. Evil is a real thing in the world

Most problems in the world are due to complex structural forces, ultimately amoral factors and do not have any cleanly “evil” guy you can point to.

Most problems.

It is fashionable in our modern, “enlightened” world to downplay the influence of genuine, unabashed, moral evil, instead invoking “incentives” and moral relativism, or whatever.

I think this is (psychoanalytical) cope, the kind nerds do. Realizing you are dealing with actual, genuine evil is scary. Especially if they are your boss, friend, lover, family member or political representative. It’s so much easier to find an endless parade of excuses of why it’s not really evil, it’s just “““rational”””, or whatever.

Sure, some people are “just” “amoral” sociopaths that just have no leaning towards good whatsoever. That’s pretty bad. But it is very important to understand that genuinely evil sadists exist!

What I’m talking about are people that will optimize to cause others harm, even against their own best interests. They take inherent joy in your suffering. The more negative your outcome, the more positive their feelings.

People like this exist. They are not super common, but their effects and harms are vastly disproportionate to their frequency.

I could (and might) write a whole book about Evil, these people, their psychology and how to spot and defend against them. There is a lot of important things to be said here.

For now, I will leave you with three of my simplest and most effective trick for detecting people like this:

• They are unfathomably petty. Like seriously, you cannot imagine. They will go to extraordinary, completely irrational lengths to avenge a grudge over things that seem utterly trivial.

• They lie about irrelevant things. Normal people sometimes lie about important things. It’s not good, but it makes sense. If there is something big and important on the line, maybe you’ll get fired if you tell your boss the truth of how the project is going, yeah, sometimes normal people lie. Evil people lie about completely absurd, senseless stuff for no reason. They’ll lie about where they went on their holiday. They’ll claim to be pregnant when they’re not, at a company party. They’ll lie about what your colleague likes for breakfast. Just bizarre things that are easy (but annoying) to check, yet they lie anyways.

• They always have an excuse. It’s never their fault. They’re always the victim. Every bad thing you see them do has some explanation why this was a one off, or actually completely justified. Any individual “mask slip” will be possible to explain away, and they will try a lot to gaslight you into dismissing each individual case (and try to make you look crazy if you try to alert others to these dangers).

These aren’t 100% or enough to always detect evil, but they are extremely powerful heuristics. Ignore them at your peril.

Beware, and be ready to fight.

29. There is no such thing as Chaotic Good

There are exactly three alignments: Lawful Good, Lawful Evil and Chaos.

Good is a subset of order. There can of course be ordered Evil, and it can be a very terrible kind of Evil, but there can be no un-ordered Good.

Good is a fragile, unnatural property. Nature is not Good. Nature is ruled by entropy, and entropy is Chaos.

Love, happiness, safety, beauty are ordered phenomena. They cannot exist in Chaos.

Chaos is hell. It’s really, really bad.

If you deregulate everything, give everyone libertarian freedom, you don’t get capitalist utopia, you get Somalia (or worse). You get roving warlords pillaging and committing unspeakable violence (and that eventually likely evolve into states again).

If you bring Chaos, all that is left is death.

30. Lots of people like being oppressed/abused, and it fuels evil

This is a weird one, and this essay is not good. It’s quite bad actually. But this is important.

To be crystal clear, oppression is bad, and most people who are being oppressed are victims that deserve to be helped and are not at fault.

B U T

This is not always the case.

One of the most influential (to me) books I have read (and still not finished, which the author would find wonderfully ironic) is “Sadly, Porn”, by Edward Teach. It’s a book of raving, vitriolic pornographic projection and psychoanalysis. It’s very good, but not for everyone (or maybe anyone). Amusingly, every single review of the book seems to be wrong. Which feels very fitting if you know the book.

One of the ideas of the book I found most insightful for me, is this idea that people crave being oppressed. That they will go out of their way, of their own free will, to find the most evil creepy sociopath vampire in his castle and beg him to please, oh please oppress them. And the vampire says “well, ok, I guess why not, if you insist, I do like how blood tastes.”

This isn’t the first time I have encountered this idea. It has floated around the vaguely rightwing/nietzschean ideasphere for a long time. I had always dismissed it out of hand as just cope by evil people that want to feel like they are good people. “No, these people actually LIKE when I am evil, I am doing them a favor!!” Classic authoritarian bullshit.

I have recently had to reevaluate that judgement (partially).

It’s true that oppression is bad, and oppressing people is bad. (News flash: If you oppress people, even if they ask you to do it, it’s still bad!) Yet, there is a thing where people will go out of their way to both be oppressed, and to avoid escaping their oppression when given the opportunity.

Some of it is learned helplessness or whatever. But there’s another effect that I had not previously accounted for.

This could be a book in itself, but basically, especially in the west, people go to extraordinary lengths to avoid being blamed. Once you internalize this, a lot of strange phenomena start clicking into place.

The classic outcome of this is that people avoid being strong, or fetishize being weak. Because if you are weak, then you can never be blamed, you are never responsible. Seeing weakness as equivalent to, or precisely what makes, goodness. This view is very repulsive to me. (Here is a related post by a friend of mine)

This also relates phenomena like the Copenhagen Interpretation of Ethics. If you interact with a problem, you could be blamed. So best to not even try to solve problems. But “not even trying to solve problems” could be blameworthy! So what’s the solution? You just have to make yourself so weak that you couldn’t even try to solve any problems, even if you wanted to. Perfect blamelessness nirvana at last.

And one of the best ways to achieve this weakness nirvana is to get someone stronger and more evil than you to oppress you. I mean, if you’re being oppressed by an evil vampire, it’s not your fault when bad things happen in the world! It’s the vampire’s fault! And of course, you are totally 100% against the vampire, but you know how it is…

The most crystal clear example of this in the modern world is engineers working at big tech. I’ve already said this previously, but if you work on Meta recommender algorithms or whatever, you are a bad person doing bad things. But working on recommender algorithms is fun! So how do we close the loop?

Find an evil, evil vampire (CEO, managers, whatever) that takes all that blame and “being responsible” nonsense off your shoulders. Don’t worry about it, kitten. Just make the numbers on this chart go up, and we’ll do all the scary politics and “being blamed for things” for you!

There might be legitimate worlds where you are actually, genuinely crushed under the boot of some totalitarian corporate or government oppressor. And if that happens, yeah that’s just a really fucking shit situation, good luck.

But that is so incredibly not what is happening here. Top tech talent are some of the smartest, richest, most flexible people in the world. They could figure out other things to do. Or even better, they could put some of that IQ to use into making the world better instead of feeding it to vampires.

One of the most impactful inventions of the late 20th century was when sociopaths figured out how to domesticate nerds, or how nerds figured out how to get sociopaths to adopt them as pets.

Nerds figured out they can give the sociopaths something they want (technological power and obedience), and in return sociopaths give them that sweet, sweet oppression and blamelessness they crave. They never have to feel bad about playing with their toys, or what those toys end up being used for. That’s someone else’s problem.

You were just following orders.

31. A better world is possible

I used to be a “techno optimist”, a “transhumanist”, even. I grew up in a culture where technology promised to solve all the terrible, scary problems humanity was facing.

Everything will turn out ok, just ignore that nagging sense of doubt and ambient disgust at what kind of future this promises. Just build more technology, faster. Faster. Even faster.

It’s for the revolution singularity, comrade, any feelings of doubt or disgust at what this future looks like are just irrational. It’s rational to want humans to be replaced, to maximize utility. What are you, some kind of luddite? You want your children to have a nice life, for there to be pretty art, for you to have time to relax and grow, to take things slow? How inefficient, comrade, we could instead be converting your family’s atoms to hyperoptimized giga hedonium. You of course don’t value your own family more than all the trillions of counterfactual happy beings that could be created instead. You wouldn’t disagree with that, would you, comrade? That would be irrational.

Your values are contradictory, and that’s ok. My values are contradictory, and that is ok.

I want to take things a bit slower. I’m not a transhumanist (not the San Francisco kind, at least), I’m a humanist.

I don’t like the utopia that transhumanists are selling, and for so long, I just couldn’t feel what a humanist utopia would feel like, or whether that was actually a thing, or a thing you were allowed to want.

Recently, I’ve started to really “feel the humanism”, what a well ordered, friendly, wholesome human society could look like. That masters and wields technology wisely and justly, rather than letting technology rule it.

You don’t have to sacrifice your values on the altar of “progress”! You can love things just because you love them, not because they are the most “efficient”, you can enjoy things, you can do things for no good reason at all. You can smell the flowers along the road, you can enjoy your weird little niche art that’s only for you and your friends.

You can have a world like this, and you should demand it.

Human values are complex, multifaceted, contradictory. It is the primary goal of every extremist to grind down the rich beauty of humanity’s soul into the simplest, most disfigured shape it can. Because then it will be so easy to get what you wish for. If only you wished for something you didn’t actually want.

Historically, the Human Project was one spread over generations. You didn’t strive to solve all the problems within your lifetime so you could enjoy them, instead you chipped away at the overwhelmingly huge problems bit by bit, so one day your children’s children could enjoy the fruits of a brighter future. This idea of “progress over generations” features heavily in humanism, and in concepts such as karma.

But imagine if the enlightenment humanists, the greek philosophers, the buddha, had access to the internet. Coordination and the Human Project become a lot easier and faster if you can instantly send messages to anyone in the world for effectively free!

What would the humanists do if they had the technology of today? Could we see the fruits of our labours even within our lifetimes?

We can build a new humanism for the 21st century, today.

For the first time, the final stage of the Human Project is in sight. The project of peace, prosperity and the pursuit of happiness for all.

It can be done. The Project is not yet completed, but it can be.

Final Reflections

I don’t have many regrets. Maybe I should call my mom more often.

I had some terrible things happen to me in my life, things that could have been Game Over, but they don’t really bother me anymore. You can just grow and get over things. You are not your trauma.

I have been very blessed. I am in good health, I am surrounded by people I like, and working on problems I care about. My job is very demanding, so sometimes I get put on an emotional rollercoaster, but I wouldn’t have it any other way.

Overall, I’m very happy. I have everything I want. I am very grateful. I wish this for everyone.

I don’t know if I will live to see 40.

Or if I do, whether humanity will still be in charge of this planet. I don’t expect so if things continue as they currently are. Maybe I’m wrong. I hope I am. But it’s not looking great.

The world can still be saved, but not for much longer. I think we will likely see Game Over in the next 2-5 years.

I am very grateful to have gotten to such a wonderful state in my life before that point. I am grateful, but the battle is not over, it is not yet time to rest.

The shape of the Problem has been clarified in my sights, I can see it more clearly than I ever could before. And for the first time, I think I see how to fix it.

I don’t know if I can do it, definitely not alone. Building a better humanism, a prosperous and peaceful world, for everyone, to master technology before it masters us, to wield our powers wisely and justly to enable our universal pursuit of happiness, is a tall task.

The odds were always against us.

But if there is one thing I believe, it’s that this world is worth fighting for. It’s worth it, for all the people enjoying their lives, all the people suffering, all the children and happiness still to come. For all of them, for you, for me, it’s worth fighting.

I think I might know how to save the world, and dammit I’m gonna try.

— Connor, 2025

I might turn thi…

Read more

It is mandatory to start any essay on AI in the post-ChatGPT era with the disclaimer that AI brings huge potential, and great risks. Unfortunately, on the path we are currently on, we will not realize those benefits, but are far more likely to simply drown in terrible AI slop, undermine systemic cybersecurity and blow ourselves up.

We believe AI on its current path will continue to progress exponentially, to the point where it can automate, and summarily replace, all of humanity. We are unlikely to survive such a transition.

Powerful technology always comes with powerful risks, but this does not mean we have to wait idly for the other shoe to drop. Risks can be managed and prevented, while harnessing the benefits. We have done it with aviation, nuclear and other risky technologies, and we can do it again. But currently we are not on track to do so with AI.

What are we doing wrong with AI, and how can we do better? Given where we are currently with AI technology, what would it look like to actually build things safely, and usefully?

We think the answers are downstream of practical questions of how to build cognitive software well.

AI is often seen as a weird brain in a box you ask questions to and try desperately to cajole into doing what you ask of it. At Conjecture, we think about this differently. We want to take AI seriously as what it is, a software problem.

What would it mean to take AI seriously as software?

Part 1: Cognitive Software

The field of AI is weird. AIs are not like traditional software. They are more “grown” than they are “written”. It’s not like traditional software, where an engineer sits down and writes down line by line what an AI should do. Instead, you take a huge pile of data and “grow” a program on that data to solve your problem.

How these “grown” programs work internally is utterly obscure to our current methods of understanding, similar to how e.g. the human genome and its consequences for health are still extremely murky in practice. Even if we have the full genome of a patient sequenced, while we might be able to notice a few gene mutations that are well known to have certain effects, most of the genome and its functioning is completely unintelligible to us. We are in a similar situation with understanding neural network based AI systems.

This weird fact leads to much of the downstream strangeness of the AI field. We tolerate types and frequency of errors that would be unacceptable in any other context; and our methods for ensuring safety and compliance are pitiful in their efficacy compared to what would be needed given AI’s transformative potential.

AI, what is it good for?

The thing we want AI for, the thing ultimately we are using it to do, is to execute “cognitive programs”, to build Cognitive Software, so we claim.

When I say “cognitive programs” or “cognitive software”, what I intuitively mean is “the stuff you can get a human to do, but not (currently) a computer”, or “anything you could write down on a sheet of paper and hand to your intern and expect them to be able to do it reasonably well.” Things that we can’t (yet) formalize on a computer with traditional computer code.

This is what we want from AIs. “Sheets of paper with instructions handed to interns” are not currently executable by computers, but we would like them to be. These kinds of instructions are what most companies and human cognitive labor are built upon.

Our traditional methods of software development have not been up to the task of solving these problems, and so we have seen the emergence of methods of Cognitive Engineering, most famously neural networks, LLMs and Prompt Engineering. But the field is currently nascent, informal, and full of slop. The key to both an economically abundant and safe future is developing and wielding a mature field of Cognitive Engineering.

The way to develop the field of Cognitive Engineering is to think of AIs, and the workflows we make with them, not as magic brains that we talk to, or as inscrutable agents, but as software that needs to be developed, tested and maintained securely and effectively.

What can this view teach us about building better, safer and more effective AI systems?

A Tale of a Database

There is a really (morbidly) funny story, from a user on hackernews, who used to work at Oracle. For those unaware, Oracle is a legacy software company, mostly providing extremely expensive and baroque software and services to massive old companies. Oracle sells a database, the Oracle Database, and in the post, the user talks about how the codebase for this is one of the single worst codebases known to man.

It is millions of lines of poorly documented, confusing, horrible mess of code. No one knows how it all works, it’s all a messy disaster and it all relies on thousands of “flags” that all interact with each other in weird and undocumented ways. It’s hell!

And so the only way Oracle can do anything with this codebase is that every time they change a single line of code, they have to run literally millions of tests, which takes days on their cluster.

And every such change breaks thousands of tests, so you have to then go through each one, fiddle with all the flags, until eventually, at some point, you’ve found the right magic incantation of settings for your edge case, and can submit your code, which gets reviewed and merged sometime months later.

This is a terrible way to build software! Absolutely terrible! It’s not just extremely inefficient and costly, but there is also just no way to actually find and fix all possible bugs or vulnerabilities.

It’s simply impossible to design software that is safe and effective this way, it can’t be done, it’s too complex, no one understands the code well enough. I can guarantee you there are numerous fatal security flaws hidden in that codebase that just cannot ever practically be discovered or fixed, it’s just too complex! Complexity is the number 1 enemy of security.

This is not how we want to design real software for real world applications, especially mission critical or in high risk environments!

The Punchline Should Be Obvious

And the punchline to the Oracle story is: This is how we currently develop AI, but worse!

At least Oracle had a codebase, we don’t have a codebase at all, no matter how terrible! Our AI is a neural network, a huge blob of numbers, that we can’t look inside of or understand!

The way we currently build cognitive programs is to throw the largest pile of slop we can find into our GPUs, run inscrutable linear algebra over it, and then ask the resulting LLM to solve our problem. And then if it fails…what?? Try again? Ask more nicely? Feed it more slop?

And then maybe, if we’re really following “best practices”, we run a huge suite of “evals” on our model and squint at the numbers and see if they move in a good way or not. But these are not tests! At least at Oracle, they can write tests that test each little part of the code in isolation, or each specific edge case.

But we can’t do this with AI, because we don’t know how the internals work and how to test them properly. We can’t find every edgecase, or test every “subpart” of the neural network in isolation. And “fixing” one eval often breaks other ones, with no way to predict when or why. So we’re just guessing, and things can (and do) catastrophically break in ways our evals supposedly test for, constantly!

There is no process, no science, no systematic method to debug or understand why your prompt or AI didn’t work, or to find what edge cases would break it. It’s completely brute force, trial and error, no better, worse even, than Oracle’s magic flags!

This is a terrible way to make complex software! There is no way to make this reliable, safe and effective, even if we really really tried! (and most people are not even really trying very hard)

Part 2: AI Slop and Complexity Debt

We can argue about the fine details of how much value AI has or has not brought to various parts of the economy, but one thing it has undoubtedly brought us is unfathomable amounts of utter slop.

AI has not just brought slop to your parents’ facebook feed, but also to software engineering. Both through the direct writing of dubious code due to (ab)use of coding “assistants”, and more directly by the AIs themselves becoming critical components of the software systems. The use of AI has been dramatically increasing the complexity and attack surface of software.

Complexity is the enemy in all domains, not just in developing software. Take a more general domain: Responsibility. As the complexity of a system, software, bureaucracy or otherwise, grows, it becomes harder and harder to assign responsibility to anyone, or anything, involved. Everyone knows this effect from interacting with massive bureaucracies: they are inhumane, there is never anyone in charge that is actually responsible for solving the problem, and so problems don’t get solved and no one is punished for it.

Imagine if for example Facebook was using a handcrafted recommender algorithm, made of code, and someone decided to add a line of code along the lines of “if user is part of $ETHNIC_MINORITY, downgrade their popularity by 10%”. If this happened, it would be easy to prosecute, the line is clearly findable and visible to a court. And even better: We could find which person wrote that code, and every person in the line of command that resulted in that line of code being written, while exonerating anyone not involved. This is great, we can detect when a bad thing happened, find exactly who and what is responsible, and make the necessary adjustments.

But, of course, the Facebook recommender algorithm is not made of clean code, and is instead a massive deep learning blob. So now, whenever it turns out that an unfavored political group was getting deboosted, they can (and do) simply wail “it’s The Algorithm!! It’s not our fault! How could we be responsible for the organic and unpredictable effects of people interacting with The Algorithm??? Should poor innocent white collar software developers and managers go to jail for what The Algorithm did???”

The antidote to this is simplicity (or at least, well managed complexity). Complexity shields people and systems from accountability and makes the system resistant to being changed and fixed. The more complex a system, the less accountability, the less responsibility and the less humanity.

Conversely, the simpler a system is, the easier it is to make a reasonable effort, to prove good faith and exonerate oneself if something actually goes wrong. Simplicity sets up the incentives so that people are incentivized to not fuck it up, because they would be personally responsible.

As an IBM presentation from 1979 famously said: “A computer can never be held accountable. Therefore a computer must never make a management decision.” And yet, computers now manage all of our online social relationships and media. And thus, no one is being held accountable when things go wrong.

The Strange Decoupling and Software Senescence

Perhaps the strangest thing about AI is how much it has decoupled the capabilities of your system from your understanding of the system.

Understanding is the core to simplicity. The more you understand your system, the more you can understand its indirect effects and externalities, the safer (resilient to accidents) and secure (resilient to attackers) you can make it and the easier it is to predict how it will act outside of its normal range (generalisation).

The more you understand a system, the simpler it becomes. “Perfection is reached not when there is nothing left to add, but when there is nothing left to remove.” When you don’t understand something, it’s muddy, easy to delude yourself, super complex. As you understand it more, things become simple, sharp, clear.

The better you understand a system, the easier it becomes for other people to work with, the easier it becomes to transmit and teach and build upon, to integrate with other systems. As you understand the boundaries and limitations of your system, integration becomes natural.

The typical flow of science is something like: “Messing around with some small thing” -> “You gain more understanding, which also gives you more capabilities, and all the other nice things above” -> “You build big reliable projects using your new understanding”

Generally, given a certain amount of understanding, there’s only so much you can do. Capabilities are bottlenecked by understanding.

For instance, in traditional software, as you add more and more capabilities to your software, it becomes more and more complex and brittle, unless you understand and manage its complexity very well, until eventually it becomes so complex that adding more capabilities or fixing bugs becomes an impossibly daunting task, and you are stuck. Many legacy software companies find themselves in this unenviable position. Call this “Software Senescence.”

But this doesn’t just happen in software, it applies everywhere. If you push too far beyond your understanding, you quickly get signal from reality that you are screwing up, and things start to break.

In AI, things are very different. You can in fact get more capabilities without increasing your understanding, just by shoveling more data into the GPUs. There is nuance to this process, of course, but it’s important to understand how different and more brute force this is vs the careful management of complexity you have to do with traditional complex software. You can’t just slam ever more lines of code into a codebase to make it better (despite the best attempts of many large corporations).

The “scientific” process of AI looks more like: “mess around with huge things” -> “gain no new understanding” -> “gain new capabilities but none of the nice properties above” -> “build big dangerous things”

This is extremely perverse. All of our natural expectations of this process are reversed. We intuitively assume that as a system becomes more capable, it comes hand in hand with better understanding and all the nice properties that come with that. If I see someone has built a much faster, more maneuverable and capable airplane than I can build, I assume he understands airplanes better than I do. But here, AI subverts our expectations, and we not only don’t gain the understanding we expect, but lose ever more understanding as capabilities increase.

Algorithmic Carcinogenesis

If the natural lifecycle of software terminates in arrested senescence, an ungraceful but not worsening stasis, then the natural lifecycle of AI leads to a form of algorithmic cancer. And it has been eating us alive for a while now.

Algorithmic cancer is an uncontrolled and unconstrained tumorous growth that infects everything it touches and crowds out healthy tissues, just as AI slop is crowding out true humane creations, and how social media and recommender algorithms killed the diversity and creativity of the old web before.

It’s pretty viscerally intuitive that there is something gross about the proliferation of low effort, mediocre AI content. Just try using Google text or image search for any common topic, and you can immediately see what I mean.

Pictured: A Google image search for “baby peacock.” Each image framed with a red box is a fake AI image. Baby peacocks do not have the extravagant tails associated with adult male peacocks. Knowing this makes these AI generated pictures particularly egregious and easy to recognise.

Having “more content” is not good, the same way having “more cells” is not good. A flood of the mediocre crowds out the actually good, what you are actually looking for, and rewards the lowest possible common denominator, anything that can trick or grab your attention for even a second.

The “demand” for cute baby peacocks with extravagant tails drowns out the pictures of actual baby peacocks, let alone the few unlucky human artists that put in the time and effort to accurately represent cute but realistic baby peacocks. Think of the work that goes into creating scientifically accurate artist’s impressions. Instead of putting in the hard, subtle labor of getting the visualization right, AI serves up the preconceived notions we already have and drowns out all other considerations we humans care about.

There has been a palpable sickness in the internet for quite a while, at least for the last 15 years or so. Gen AI slop is the most salient recent expression of it, but it is not where it started. Machine Learning’s first use wasn’t Gen AI. Lots of its earliest funding (at places such as Google, Facebook and YouTube) was in order to develop The Algorithm for social media.

Slop is not new to the AI era. Do you remember listicles? Buzzfeed? The transition from the wild, but human-built, west of Web 1.0 and early Web 2.0 to the homogenized, centralized, algorithmically curated walled gardens of social media? Machine learning was integral to the slop focused curation from day 1.

The mercy that Software Senescence grants us is that it at least provides something like a natural limit to the growth of cancerous software (even though large corporations spend a lot of money to counteract this effect and keep their virtual tumors alive and spreading). If your burgeoning tumor grows beyond your ability to manage, it at least doesn’t get much worse before utterly breaking down and (hopefully) being put out of its misery by either you or your competitors shortly thereafter.

And this again generalizes far beyond software. Historically, if you didn’t understand the chemicals you were working with, you would more often than not end up inhaling poisonous fumes or dying in an explosion, as was the fate of many alchemists and early chemists (remember: Every warning label on a chemical was, at some point, found out the hard way). There was a strong pressure to actually understand what you were dealing with, leading to all the fantastic benefits of modern science and civilization.

In the past, the feedback loop was short and direct. With AI, the effects are far more insidious: We all have a creeping feeling that our social life has deteriorated since the advent of social media, but it’s hard to pin down exactly what went wrong and when, precisely because the whole thing is so complex and inhuman.

Algorithmic cancer is dangerous because it doesn’t have a natural limit to its lifespan, its spread, or its virulence. If there is any further compromise on quality that can be made in exchange for more growth, cancer will take it. Every last bit of human soul snuffed out for maximal engagement. The cancer is in our media, our art, our software, our soul, it is everywhere, and it is spreading. And AI is its perfect vessel.

The enshittification and carcinogenesis of the internet has been supercharged by people building and deploying shitty, myopically designed and complex AI systems en masse. This is not the mere result of “The Incentives”, “Moloch” or “Technological Progress” at play. The people and companies involved have hired lobbyists to explicitly lobby governments and regulators to be allowed to keep doing this. This is not a natural phenomena, it is a deliberate pollution and toxicity induced carcinogenesis.

Cancerous tumors bring a lot of “growth” (by some metrics…), while making everything around them worse, and setting up the body for catastrophic, terminal failure. This is what is currently happening to our entire information ecosystem. Sounds like a great VC investment if you ask me!!

Neither New Nor Unsolvable

Is Conjecture somehow the first people to ever think about all this? Of course not, not by a long shot!

Healthy standards exist in many, if not most, fields. Other fields of engineering routinely predict and bound the effects of big things before they happen or are built: Bridges, airplanes, nuclear reactors… Anything that matters usually goes through a process of “stupid small thing -> science -> reliable big thing.”

We can quibble about how much science or safety is needed (FDA lol), but this is how good things happen in our civilization.

On the other hand, in machine learning, we just immediately deployed social media algorithms to billions of people without any constraint, oversight or science of what would happen. ChatGPT was deployed to 100 million people in less than 2 months. Anthropic just recently deployed full AI agents running on your computer. All of this is further, reckless carcinogenic pollution of our shared information ecosystem. All websites need to ramp up their security in response to these new systems, and we can now never be sure whether someone we are talking to online is really a human or not.

Sometimes people say things like “well so far AI hasn’t led to any large scale damage.” And I just couldn’t disagree more. Any time I am on social media now and see someone well articulately responding to my post, I can no longer be sure if it is a real person or not. AI generated images clog up search engines and crush artists, and social media recommender systems savage the mental health of the younger (and older) generations.

AIs have made a ludicrous number of people addicted to social media and waste their time. Instead of forming real human bonds, participating in civic and political life or building families, people made addicts by The Algorithm scream on social media. Instead of learning to make art with the wonderful help of the many resources on the internet, people are just resharing deepfakes.

All of this is explicitly the result of AI optimization. The few who do not fall prey to this do so despite AI, not thanks to it. This is the opposite of what technology should be!

The goal of technology is (or at least, should be…) human flourishing. We should be getting nice things, as people ascend Maslow’s hierarchy. But here, we are building technology specifically designed to alienate people and crush the human spirit: more complexity, less meaning (“it’s not our problem if people lose jobs or their communities are fractured”), appealing to base instincts, and all of this en masse and delivered directly to your smartphone, 24/7.

If we were to tally up the amount of effort, time and sanity lost to these effects…how high would that cost be? My estimate is it would be absolutely massive.

And who is paying for that? The people building and deploying these AI systems? No, it’s you and me. The river is polluted by the chemical companies and we are left drinking the toxic, carcinogenic water, while they lobby the government to not regulate them.

Remember SB-1047? It was a modest AI regulation bill that specifically asked for liability in cases where AI causes more than $500mln of damage (and no liability below that). And yet, tech companies viciously opposed, and ultimately killed, this bill. You can start to think about why that might be.

But it doesn’t have to be this way, it never had to be this way, and it wasn’t this way.

Historically, technology led to amazing things, and drastic improvements in human flourishing. The crises of fertility, housing, meaning, etc are fairly recent. This was not the norm for most technology for most of human history. Nobody expected this would be what technology would bring. Economists a hundred years ago were expecting a humanist post-scarcity by this point.

We have done great engineering in the physical world. Airplanes and nuclear reactors are astoundingly safe and provide much value to humanity, without demeaning and polluting our souls and societies.

So how do we embody this vision at Conjecture? The forefront of innovation in the 21st century has been software. Bits, not atoms. And what does the frontier of software look like? Javascript webapp framework slop as far as the eye can see.

Can we do better?

At Conjecture, we do Software, and history is also rife with examples of great software, principles and development practices that have stood the test of time and help us do the science to make our systems safe, secure and humane.

Famous examples include:

• The Unix Philosophy. First laid out in 1978, these deep principles of modularity in software design can be translated to many different contexts and are often synonymous with what “good software” is.

• The Relational Model and SQL. First proposed in 1969 (!), the Relational Model was a theoretical model, motivated by formal logic, to solve hard problems of how to store and retrieve complex structured data. This model turned out to be so good that basically all widely used, extremely performant databases even today, more than 50 years later, use this model. Even 50 years later, in an insanely competitive field of research, the RM dominates.

• Type Systems and ML/Ocaml. Generally, there are 2 points at which you can catch a bug in your program: During running (asserts, validation), or after a program has run (debugging, looking at logs, pain). But there is a third, magical thing one can do, which is to catch bugs before running your program. Type Systems are one of the best and oldest methods for doing this, incorporating deep, powerful principles of formal Type Theory into programming to solve practical problems in a way that feels intuitive and deeply integrated into the language. It allows for fantastic UX for static analysis and formal verification. Programming languages that descend from this tradition, most notably ML (no relation to Machine Learning) and its descendants such as Ocaml, are some of the most amazingly well constructed and rigorous programming languages ever made. To this day, ML (which dates from 1973) is one of the very few languages completely specified and verified using formal semantics.

Good software engineering is Health, it is managing the health of information and software systems. And luckily, we have in fact learned a lot about how to build and manage the complexity of software over the decades. We should work hard to apply these lessons our senior engineers have learned via trial by fire to our new class of complex cognitive machines, or risk the next generation of software tumors crowding us out, and not just in the online realm.

What could it look like to develop principles and concepts as powerful as these for AI?

Part 3: A Better Future and How to Get There: A Roadmap

A better future is possible. We can learn to build AI as the complex software it is and harness the lessons from the history of software engineering. This is what we build at Conjecture.

We build the tooling to turn developing cognitive programs into a science, to treat it as a software engineering problem, rather than as negotiation with a weird little slop imp in a box.

As the world drowns in slop, our counter is to be extremely opinionated on all points to fight back against unsound defaults and norms. There are many places where things can go wrong, and our roadmap tackles each one, one by one.

Our roadmap can be roughly split into 5 phases, each building off of the previous, and becoming increasingly ambitious and experimental, taking us further and further from the comfortably unsound practices of today to a world of 21st century cognitive engineering.

This roadmap is a sketch, and necessarily will change as we progress. Realistically, the real world is also far less linear than this simplified map might make it seem. We have already done a fair amount of work on Vertical Scaling (Phase 4) and Cognitive Emulation (Phase 5), and lots of work of Phase 1 and Phase 2 happens in parallel.

Phase 1: Foundational Infrastructure

We have been doing research on cognitive algorithms for almost two years now, and so, so many problems in AI development start way, way before you even touch a neural network.

What we have found again and again is that more often than not, one of the biggest pain points of our work was for researchers to just have a nice interface to quickly write their own AI scaffolding without dying to Python research-code hell. Over and over again, we were slowed down more by poor devops than we were by research ideas or execution.

And especially for non-research code, whenever we wanted to build a useful AI app, like a writing assistant, or an internal Perplexity-like AI search engine, we would get stuck in moving from experiment to production. We could write our own cognitive software and impressive heuristics, but then inevitably would end up spending inordinate amounts of time managing database connections, sessions, authentication and other odds and ends that were just distractions and hindrances from actually developing and deploying cool stuff.

So the first step in building a 21st century cognitive software stack is solving these 20th century devops headaches for good. All the most clever cognitive code is useless if the underlying infrastructure doesn’t work, or is hell to use.

Our new platform, tactics.dev, takes solving this problem to the extreme. With Tactics, we take care of the backend, combining some of the best ideas of services like firebase and cloud functions, with first-class AI support.

Using Tactics, anyone can create a tactic with a few clicks, and as little as a single line of code (here to ask an LLM to give us a fluffy animal):

selected_animal = $do(“give me a random fluffy mammal”)

…that’s all that’s needed to have your API up and running! You get parallelism, auto scaling, LLM ops, devops, secure authentication and more straight out of the box, no setup needed!

This makes iterating on AI-first workflows and deploying them to production easier than ever, no slowing down or getting caught up in managing fiddly backends, just focus on what matters.

Tactics already has many useful features such as:

• First class LLM support. Switch between model providers and manage your API keys with ease.

• Every tactic is an API. The moment you have written your tactic, you have a REST endpoint you can hit.

• No async/await pollution. Just make llm calls and put the results into variables. It just works.

• Just assume the database is always there. No need to open a connection, no dangling closes.

• Sessions and authentication are directly integrated, no setup required.

Tactics.dev is rapidly developing, and we want your feedback for making it better! Deploying an AI app should be a breeze, so you can focus on building the hard parts of your application, and we need your input to help make the experience as good as it can possibly be.

Try tactics.dev today!

Phase 2: Cognitive Language Design and Computational Control

Sometimes when thinking about the massive potential of AI, people talk about “aligning” AI systems to humanity’s wishes. This is often seen as an impossibly esoteric field of study that resembles abstract philosophy more than any actual form of engineering.

We disagree with this, and think there is already a rich, powerful field of study concerned with exactly the problem of how to express a human’s wishes to a computer efficiently in such a way that they get what they wanted: Programming language design.

When people talk about coding in the AI era, they imagine coding with the exact same programming languages and paradigms from 20 years ago, just with a little AI assistant talking to you and regurgitating javascript boilerplate for you.

This is obviously ridiculous and dramatically lacking in imagination and ambition. We need to go back to the drawing board. What does programming language design for the cognitive software era look like? What are the new abstractions? The new primitives? What if you design your entire language around deep integration with, and management of, complex cognitive systems, rather than treating them as a clunky API pasted on top of your existing software semantics?

We have developed our own programming language, CTAC (Cognitive TACtics), with which we are doing exactly this kind of exploration. It’s somewhere halfway between a traditional programming language and a prompting language, allowing for easy and strongly integrated access to the LLM backend. Language design is very hard without sloppifying the language (as can be seen with ~all of the most widely used languages today), so we are still iterating.

But why a programming language? Why not just a library or framework in a more popular language? There are a number of reasons.

First, you always will inherit the mountain of slop complexity that comes with any existing language and its package and tooling ecosystem. And lets face it, one glance at any programming forum is enough to tell how people feel about the quality of the languages and ecosystem of the most popular languages (hint: Not Good). These languages and their tooling simply come with decades of organic growth and accumulated technical debt, which makes supporting and building on them hell.

Second is that we fundamentally want to do things that are hard or impossible to do with current languages in “user land” alone. A good inspiration for our thinking is Ltac, the “tactics language” of the extremely powerful theorem prover Coq (You may notice an etymological inspiration for our own tools).

Ltac is a stateful language for managing the proof context. It lets you manage your current assumptions and conclusions that are still left to prove.

There are many control flow options and heuristics in Ltac that are not part of common programming languages. An example is the keyword “try”, which tries a tactic, and if it fails, reverts the entire proof state back to what it was before, and tries again. Doing this as the user of another language would require having access to all of the state of the whole program, which can be done, but is extremely painful. In Ltac, it’s just a single keyword.

Similarly, there are many high-level heuristics and control flows that you would want to use when building cognitive software, and we want to integrate great primitives to do so into the language itself.

There are many other patterns of control flow like this, such as:

• Semantically aggregate the result of applying a tactic to a folder/20 websites/etc

• Pick the tactic that is most relevant given a context and guardrails

• Look at the recent execution trace, and see if things went wrong and execution should be interrupted until a human takes a look

We have often wanted to implement features like these cleanly. It can be done in other languages, but it is often extremely ugly. We have often had to struggle with lots of boilerplate from badly integrated libraries, or suffered from programming languages’ lack of strong control flow primitives. If you wanted to implement proper advanced control flow in a library/user-land correctly, you’d need languages with features like typed Effects and well-typed Metaprogramming, and for some reason in the year of our lord 2024, most languages still have neither.

But, we don’t need typed Effects or well-typed Metaprogramming to implement such things in CTAC, because we control the language! So when we want to implement new control flow primitives, we can just directly modify the interpreter! (Though we may still be coming back to implementing a proper typed Effect System…)

And third, there are new abstractions and primitives to be discovered and implemented for the cognitive programming era! We are working on finding these right abstractions, and our control of the language allows us to iterate on and smoothly integrate these new primitives.

What could new, cognitive primitives look like? We could natively track uncertainty throughout program execution, and bound the context of the cognitive coprocessers precisely and painlessly. What if we had keywords such as “reflect”, that allows the cognitive system to look at its own recent execution trace and write down its thoughts about it, or “reify”, going through a trace and distilling it into a tactic, all directly and deeply integrated into the language?

CTAC is still far, far from feature complete, and should be seen as an early proof of concept. Phase 2 is going from where we are now to a true 21st century cognitive programming language. Having full control over the language will allow us to implement features and build tools that just aren’t possible otherwise, or that would be crushed under the weight of inherited technical debt from other, older languages.

If you want to try out CTAC, you can do so on Tactics.dev, here!

Phase 3: Horizontal Scaling: Scaling Without Sprawling

Now that we can write, execute and deploy individual cognitive programs effectively, how do we scale? How do we get from small chunks of cognition to large, powerful systems that do what we want and cover the scope of what we care about?

There are two kinds of scaling: Horizontal and Vertical. Horizontal Scaling is about breadth, increasing the scope of different tasks of similar maximum complexity and putting these parts together into a larger system while keeping it coherent and manageable. Vertical Scaling is about depth, increasing the depth and maximum complexity of individual parts of the system. Horizontal is about solving more tasks, Vertical is about solving harder tasks.

We start with Horizontal Scaling. Adding more and more simple features, intuitively one might think, should not result in an overall complex system, but this is almost always what happens. This is the problem of Horizontal Scaling. How do you build and maintain many different cognitive components? How do you make sure they remain coherent and compatible as the scope and size of the project grows? How do you maintain oversight of the whole system and its capabilities, and confidence in its safety?

As a motivating example, let's look at one of our early inspirations for our approach to cognitive engineering: Voyager.

Voyager is an amazing paper, I really recommend you read it and check out their videos if you haven’t seen it before. The basic goal is to get GPT4 to play Minecraft. To do this, they build a pipeline of prompts that generate snippets of JS code (“skills”) that interact with an API to play Minecraft to solve various tasks. Crucially, as the tasks become harder, the new “skills” can refer to and reuse previous “skills.” The neural network never directly interacts with the environment, only through these little “shards” of code. Lets call this type of system a “Crystalline Learning System” (CLS), as it "crystallizes" its learning into discrete shards of code, rather than encoding everything in fuzzy neural weights.

Lets consider a generalized CLS, that interfaces directly with a keyboard and mouse on an open ended computer rather than just with Minecraft. By default, this is quite unsafe, there is no bound on what such a system might learn or do. But notice there is actually little that is needed to make this bounded and manageable: Its artifacts are just regular JS programs that are quite legible to people, and we can apply all kinds of complexity and legibility measures to them (both of the traditional and AI variety).

This is an approach to Horizontal Scaling: Without needing to make the base model a more and more powerful blackbox, we can extend the capabilities of the system in a way that humans can understand and control. You could even have the humans write these cognitive shards, or tactics, directly!

The first and most important part of any development and debugging loop is the human in the loop. A developing codebase is not a static thing, but an interactive system with its human developers, and giving the developers the tools and affordances to interact with and intervene on the system when needed is crucial.

There are 3 possible times for human in the loop:

• Static analysis: Review the code before it is run.

• Runtime analysis: Monitor the code as it runs and pause its execution when needed.

• Trace analysis: Look at the trace of execution after the fact and figure out what happened.

All of these points of human-machine interaction can be turbo-charged with strong principles, good tooling and, if done right, AI to automate what has been crystallized from experts.

The key to doing Horizontal Scaling is to get this right. If you do this kind of stuff, especially integrating AI in the devloop, naively, you just get slop. If you use tactics that just generate super sloppy new tactics that don’t compose well, that do not preserve invariants and implicit properties, everything goes to complexity hell really quickly.

You can see this effect in action with Voyager: Taking a look at some of the skills Voyager developed during a run, it has some very sensible skills such as “craftChest.js” and “mineCoalOre.js”, then ends up with badly abstracted slop like “catchFiveFishSafely.js”, “mineCopperOreWithStonePickaxe.js” and “mineFiveCoalOresV2.js”

We need to apply strong safety engineering and software architecture principles to do this in a way that scales horizontally cleanly and properly, rather than ensloppifying everything. This is the key problem that Phase 3 must solve.

And there is a lot here to be done! Better type systems, time-travelling and multiversal debuggers, prompt tracing, neural debuggers (programs that analyze neural networks, attention patterns, etc, directly to guide debugging of prompts. We have developed quite a bit of this in house, but they’re not quite yet ready for wider release), and, eventually, properly constrained and reliable AI reviews and tests.

But we must go further than this. Bring back complexity measures like cyclomatic complexity, and improve on them with all the knowledge we’ve accumulated since they were introduced! Have you heard of FRAP? It’s a book from the future about writing correct software that somehow ended up in the early 21st century instead of the 22nd century, where it belongs. We should apply it to building cognitive software if we want to scale and manage our complexity!

Phase 4: Vertical Scaling: Deliberate Teaching and Cognitive Provenance

Proper Horizontal Scaling is already insanely powerful, and allows us to cover a vast and expanding scope of relevant tasks, especially when fueled by the powerful cognitive engines (LLMs) already in existence today.

But sometimes, you need the underlying cognitive engine to exhibit new behaviors that shards/tactics are just not a good medium for, such as:

• Automations that should be quick and not need to go through the equivalent of 17 post-its.

• Tasks involving “vibes” or styles that would be too painful not only to put into code, but also to maintain and edit in a coherent way.

• Facts that should be baked into the model for when it performs other tasks.

Right now, our solutions to these problems are terrible. Whether it is pretraining or finetuning, we just feed the model piles of data and hope that the increase in measurable evals translates into better performance on the tasks we care about. This goes against our philosophy, which is to craft and understand what is happening when we add capabilities to our system.

What we would want is a training procedure that lets us understand what the model does and does not know at each step. What might that look like?

One would need some kind of “deliberate teaching” procedure by which one can selectively add pieces of knowledge, check if it was integrated into the model, and how it was integrated. You would want to be able to know how the new knowledge generalized and interacted with other patterns and knowledge, and revert back in case something happened that was not intended.

Such a procedure is the fundamental cognitive building block for constructing vertically scalable AI systems in a safe and controllable way. But we still have the problem that if we build off of already pretrained models, we can’t know everything that is already in them, and reverse engineering seems hopeless.

Taking things a step further, the speculative ultimate result of applying this constructivist spirit to pretraining would be what we call “Empty Language Models” (ELMs), models that are trained first exclusively on patterns but not on facts.

Once you have a base model that is “empty”, you can then one by one add the relevant facts and information you want to use in your application context. And throughout this procedure, you monitor and control the learning and generalization process, such that you know not just what information went into your model (Provenance), but also what generalizations it learned from that data (Cognitive Provenance).

Through ELMs, we can even speculate of creating a kind of “Cognitive Fourier Analysis”, where we can separate the “low frequency patterns” of generalization from the “high frequency patterns” of fact recalling, and then use this information in our debugging and software development loops.

This is the philosopher’s stone of how to turn AI into software engineering. The ability to add and subtract information and monitor the patterns that your cognitive system has access to should make possible not just Cognitive Provenance, but also cognitive unit tests, formal guarantees, and more.

In practice, sadly, developing a true ELM is currently too expensive for us to pursue (but if you want to fund us to do that, lmk). So instead, in our internal research, we focus on finetuning over pretraining. Our goal is to be able to teach a model a set of facts/constraints/instructions and be able to predict how it will generalize from them, and ensure it doesn’t learn unwanted facts (such as learning human psychology from programmer comments, or general hallucinations).

This research is highly speculative at this time, but we have made significant progress on this agenda, and think we know how to get all the way. This is not likely to be in public facing products in the short term, but if you are interested in this, get in touch!

Phase 5: Cognitive Emulation (CoEm)

There is a lot of software architecture and ML research to be done, many systems to build, problems to solve, and this will take time. But it’s worth speculating a bit further.

Ultimately: Where does this all lead?

Right now, we are focusing on short, self-contained programs that can be integrated in larger environments. But how do we scale all the way? How can we build humane and legible cognitive edifices that scale gracefully both in scope and complexity all the way to solving our real, human problems?

Our claim is that the way to get there is to write cognitive programs that emulate human cognition. “Cognitive Emulation”, or “CoEm”.

The way humans solve problems, our type of cognition, is a very specific kind of cognition, and it is different from just telling an AI to solve a problem by whatever means it wants. AI cognition is very different from, and inscrutable to, humans!

In mathematics, there is a notion between researchers of “proof techniques”, tricks and techniques to aid in solving tricky mathematical proofs and problems. Curiously, there is no article on proof techniques on Wikipedia, because these techniques mostly spread via oral transmission.

But some of these proof techniques have been studied in such great detail that they do feature in their own first-class Wikipedia articles, such as forcing, Ehrenfeucht–Fraïssé games or defining greedoids to prove optimality of greedy algorithms. A major milestone of the CoEm agenda would be to systematically distill intuitions such as these to the point they can be reused in cognitive software at scale.

Teaching is an interactive process, there is a back and forth. Books are not very good at this, their linearity can never adapt to the combinatorially explosive nature of possible interactions and possible readers. Iteratively building an interactive teacher in a humane way, that doesn’t hallucinate, where both the students and the teachers master the topics together as they interact, is at the core of the CoEm aesthetic.

The more varied people you can teach something to, the more deeply, you, as a teacher, understand the concept.

With DL, you just feed more data, lol. In a CoEm world, the experts building the systems themselves would be enriched over the course of their many interactions with the systems they build.

And emulating human cognition gives us extremely economically valuable systems (the current economy after all is built entirely on this type of cognition!) that our institutions are already adapted to, and that we can understand and audit and make safe.

Software is a huge topic, involving many different types of thinking, and has amongst the shortest timespans from “idea” to “real thing that people can use”. This makes it a very special field, there are few others where a kid with a laptop can realistically build and deploy something novel that can improve the lives of millions. This makes it so much the sadder that DL has been used to enshittify, rather than enliven, it. With CoEm, we should get better software than the ones we’d write without AI, not worse.

Building well architected CoEm systems would allow us to get the wonderful economic benefits of AI software, while maintaining systems that can be understood and integrated into already existing human institutions cleanly.

We have a lot to say about this, what human cognition is, how to achieve it, etc, but that will have to wait for another day.

Conclusion: A Future Worth Building

This all assumes of course that we make the sensible choice to not go straight for ASI, the same way we eschew building other catastrophic technologies. A Narrow Path lays out what that might look like from a policy standpoint.

If we do, we firmly believe that one of the greatest bottlenecks to getting to a future of incredible economic, scientific and even just entertainment abundance is humanity’s collective poor practices in designing, building and maintaining software. If we carry these practices forward into AI, the outcome will be even worse.

This makes us optimistic rather than pessimistic. Not because humanity has a particularly good track record of building complex software well (just look at any software vulnerability database and weep), but because the same solutions that will make AI systems beneficial will also make them safer.

We have decades of hard-won knowledge about building and maintaining complex software, and concrete examples of success in other critical domains - from flight control systems to nuclear reactor design. If we applied the levels of rigor that we apply to flight control systems or nuclear reactor design to AI systems (or hell, even just got Adam Chlipala to do it)…yeah, that could really work!

This doesn’t solve the problems of misuse or negligence. You can still misuse software, or write bad software on purpose. But what we need is to apply the rigor of engineering to AI development, develop and standardize best practices for cognitive software engineering, and make it not just possible but straightforward to write good AI software.

The path to beneficial AI isn't mysterious - it's methodical.

This is where you come in. Whether you're a developer looking to build better AI applications, a researcher interested in cognitive software architecture, a company wanting to deploy AI systems responsibly and reliably, or someone concerned about the future of AI development, we invite you to join us in building this future.

The journey starts with practical steps. Try our tools at tactics.dev and share your experiences. Engage with our research and development principles. Help us develop and refine best practices for the field. Share your challenges and insights in building cognitive software.

The future of AI doesn't have to be mysterious black boxes or uncontrollable systems. We can build a future where cognitive software is understandable, reliable, and truly beneficial to humanity.

Ready to start? Sign up for our alpha test, follow our progress on X, or reach out to us directly. Together, we can transform AI from black box to building block.

The future of AI isn't predetermined - it's engineered. Let's build it right.

And this applies both defensively (your private home) and offensively. If you want to reconfigure a hostile nation state to have less Power over an oppressed minority or disputed resource, that’s generally considered a Security issue.

Cyber Security is an important subfield of Security more broadly. How do you keep digital assets safe, access and authentication well configured? You don’t want random ransomware gangs to have Power over the encryption status of your hard drive.

Buying Cyber Security

Lets say you’re an up and coming digital business that wants to purchase Security for your digital assets by moving them into some well managed cloud. You are pitched by two companies:

• AWS, the largest and oldest player in the field

• Some Upstart™ Inc.

Which one do you expect will have better Security?

Well, the obvious answer is AWS. They’ve been around for a long time, are trusted by many, are very experienced, etc. Nothing controversial there.

But what process actually underlies this intuition?

• Is it that AWS has some kind of awesome, cutting edge, super security software? No, unlikely, if anything they are probably tethered to tons of horrible legacy systems they have to keep maintaining.

• Is it that AWS has the best talent? Sure, there’s definitely some of this, but did you look up their head of security’s CV and decide based on that?

• Is it that AWS has the best security processes? Maybe, but did you make the decision based on them describing their code deployment checklists?

• Is it that AWS has been hacked less? Surely not, the upstart is likely to have had much fewer critical security incidents than AWS has, and AWS is being targeted directly by the most dangerous threat actors 24/7.

What I want to point out is an underlying intuition I call Security by Blood.

Security by Blood

The reason we trust AWS’ security is because they have Paid in Blood, they have made every possible mistake, gotten hit by every possible threat actor and are still here.

AWS is trusted for security not because they have the best formally verified software stack, or because their head of security is a genius, or even because they don’t get regularly hacked. It’s because they’ve seen everything, been hit by everything and have built, iteratively, agonizingly, a robust enough system.2

This is not the only way one could, hypothetically, reach Security. There are methods such as formal verification that could be used to prevent hacks from ever happening in the first place.3

But in practice, this is just too hard/expensive to do, so people don’t.

Punchline: Everything is like this

The punchline is that AWS isn’t unique, everything is like this!

Security is most often gained not by the best planning, the best processes or mechanisms ahead of time, but by agonizingly throwing yourself against the problem over and over and iterating without dying.

If you can survive a lot of Bleeding without Dying, you can acquire Security.4

But if the Bleeding kills you, you need another method.

My colleagues and I work on the problem of catastrophic risk from AGI. We have recently put together a (mostly) complete account of this topic.

The Compendium is a bit different from most other texts in its thoroughness, accessibility to non-technical readers and a focus on the political, social and ideological aspects of the current state of affairs.

It’s style is far more matter of fact than this blog, but I still hope you consider giving it a read.

I will republish here the foreword I wrote for the Compendium, written in a style more familiar to this blog’s audience:

Compendium - Foreword

A few million years ago, something very strange happened. 

Through minor genetic tweaks, an ancestor of the modern chimpanzee split into a new line of species: Homo, humans. This new chimp variant was odd in several ways: it learned to stand upright, lost most of its fur, and grew a bigger brain. This bigger brain was not really all that different from that of his chimp cousins, just scaled up by a factor of about three.

If you had seen this odd, half naked chimp with a brain three times bigger than its cousins’, and you would have to guess what this new chimp will do, what would you have said?

Maybe you would have expected it to be a bit better at collecting termites, or throwing rocks more accurately, or have more complicated status hierarchies. But that 3x scaled up chimp ends up building nuclear weapons and going to the moon. Chimps don’t go one third of the way to the moon, they go zero to the moon; humans go all the way. 

We still don’t exactly know how or why this happened, but whatever it is that happened, we call the result General Intelligence. It is what has allowed our species to build the magical glowing brick that you are looking at right now to transmit the words of another chimp descendant located halfway across the world to your eyes and brain. 

This is crazy.

General Intelligence is what separates human from animal, industrial civilization from chimpanzee band. It is unlikely to be a discrete all-or-nothing property, but it sure is suspicious that you go from “zero going to the moon” to “all of going to the moon” within a 3x difference in brain size. Things can change quickly with scale.

Our intelligence makes us the masters of the planet. The future of chimpanzees is utterly dependent on what humans want to do with them. If we want to give them infinite food, incredible medicines they can’t hope to understand, and safety from any predators, we can. If we want to keep them in zoos, or hunt them for sport, we can. If we wanted them extinct, their habitats paved over with parking lots and solar cells, we could.

This kind of relationship, of complete domination over another, is the natural balance of power between a much more intelligent creature and a less intelligent one. It’s the kind of power an adult has over a small child, or an owner over their pet. The arrangement may or may not be beneficial to the weaker party, but ultimately, the more intelligent and powerful agent decides the future. A pet doesn’t get a say in whether they get spayed or not. 

Luckily, there are no other species out there running around that might be even smarter than us.

-

But that is changing.

Currently, the future belongs to humanity, for better or for worse. The planet and stars are ours to do with as we decide. If we want to drown ourselves in pollutants and a warming climate, we can. If we want to annihilate each other in nuclear war, we can. If we want to become responsible stewards of our environment, we can. If we want to build global abundance, limitless energy, interstellar travel, transcendent art and a rule of just law, we can. 

If a new, more intelligent species were to appear on Earth, humanity would surrender its choice over what future we want to make manifest. The future would be in the hands of the successor, and humanity would be relegated to a position no more admirable than the one chimpanzees inhabit today.

No such more intelligent species exist today, but they are being built.

Since its inception, the field of artificial intelligence has aspired to construct artificial minds and species as smart as, and then even smarter than, humans. If they succeed, and such systems are built, humanity will no longer be in control of the future, and the decisions will be in the hands of the machines.

-

If you don’t do something, it doesn’t happen.

This might seem so obvious it’s barely worth bringing up. Yet, you might be surprised how often people, probably including you, don’t really believe this. 

If we want the future to go well, someone needs to make it so. The default state of nature is chaos, competition, and conflict, not peace. Peace is a choice we must strive for, a delicate balance on the edge of entropy that must be lovingly and continuously maintained and strengthened. Good intentions are not enough — it demands calm, cooperative, and decisive action. 

This document is a guide to what is happening with AI, and offers a playbook for nudging the future into the direction you want it to go. It is not a solution, but a guide. A book cannot be a solution, only a person can.

What is AI? Who is building it? Why? And is it going to be a future we want? (Spoiler: No) There are so many things happening every single day in the field of AI, not to speak of geopolitics, that it seems impossible to keep up with, or to keep focused on what really matters: What kind of future do we want, for ourselves, and for our children?

We must steady our focus on this, and not let ourselves be distracted by all the noise and demoralizing nihilism pelting down on us from all sides. We need to understand where we want to go, chart a path there, and then walk this path. 

If we don’t do something, it doesn’t happen.

-

The default path we are on now is one of ruthless, sociopathic corporations racing toward building the most intelligent, powerful AIs as fast as possible to compete with one another and vie for monopolization and control of both the market and geopolitics. Without intervention, humanity will be summarily outcompeted and relegated to irrelevancy by such machines, as we did with our chimp cousins. 

A species of intelligent beings born from the crucible of sociopathic market and military competition will not be one of loving grace, and will have far fewer qualms about paving over humanity’s habitat with solar cells and parking lots. Despite humanity’s flaws, we still have a heart, we have love, even for our chimpanzee cousins, somewhere, sometimes. Machines of hateful competition need not have such hindrances.

And then that’s…it. Story over. Humanity is no more.

There is no one is coming to save us. There is no benevolent watcher, no adults in the room, no superhero that will come to save the day. This is not that kind of story. The only thing necessary for the triumph of evil is for good people to do nothing. If you do nothing, evil triumphs, and that’s it. 

If you want a better ending for the Human Story, you must create it. Can we forge a good, humanist future, one that is just, prosperous, and leaves humanity sailing into a beautiful twilight, wherever its final destination may lie? Yes. But if you don’t do it, it doesn’t happen. 

The path we are on is one of going out with a whimper, not of humanist splendor. It is embarrassing to lose out on all of the future potential of humanity because of the shortsightedness and greed of a few. But it wouldn’t be surprising. A human story if there ever was one.

-

The ending to the Human Story isn’t decided yet, but it will be soon. 

We hope you join us in writing a better ending.

- Connor Leahy, October 2024

But just because a system can solve a problem does not mean it will. 

The world’s governments could solve climate change and world peace, but will they? It’s kind of strange, isn’t it? We’ve built institutions that are powerful enough to solve global crises like these but they…won’t do it? But we built them! Why won’t they do what we want? And what does this tell us about AI?

Being Good

What does it mean to “be good”? Nevermind the philosophical hairsplitting, just intuitively, there is a sense of how there are some kind of things we want in the world that are “good”. Peace, happiness, freedom, prosperity, whatever. And we can think of anyone, or anything, that is acting to bring our world closer to those ideals to be “good”1.

But as you can surely tell just from thinking about this for 15 seconds…it’s really, really not that simple!

How many times historically have we learned the lesson that taking even a well meaning, but naive, conception of what is “good” and applying it forcefully is usually not a good idea? There are so many edge cases!2 How many disagreements do even completely sane and reasonable people3 have about what is truly “good”?

Well darn, this seems tricky.

So the first thing for us to notice is that, even under ideal circumstances, choosing what to do to “be good” is really hard4, because it involves all kinds of sticky questions about morality, emotions, compromises, etc.5

Being Good at Something

Now there is also another meaning of the phrase “to be good”, which is to be good at some specific skill. “To be good at basketball” or “to be good at programming” or “to be good at making money” or whatever.

Immediately, it feels like this is so much easier than what we were talking about before. What does it even mean to “be good”? But being good at basketball? That’s something we can understand, measure and improve!

And AI loves things that you can understand, measure and improve.

You’ll see again and again news articles breathlessly talking about what new game, or benchmark or task AI has aced, getting higher and higher scores, learning new skills, etc. This is exactly what we build AI to do, and it’s easy, fun and profitable! If you’re a company and want to buy AI, you’re buying it because it has some kind of useful skill you want to exploit, why else would you buy it?6

So by default, basically all types of progress we make on AI is making it better at doing some kind of thing, things that ideally are measurable and improvable! Whether that’s points in a game, money in a bank account or whatever.

Being Good is a Handicap

So when we put these two observations together, we notice an unfortunate situation: Making our AI better and more powerful at all kinds of skills is easy, but getting it to be good, to do what we want, is way harder! How do you easily measure “being good”? What would that even mean? Can you write that down in math, please?

And it’s even worse than that! Being good directly goes against being good at most things! It’s much easier to make money if you’re willing to lie and cheat.

Goodness is a constraint on our AI’s skills, it’s making it handicap itself to act more like we like. But this means that any advances we make on making the AI more powerful will almost certainly also make it better at doing things that are not good, such as lying or hiding its true intentions from us!7

AI, by default, will not have human emotions like us, why would it? It won’t feel bad about lying or breaking rules, it won’t feel bad about going behind your back. If that makes it win, that’s what it does, that’s what we built it to do! 

Being Good is Hard

So where does this leave us? It’s like we have a tug of war pulling us in two directions.

In the one direction we want to pull towards AI systems that “are good” and do what we think is good. On the other, we have all the work to make our AIs smarter, more general, better at winning and making money.

The former is handicapped by being so, so much harder and complicated, while the latter is simple, tractable, and each improvement makes you a ton of money, so it also has 100000x the funding compared to being good.8

The danger from AI does not come from some kind of “evil property” that must be “removed” from the AI. It comes from competence, competition and the inherit complexities of what we really want.

In this series, I want to build up to talking about one of the most important classes of forces shaping the world: Cartesian Demons.

Cartesian Demons are a common but advanced1 and dangerous class of Demons. They are entities/processes that pierce the Cartesian Cut (that “reason about the internals of your mind”) and are Demonic (they are adversarial to Human Values).

We will need to build up quite a bit until we can tackle them directly.

Since this is the first series I am writing on High Powerlevel™ concepts, we will also need to define a lot of common ground on how to reason about these kinds of things and what the lingo means.

We will cover:

• The basics of the Supernatural and Magick/Memetics as the Virtual and Computer Science (this post)

• The Origin of Demons in The Land Between and Pseudo-Bayesian Epistemology

• Daemons and garden variety Psychofauna

• (Lesser) Demons

• and finally: Cartesian Demons, their widespread effects, their ecology and how to interface with and Hunt them.

Introduction

It is tempting to view the world as made exclusively of physical Things. People, rocks, broccoli, paroxetine anhydrate2, whatever you encounter in day to day life.

But, in practice, while trying to model the world through sheer autistic dedication to physicality alone is admirable, it is not in fact a good and practical model of how many things in reality work. Great for stars and siege weapons, not so great for human life.

No, our practical lives contain another critical type of Thing: Software.3

The Supernatural and the Virtual

“Software” is a pretty recent term (coined in 1958), and while it is for sure the case that “software” is important to our modern, digital world, it is still usually classed firmly in the “physical, rational” category, and opposing it to that might feel a bit odd.

No, the traditional Human™ counterpoint to the Physical™ is the Supernatural™.

Here is a platonic example of a conversation between a Believer B and a Skeptic S:

• S: What is the Supernatural™?

• B: Well, it’s something that is “super” to the “natural”, something beyond the understanding and confines of physical reality and science.

• S: Ah, well, then this Supernatural™ is an epiphenomena that doesn’t interact with physical reality? Well, then I don’t care, because I live in physical reality.

• B: No no, it can interface with physical reality! Ghosts and magic can do all kinds of crazy stuff!

• S: Oh, it can interface with physical reality? So it is physical and therefor within the realms of science to study?

• B: No.

• S: No?

• B: *thinking of a slur to call S*

This is the strawman version of what people mean when they talk about the Supernatural™: an incoherent confusion to how causality, science and reality work.

But trying to explain everything that could be learned here away with such a strawman is epistemic cowardice!

It’s clear that the concept of the Supernatural has persisted throughout humanity’s entire existence and constantly pops back up, including in our modern, “rational” world. There is clearly Signal here, something is going on. Now that Signal might just be “people are stupid/confused”, sure, and there is a lot of that.

But I want to present a stronger, more coherent version of what it might mean to be talking about the Supernatural in a meaningful way.

Taking the Supernatural seriously

So what would a serious version of the concept of the “Supernatural” look like? What are people actually pointing at when they use this word?

As a first step on our Expedition, lets look at what people usually point to when in this context. From Wikipedia4:

“The term is attributed to non-physical entities, such as angels, demons, gods, and spirits. It also includes claimed abilities embodied in or provided by such beings, including magic, telekinesis, levitation, precognition, and extrasensory perception.”

Angels, demons, gods, spirits, magic, telekinesis, levitation, precognition and ESP. All things that would be pretty important to understand if they existed!5

Let us focus on Spirits as the most core example of a Supernatural phenomena.

Spirits are:

• Immaterial, they have no physical body

• Have various abilities, such as the ability to think, take actions, do tasks and/or speak

• Can be summoned by a Sorcerer by using various Arcane Rituals and Spells

• Can be hard to control and cause a lot of trouble if the Sorcerer makes a mistake

Take this excerpt from the Esoteric Tome™ known as “The Wizard Book”:6

[…] we conjure the spirits […] with our spells.

[…] much like a sorcerer's idea of a spirit. It cannot be seen or touched. It is not composed of matter at all. However, it is very real. It can perform intellectual work. It can answer questions. It can affect the world […]. [The methods] we use to conjure […] are like a sorcerer's spells. They are carefully composed from [symbols] in arcane and esoteric […] languages that prescribe the tasks we want our [spirit] to perform.

[…] Thus, like the sorcerer's apprentice, novice […] must learn to understand and to anticipate the consequences of their conjuring. Even small errors […] can have complex and unanticipated consequences.

Wow, yeah, that sure does sound like Spirits doesn’t it? Is this “Wizard Book” for real? It must be some kind of quack superstitious nonsense!

I assume you can see the punchline a mile away.

Read more

What do I want to talk about on this blog?

I could do the usual tech blog thing of only sticking to things I’m extremely confident in and can back up with tons of evidence and proof. But this comes at a cost, in that it greatly restricts the kinds of things I can talk about. There are many very important things in life that are worth talking about that are simply not easy/possible to discuss in a completely rigorous/formal way. And in the past, I have gotten signal that there is appetite for me to talk about things on the “edge”.

In this blog, I want to be able to talk about very high signal things that are on the edge of rigor, dipping into the territory of “mystical” or “esoteric” (I’m gonna define what I mean by that in a second). I think these are extremely real and useful things to be thinking about, but they aren’t quite strictly true in the most narrow definition of the word.

There are high signal/variance things that are not (yet) well digested into a rigorous framework that allows it to be consumable without using high-risk “mystic” tools of communication. I want to talk about some of those things, knowing the risk that many people will misunderstand or bounce off what I’m talking about. C’est la vie.

If you want to risk some of the more exotic Expeditions to the Far Lands1, read this post to help you understand how I write such posts and how you can extract useful (or at least amusing) signal.

I will be putting my experimental posts by default behind a paywall, mostly to filter readership for people actually interested in the topic. These aren’t really meant for a mass audience.2

Some of it is science, some of it is fiction, most of it is somewhere in the middle.

Join me for the Expedition.

Rigor vs Signal

What is Mysticism? Well according to wikipedia it means “becoming one with God or the Absolute, but may refer to any kind of ecstasy or altered state of consciousness which is given a religious or spiritual meaning”.

I will now do what’s called a “pro gamer move” and ignore that definition entirely.

When I refer to “Mysticism” or “Esotericism”3, what I mean is something like “topics that are on the edge of rigor/science/sanity, neither necessarily wrong nor right, unformalized, with lots of sharp intellectual edges; and the tools and styles of communication related to them”.

As you may be able to tell, that definition itself is far from formal.

Language and Truth

The more right4 you want to ensure you are, the more rigorous and the more constrained your language needs to be, both in terms of what topics you talk about and what language/symbols/grammar you use to talk about them.

If you want to ensure you’re never wrong5, we actually have ways to do that!

It’s called Logic6, and is a fascinating topic I’m sure I’ll end up writing about sooner or later.

But as I’m sure you can intuit, using formal logic and writing formal proofs is a huge pain in the ass, and is very constraining. If you’d try to have a normal day to day conversation exclusively in a formal proof language, you would not get very far.7

There are simply many, many concepts we talk about constantly that are not very formal. And, as an unsurprising corollary, humans are wrong and miscommunicate constantly.

So if I restrict my language to rigorous things only, there will be a massive amount of real signal I cannot communicate to people. Try forming an emotional bond with people without ever using informal or vibes based communication, good luck!8

Getting Signal from Wrong Things

The further along towards the informal end of the spectrum we go, the more we will get “wrong” things, and especially wrong explanations.

There is a common “nerd” reaction to informal/esoteric stuff: Dismissing everything that is “wrong” completely, as embodied by the platonic “reddit atheist” that encounters e.g. the topic of theodicy and therefore concludes that the common monotheistic worldview cannot be correct (which is a correct inference), but then goes on to dismiss that there is any signal or information or wisdom to be had. This is, as we say in the business, “monumental cope”. 

Obviously there is signal in ancient (or just “common”) traditions, social structures, history, practices, etc.9 It may not be the signal the practitioners themselves think it is, but, if you’re not an epistemic coward, there is a lot of things to be learnt.

As a concrete example, I’ve had certain back problems for most of my adult life, and have tried various forms of physical therapy for it, to little success. A little while back, I tried something called Rolfing. The explanation Rolfing gives for how it works is, to put it lightly, batshit crazy. It’s about “aligning the body’s energy field with the Earth gravitational field” or some shit like that. But, it made all my back pains go away and my posture has improved dramatically with no effort on my part.

So from this should we conclude that there actually is a “body energy field” that is being “aligned” to the Earth’s gravitational field?? Well, no, duh. But obviously something interesting is happening here.

Now we can argue about what the “actual” explanation for the phenomena is, whether it’s “placebo effect” or “they actually treat the fascia properly which almost no other physical therapy does, completely coincidental to the energy healing stuff”10 or “space aliens”, but you should be able to learn something and investigate this phenomena. 

Or think of Buddhism. In Buddhism there are ghost guys with tiny mouths that are really hungry (???). But obviously there is a shit ton of signal to buddhism and meditation11 , tiny mouth ghosts notwithstanding. Buddhism is, in my humble12 opinion, such a fascinating marvel, as the arguably most serious attempt to actually study and categorize mental phenomena and practices in a prescientific framework the world has ever seen.13

Now that doesn’t mean the Buddhists necessarily have the right explanations for everything they do and say, but I would trust a Tibetan monk or Zen master far more to actually get my brain to an “enlightened” state in one piece (if that’s what I wanted) than any modern scientist. They obviously know something.

The Abyss Stares Back

There is a common trope that I’ve seen play out a number of times. It goes something like this:

• Young person either grows up with no exposure to esoteric/religious topics, or grows up religious but rejects it.

• Person becomes a fervent defender of atheism/science/rationality/etc.

• As person gets older, they have some experience that cracks this perspective (psychedelic trip, witchy girlfriend, alt medicine success, cult indoctrination, atheists being very cringe…)

• Person is so overjoyed to be a bit more “open minded”, and gets so much enthusiastic support from the “(too) open minded community” that they flip (horseshoe theory style) to believing all kinds of esoteric stuff.14

• Person is now (permanently or temporarily) batshit crazy. Sad.

There is a cost to be paid in allowing more mystical reasoning into your epistemics. The more you move what you accept as “valid” to the mystical end of things, the less resistance you have to being wrong, being duped or just going crazy.

If you stare into the abyss, the abyss also stares into you. If you delve into the esoteric, the not-quite-true, then you will also likely start to become not-quite-true.

Of course, what this solemn warning15 really is, is a Deep™ way to say “Warning: skill issue”. 

Unintuitively, formal proofs are really easy to not fuck up. Every single step is completely formal and well defined, there is no misunderstanding.

The further you go to the other side, the more Skillful you need to be to not fuck yourself up and go crazy. This is why most mystic traditions present themselves as “paths”16, where you start with the easier things first, until you’ve built the skills to handle the more advanced stuff without going crazy/getting confused and shooting yourself in the epistemic foot.

One principled epistemologist and 100 witches

There is another risk associated with discussing esoteric topics, which is that most people who are interested in esotericism are interested in it for stupid reasons, or more prosaically, are really just not very good at it. Or, even more prosaically, simply suffer from some form of mental illness.

So you get the effect that if you are a principled Expeditioner on your journey to figure things out and ground things in proper rational worldview as much as possible, you will quickly find yourself in the company of 100 people that are doing it because they like LSD, have very strong opinions about polyamory and the age of consent or stopped taking their antipsychotic medication recently.

This is the classic “one principled libertarian and 100 witches” problem. The more you stretch what you tolerate, the more you let in people who want to (ab)use that tolerance for less than good reasons, or that are just crazy.17

There is no general solution to this problem, just something important to keep in mind if you poke your head down these paths: Most people treading them are Lost.

Tools of the Trade: How I write esotericism, and how to read it

I use a standard “toolbox” when I write about esoteric topics.18 There (obviously) isn’t one true correct way to talk about these kinds of topics, and every mystic will have their own tricks and style. Here are some of mine that I will be using and will be important to understand in order to understand how I write esoteric stuff.

Tool 1: Binding Words and Natural Language DSLs

Words were made for man, not man made for words.19

What do words mean? Are their meanings independent of their context? Well, no, but also yes. We have these things called “dictionaries” that tell us what words are “supposed” to mean. This is all well and good, and in most scenarios, I am greatly in favor of using words to mean what the consensus says they mean. But…

English20 is really just not a very good language to discuss esoteric concepts effectively. It lacks so many grammatical constructs21, words and even punctuation I wish I had access to to express finer grained concepts.22 

But alas, there is no perfect language. And this is a problem that crops up in formal languages, in particular programming languages, a lot, and one of the best ways to address this problem is the use of Domain Specific Languages (DSLs).

A DSL is a smaller, more specialized language embedded in a more general language, where some of the words can mean different things from what they usually mean in the more general language. This is common in programming, but also in all kinds of other fields, math being one of the worst/best offenders. The word “group” is a pretty common and very well agreed upon word meaning-wise, but oh boy does the meaning in math have ~nothing to do with what it commonly means. 

This is often a great initial hindrance to learning higher math, where you have words like “Group” and “Knot” and “Ring” thrown around like they mean something normal.

And I want to come out in full defense of this behavior!23

We don’t have enough words in English to cover every possible thing we want to talk about in every possible context, and so I am fully in favor of “rebinding” words to new meanings in different contexts, and this is exactly what I’m going to be doing a lot of.

To do this, I will be mercilessly exploiting a degree of freedom that is left in written English:

Whenever I take a word that is normally lowercase and I instead write it as Uppercase, this is syntax for “this word has been rebound to some context-specific meaning”. Which also means don’t interpret this word in its usual way, read the rest of the post to understand what it’s supposed to mean!

Hey, look at me!

Hey!

Read that again!

If you link a piece of one of my esoteric posts that includes high context DSL words like this out of context, then -1000 aura cursed upon you! Read the whole damn thing or I place a hex upon you! May your life feed be cursed by the following of 1000 vengeful spirits AI bots!

DSL Syntax Glossary

• Capitalization (e.g. “[...] they are Lost.”) = This word is bound to some other concept than usual, read the surrounding context to understand. Very often, it will deliberately be left undefined, in which case you are supposed to Think about what it is supposed to mean.24

• ™ symbol (e.g. “that is True™”) = Tongue-in-cheek. I’m joking/being cheeky about this word, don’t take it literally, I probably mean the opposite. Often meant to be mocking.

• Triple quotation (e.g. “““the truth”””) = Big airquotes, not to be taken too literally, like I’m aiming at something high that I am obviously not actually reaching. Not meant to be mocking.

• SpongeBob case (“e.g. bUt CoNnOr WhAt AbOuT [...]”) = I am mocking whatever hypothetical strawman is saying this. Imagine this being said in a really stupid and annoying voice.

• Footnotes: I really like footnotes, deal with it.25 They are usually either extended elaborations that are optional for understanding, or more often just meant to be amusing/funny side observations/commentary. They should be taken at least one level less seriously than the main text.

This syntax is not Complete, in the sense that there will still be lots of vagaries of natural language, and many not-serious things will appear even outside the designated Silly Syntax™.

Tool 2: Pointing in Concept Space

What is the purpose of using all these funky words and esoteric stuff? What am I trying to do? (Pseudo-) Mechanistically?

I like to imagine there is a huge “concept space”, where every possible idea or concept is a point (or cloud of points) somewhere in this super high dimensional space. 

A lot of science/formal writing is detailing a specific(-ish) point in concept space, laying out exactly where it is, what properties it has, what to do with it, etc. (One of) The purpose(s) of esoteric writing is to point to things26 that are not that crisp.

This can often look like me doing the conceptual equivalent of “Hey! You see this weird corner over here in your head? Haven’t ever thought about that area before, huh? Go poke around there!! I promise it’s worth your time!”

A common (and fair) reaction to this is: “Why don’t you just show me the thing you want me to find instead of dropping me off in the conceptual boonies and leaving me to fight for my conceptual life??”27

First, stop complaining, enjoy the free Travel. Second, here comes the Thinking Stick™!

The true answer is some mix of:

• We have limited time

• We have limited attention

• We have limited patience

• We have limited intelligence

• You have limited trust in me28

• We have other things we could be doing

• Skill issue (mutual)

Constructing a fine, crisp, “memetic bootloader”29 that you can just drop into an arbitrary person’s head that then actually gets them to the idea you want them to get (instead of, say, a nearby, seductive but wrong idea30) is incredibly difficult, resource intensive, and sometimes just impossible. With more time, trust, energy, etc, a shittier bootloader can go further, but there is basically no limit to the amount of intellectual resources you could invest in better and better explanations/bootloaders.

Making people understand things is hard. It’s a miracle of memetic infrastructure that over generations we have discovered so many explanations for really complicated stuff that actually work pretty well.

Where we’re heading, we don’t have the luxury of well paved memetic roads, so vague memetic hand gestures and memetic maps-sketched-on-the-back-of-a-napkin will have to suffice.

Hope you’re good at Pathfinding!31

Tool 3: Funny Word

I like funny word. Funny word good.

I use many funny word when non-funny word could be use. This make me bad academic.32

If there is a “sane” option for a word (e.g. calling it an “explanation”) and an insane option that is funnier (“memetic bootloader”), there is a tight calculus that needs to be run in my head to convince me not to say the funnier one.

This is especially useful as a “memetic signpost” to demarcate stuff like “hey, this is crazy high temperature territory, bring Sunscreen.”

Also it’s funny, and people remember things better when they are funny.

Tool 4: Chekhov’s Koan

A lot of “““wisdom””” tends to be delivered in a form that seems to be optimized to fulfil the karmic obligation of being very, very annoying when told to young and Un-Wise™ people, only for it to serve as a very ironic Checkhov’s Gun later.

The archetypical piece of “““wisdom””” is a Deep™ sounding, almost meaningless little quip, a thing an old wisened master says to their protege with a twinkle in their eye and a bamboo rod to the apprentice’s fingers. And yet, alas, eventually that very thing the master spoke of ends up happening in an ironic and unexpected way, and The Student Was Enlightened™.

The best model I have for what is going on here is that the Wisdom is an attempt to pre-register a thing the Student has yet to learn, and that can’t be taught directly, so that when it inevitably Happens, the Student will Learn. Many things are sadly of this shape. Many things you just gotta learn the hard way.33

Wisdom of this shape works as a “a vaccine that doesn't (usually) prevent you from getting sick that first time, but prepares you to recognize and fight off the virus every time after that.”34 It’s a claim to the Student that “this thing you just saw wasn’t just some weird one off, it’s an Important Pattern, pay attention!”

I call this the “Chekhov’s Koan”.

Classic examples might include:

• “The teacher appears when the student is ready”35

• “The more you know, the more you realize you don't know”

• “You must be the change you wish to see in the world”

• “The master has failed more times than the beginner has even tried”36

• “The fish is the last to discover water”

• “The axe forgets; the tree remembers”

• “To straighten the crooked, you must first do a harder thing: straighten yourself”

• "The best time to plant a tree was 20 years ago. The second best time is now."

• “The palest ink is better than the best memory”37

• “If you want to go fast, go alone. If you want to go far, go together”

• “Chance favors the prepared mind”

• “The usefulness of a cup is in its emptiness”

• “The map is not the territory.”38

All of these are extremely annoying/obvious/meaningless, until they very suddenly Mean A Lot.39

It might mean nothing right now for me to tell you:40

• “Your learning is accidental” or

• “If you only do what you like, you have no power” or

• “If you don’t do something, it doesn’t happen” or

• “It doesn’t matter what people say in private, it doesn’t exist” or

• “You are not your brain/mind/body”41 or

• “Don’t be stupid.”42

but I promise you, dear reader, if you keep reading43, and living a normal, dynamic life, that Chekhovian gun will go off sooner or later.

When you see me get very worked up about something that looks like an utterly irrelevant or bizarre detail, at least entertain the hypothesis that it might be a Chekhov’s Koan that will become important later in Act 3.

Just keep it in the back of your mind, for now.

Tool 5: It all adds up to normality

…and if it doesn’t, you fucked up!

In the Shamanic Journey44, you journey to the Underworld, discover all kinds of strange and wondrous Spirits, gain new insight and Wisdom, etc, and finally, the ultimate punchline is (spoiler alert): There are no spirits, there never were any spirits!! It’s just You, embedded in good ol’ Physics! 

The ultimate punchline to all of esoteric knowledge (and all other forms of knowledge) is that it all adds up to normality.

At the end of the day, no matter how far you ventured into the Underworld, eventually you pop back up out of the other side, exactly where you started. Apples didn’t stop falling from trees just because Einstein supplanted Newton’s theory of gravity.45

If you acquire Deep And Esoteric Knowledge™ and this leads to a fully radical different conclusion for everything about reality where you have to abandon your family and live in a commune to communicate with the space aliens that live in the quantum realm, you fucked up.46 Retrace your steps and try again.47

If you ever come away from one of my posts with your mind so utterly blown that you rethink your entire life and reality itself, you almost surely fucked up somewhere. The Ultimate Enlightenment™ shouldn’t feel epic, it should feel like “huh, oh yeah, of course, duh”. If it feels Deep™ and Radical™ and Meaningful™, you’re probably still Confused.

Ultimate gnostic enlightenment should feel boring as hell, because it’s just seeing boring reality for what it boringly already boring is.48

Use this tool as a guidepost and anchor, to double check both your and my sanity.

Conclusion

Thank you for reading the Beginner’s Guide™, the Expedition49 begins now.