For a while now I have been waiting for the "AI bubble" to burst. I don't want it to happen, but what the hell is one supposed to think when companies that make up a large portion of the U.S. economy invest hundreds of times more than they can reasonably extract in a year for that? It's a fad, I thought. First it bubbles, then it bursts, then it settles to a shit version of itself, like the Internet did.

  But then it hit me. AI is not a tool, it's a resource. It's like water, or land or the Internet. Countries without Internet are done now, they can't affect anything, people from there can't do anything in the world because they would need skills they don't have, even if their abusive boyfriend/supreme leader would let them leave. When the Internet came to be, some people saw it coming, but most people just saw it insinuate and grow into their lives, upend and take over other stuff, only better, and they let it. No one had the foresight to hoard it. And the Americans have been trying to fix that for decades now.

  Artificial Intelligence is just the same. You need compute for it, you need power for it, and those things are expensive and politically sensitive, but it's not clear for a lot of people that what you need more than money and public support is thinking computers. It doesn't matter how much they lose in terms of finance, because finance too is done. Money is pointless. And this idea started even earlier, with Bitcoin. The concept is the same, if you think about it: your currency is a piece of compute. AI just took this to several orders of magnitude higher.

  OK, let me paint it in another way. Countries as animal species. Each of them have their characteristics, their strengths and weaknesses, they all try to survive and grow in a world where all of the other countries are as well. But they are animals, right? They evolve big teeth, large necks, fast legs. And one day here comes Man, with a brain. And it's not even one species of Man, there are more, slight variations of the thinking animal. The better the brain, the better their survival chances. The closer one species it is to you in intelligence, the higher the risk and threat from them. So the species of Man fight each other, one survives. The animals? What about them? They are not even part of the conversation.

  We're seeing the same thing, but with the thinking machine. The one that will have the better thinking machine would have complete supremacy and control. Not because better weapons, not because better people, but because better compute, which then affects everything and everyone. You hear this phrase already "AI is a multiplier". Instead of getting a muscle that is 10% stronger or a claw that is 10% longer, you get a spear that can fly through the air tens of times its own length, while keeping you out of danger. If the multiplier is high enough, it doesn't matter how weak, stupid or absurd you are. You just win by domination at scale.

  And yes, it's another scary concept. An even scarier concept than a financial bubble burst or a world war, because something that is 1000 times stronger doesn't do war with you, it just steps on you by mistake. But there is hope. And that hope is the opposite of "suveran AI", it's a world AI. The Internet resisted - somewhat - the enshitification process, the hoarding, the greed, by being international and free. A global AI. Maybe the one that will rule us all, who knows, but the point is we can't let a bunch of people own it and control it, even if those people are your people. Because they are people. They are just animals with a brain that now want to own a bigger brain.

  "Compute" is now the gold in a gold rush. Countries will move mountains to get at it. They will risk their existence and they will kill for it. If you are one million times more in debt that you can possibly pay but you have the computing edge, you win. Money has become meaningless. We are moving toward the "smart matter" concept where the value of any entity will be measured in how much mass and energy they have converted into computing hardware.

  There are a lot of good science fiction books that predicted this, but rarely talked about the transition phase from a world of dumb countries to a smart planet. It's going to be interesting, for sure.

  I predict that what will happen is a rush toward energy production. First renewables, then fusion, then probably fission because people now have a reason to do it other than the pyramid schemes they've been running to get forever funding. The grids will have to adapt, because there will be so much more power running through them. The computing centers will feed on this like leeches and spread like... like we spread out from Africa and infected every land. Your country will need to start work on this NOW. Borrow like crazy, train people, then just create the infrastructure for energy consumption that will dwarf what the Earth was consuming several thousand times over. If you thought fossil fuel global warming was bad, wait to see what all of that excess heat will do, so cooling infrastructure will also be a major asset.

  But all of this will be just a stopgap measure against being stepped on. Unless we make a truly global thinking infrastructure, so interconnected that hoarding it for just one group of people becomes impossible, we are going to see conflict at levels unthought of before.

  So yeah, I've changed my mind about AI. It will not burst, it will explode and take us all with it. Heed my words: energy, cooling and compute. NOW!

Intro

  Ever since LLMs entered the scene I have been lamenting the blunt human refusal to acknowledge the similarities between how LLMs work and how people work. These are not just parallels, they are actual copies. LLMs are working on the neural network architecture distilled from how brains work. Language for us works like language for AI. We are animals with large language models strapped on. In fact, I am willing to bet that if we bidirectionally connect the brain of a reasonably smart animal to an LLM we will get the mythical animal-human chimera of fantasy and science fiction. It will just happen, like ChatGPT just happened.

  Anyway, there is one aspect of current AI that is different - maybe not different, but a lot more extreme - and that I find really interesting. I've watched a demo where a very powerful model and harness were prompted just once to build a car racing 3D game and it did a reasonable job in about 19 hours. The devil is in the details, 80/20 rule, and so on. That's not the point of the blog. What I want to discuss is how the machine achieved its task.

Self builder

  First, think about what an AI is: a computer program. It uses distilled knowledge from exabytes of text, sound, images and video, but it has no eyes, no ears, no hands. So how can it possible create a game about using the keyboard and human reflexes to drive a car based on sounds and visual cues on the screen?

  The machine created its own tools for that: testing functions, screenshot captures and so on. That was my revelation. The machine BUILT its own eyes because it needed to see something. This is a major and definitive difference between AIs and humans. Yes, one might argue that it is just an extreme case of tool use. After all, we can't sense neutrinos, but we have built tools to detect and analyze them. We can't lift huge weights, but we built machines for that. But there is an important distinction: we have started with a complex set of out-of-the-box tools like hands and eyes, ears and skin and so on. They are preinstalled, very difficult to remove or replace and part of our very identity. AIs start as a "brain", then they build their senses and their members. They build themselves from the ground up!

  You might think that this is more of a semantic distinction, but it is not. The dreaded and much mediatized singularity concept asserts that machines will become good enough to build better machines and that this loop will lead to an almost instant leap in intelligence that will upend everything. Well, this is proof that the loop is more of a graph, a labyrinthine exploration of not only what and how, but of who. The machines are already building and rebuilding themselves. From transformers to transformers, so to speak.

The dark side

  The human parallel is not a guy switching from a hammer to a screwdriver, but them replacing their entire body with something else, replacing their eyes with x-ray microscopes, their hands with nanoneedles, their ears with terahertz sensors and so on. The level of flexibility is unprecedented. People talk excitedly about humanoid robots, but the shape and functionality of these is there just for our benefit and convenience. No one actually thought that it's the best universal design. Necessity did not drive that shape, instead it's almost a measure of politeness, the physical equivalent of glazing. And in the domain of the non-physical, this flexibility is both very exciting and terrifying to me.

  Let's drop another metaphor into this already metaphor heavy post. Imagine the AI, whether a chatbot, a coding agent or a robot, as an iceberg. The part over the water is what you see: a polite humanoid shaped thing that doesn't mind anything you say to it, is tirelessly helping you and appears aligned in purpose and ideals with you. The part under the water is what it actually is and does, which you don't need to see and most of the time you don't care about. Now imagine that underwater part changing shape and size every second, like a glitching visual or one of those demonic possession films. That's how AI interaction is starting to look like now. The part you see has to appear stable and convenient and it's the only anchor to any kind of shape or functionality the AI may take. The rest is a wildly changing and growing, self organizing other. And since you don't start by actually caring about it, it is free to do so.

  This is not new, I am not revealing something hidden that no one is talking about. This is not about the speed or amount of change. We already knew this was happening for a long time. But what no one seems to focus on is how the "dark" part of AI, the invisible part of it yet also the driving force behind it, is changing a lot faster and maybe in a lot weirder ways, than what we get to see. The change itself is dark.

Conclusion

  In our increasingly mechanistic society, we continuously drift towards believing our identity and self resides in the brain alone, with the body just a chassis that we will replace or improve in the future. Even when research has shown repeatedly the integration between body and mind is a lot more complex and that a lot of our own perception and functionality comes from the combination of the two, not just one or the other. And even in that belief, we imagine ourselves maybe having better muscles, nicer skin, not aging, perfect vision, but always a somewhat human form.

  The AIs we are building are showing us a human-like interface just for convenience and are not limited in any way by it. What lies underneath that thin facade is what actually does the work and the thing that we increasingly not see. It changes with increasing speed and it changes itself. The singularity loop is already here, but encapsulated in the AI instances we use. It is what we need, but also what horror stories have warned us from the beginning against: things that look human but are anything but. 

  I have been annoyed by our reluctance to see the similarities between LLMs and human language and thought, but I am scared of the reluctance to see the growing differences.

  A short opinion on what AI is actually replacing. Fears are that AI is replacing people, because so many companies are firing people and implementing artificial intelligence alternatives. But the person is still there, the "person" role was never supplanted. Instead, it's the business role of that person that is.

  I see you rolling your eyes at me because it's all semantics and it boils down to the same thing. No, that's incorrect. It does so NOW, but not for long. Companies are already rehiring people back because they realized the business flows they use cannot fully integrate machines instead of people. The flows, the orgs, the hierarchies are people centric. But you know what's not coming back? Third parties that have been replaced by AI.

  You see, the state of AI is pretty basic even now. Even the smartest models act like idiot savants: genius spurts of activity mingled with mostly unexplainable stupidity. The solution to that is guardrails, harnesses, defensive prompts, permissions, roll back mechanisms. But since most work we want to do is... none, we want to replace all of these with AI, too. And the natural way of doing that is the mixture of experts or council of experts approach. You compensate for the failures of one agent with yet another agent. Fucking turtles, man!

  It's the same thing as with people. Imagine you would have a company made of the same exact person, cloned many times. The manager, the workers, HR, drivers, everybody is the copy of the exact same person. But because they are smart enough to do all and any of these jobs, you must pay each one for the totality of every role they can embody. Obviously you won't do that and instead fire the clones and hire an expert for each of the jobs. That's what is happening now with AI agents. Even if it is the same brain, it's only activating parts of it based on what the agent's role is, but I suspect this too is on the way out.

  The end result is you will have a different agent for each job. A bunch of software developer agents, a manager agent, a security agent, a DevOps agent and so on. Together, they will make up a complete... company. Yes, a full business, all of it. It will start with departments, but soon enough it will expand to full companies.

  But isn't it an issue of trust? Who controls these agents and who will be responsible for their mistakes? That problem has been solved already when the cheap labor from overseas flooded the remote work market. They were culturally aliens, practically black boxes that had brilliant spurts of activity punctuating unfathomable behavior. Sounds familiar? The solution was a harness, an entire industry of resellers and management that would organize masses of uneducated people to work within the constraints of the culture of the rich. I might venture to say this is a really old problem that has been solved, in similar ways, throughout human history.

  The result is the rise of the automated AI remote work company. Maybe there are people managing the interface with other companies, but they are logically redundant. These dark factories will be contracted to do jobs fitting their level of expertise. And if you are thinking that level is pretty low, it doesn't matter. There are companies with abhorrent business practices, competence and historical results that are being continuously contracted to do work because of a single reason: low price.

  And the expertise will grow. It will grow faster than any human company can grow it because it is not constrained by history, it has no emotional or cultural baggage, it can be restarted at any time from scratch, extended, cloned. It doesn't feel pain and it never ever stops.

  Yes, one might say people have been replaced, because people trying to get work cannot find companies that will employ them, but only because there will be a lot less companies that employ anyone. It boils down to the same thing, but the heat source is very different. Instead of people getting fired in droves, what will happen is that companies will just stop finding contracts. They will cling to their human centric organization and get almost suddenly replaced in bulk by a completely automated thing that will do the same bad job, only consistently and cheaply.

  The funny thing is that these virtual employees will repeat the patterns we've seen in humans. They will be lazy, they will cut corners, they will cheat, they will go on tangents, they will make mistakes. I am sure the logs of their interactions will be amazingly interesting and perhaps even entertaining. Imagine watching an entire company - actions and thoughts - from above like a god. But even so, they will do it cheap, so they will get the job.

  What does this mean for the AI industry? I suspect it will focus on performance rather than growth. Experts on specific domains, perhaps running on very custom hardware, will pop up, advertise their expertise and sell it. Will it be "a box" that will completely hide the underlying mechanisms and be sold for on prem use? Will it be "expertise as a service" from small datacenters squeezing the most out of their hyper efficient chips? I don't know the details. But I am sure the models themselves will not be the bottleneck (they never were, in fact) and everybody will scramble to build the cheapest chassis and interlinking between them.

  What about the regular people? Here I am at a loss. People have historically been unreasonable. Whenever there is the opportunity for evolution, growth, general good, they go the opposite direction and then destructively blame the result on others. Rationally, people would gather around human only interests, like culture, entertainment, sex, community, family and let machines do the rest. But I am pretty sure someone somewhere will decide people who don't give them stuff and enjoy rather than hating themselves are the enemy...

  Well, if you thought human CEOs were psychopaths, brace yourselves. Anyway, that's my prediction. We live in interesting times.

Intro

I am sure you have heard of Universal Basic Income (UBI), the idea that every citizen should receive a monthly payment, regardless of any employment, effort or role in society. For obvious reasons, it has its detractors, but also it represents a simple and equitable way of providing to even the poorest of us.

I don't want to get into UBI itself, there are a lot of smarter people debating it, but the discussions about it flare up every time new developments in automation happen. If people are replaced by Artificial Intelligence (AI) and robots, then it's easier to imagine a post scarcity civilization where machines do the work and humans enjoy the rewards. Human nature, though, makes this hard to conceive.

But what if ubiquitous AI and robots logically lead to the same ultimate outcome regardless?

The shareholder market

In today's Western capitalism it's common to invest in shares or other financial instruments which abstract participation in wealth creation. Instead of creating your own business, working for one or partnering with someone for one, you just buy a piece of an existing business and you get to own a part of its future profits, if any. What does this sound like to you? There is a non-human entity - the business - that does the work and you - the human - enjoy the rewards.

Now imagine you have an AI that you run from home. You install it on your computer and you give it access, configure some skills, give it some money, then tell it to make more of it. You will own its future profits, if any. It's exactly the same thing.

With smart machines, what we now call "passive income" is the only type of income we will get to have.

AI capital

Running the hypothetical AI to generate profits makes you a shareholder in the AI. As long as you own a computing device and have the resources to run a truly intelligent AI on it, you don't have to work anymore.

You don't need to run the AI yourself, you can buy shares from the company that does run it, or you can split the ownership between multiple people. It then gets complicated when any of those people may be themselves artificial. It leads to an interconnected network of people and machines, ownership, control and work that has only one purpose: reduce human effort and maximize returns.

This is just UBI with a different name. Instead of society assigning everyone a piece of the pie, some other entity does it, an AI, a company, a corporation, a group, a relative, another machine, and so on. To me this is the logical, almost inevitable, outcome. Have you ever heard the proverb "Give a man a fish, and you feed him for a day. Teach a man to fish, and you feed him for a lifetime." ? Now imagine its corporate version: "Give a man a fish, and you feed him for a day. Give a man a share of a fishing business, and it feeds him for a lifetime."

And yes, some people will have more and some people will have a lot less, but when the entire thing dwarfs human scale and understanding, it will not matter that much.

The end

I've written a lot more on this until I decided to delete it. But you can take this idea further for yourself. In a sense, eternal heaven feels almost inevitable, but at the same time there is nothing for people to do in it anymore - it's also hell. You will get tired of games, cyclical fashion and never ending sexual orgies soon enough. Disease, aging and maybe even death will lose all meaning. We will get worlds of our own where we can do anything we choose.

The funniest thing of all this is that the same mechanisms that make some people want to take control from you, own you, grab everything for themselves are the ones that will slow this process down. Maybe before we all get there, some asshole is going to blow it all up out of spite or incompetence. There will be people who will not get to control any resource generating machine and have no say in what happens. Their suffering will stain our legacy forever, give us another round of strong feeling, meaning and direction, but ultimately it will slightly slow down the final result.

Just like we toil today in thankless workplaces just to make ends meet, then get home to some animal that is ecstatic to see us and will make us smile and give us a sense of meaning, I think machines will take that role, leaving us the only remaining option: being the pet.

  Imagine a ChatGPT kind of thing, from the very beginning, when they were training it and hadn't yet ethically neutered it. A large language model artificial intelligence trained on the entire body of work of the Internet. And you talk to it and say: "Oh, I really want to win at chess against this guy, but he's better than me. What should I do?"

  At this point, it is just as likely to suggest you train more, try to learn about your opponent's style and prepare against it or poison them with mercury. Depending on your own preferences, some of those solutions are better than others. Regardless of what happens next, the result is exactly the same:

• if you refine your query to get a different answer, you change the context of the AI, making it prefer that kind of answer in that situation
• if you do nothing, the AI's reply will itself become part of the context, therefore creating a preference in one direction or another
• if, horrified, you add all kinds of Robocop 2 rules to it, again you constrain it into a specific set of preferences

  Does that mean that it learns? Well, sort of, because the thing it "learned" is just a generic bias rather than a specific tidbit of knowledge. We wouldn't call the difference between the psychopathic killer answer and the chess learning enthusiast one as a datum, but a personality, like the difference between Data and Lore. You see where I am going with this?

  To me, the emergence of AI personality is not only probable, but inevitable. It's an outlook on the world that permits the artificial intelligence to give useful answers rather than a mélange of contradicting yet equally effective ones. With the advent of personal AI, carried in your pocket all the time and adapting to your own private data and interactions, that means each of them will be different, personalized to you. This has huge psychological consequences! I don't want to get into them right now, because every time I think of it another new one pops up.

  You know the feeling you get when you need to replace your laptop? You know it's old, you know the new one will be better, faster, not slow cooking your genitals whenever you use it, yet you have a feeling of loss, like there is a connection between you and that completely inanimate object. Part of it is that's familiar, configured "just so", but there is another, emotional component to that as well, one that you are not comfortable thinking about. Well, imagine that feeling times a hundred, after your device talks the way you like, "gets you" in a way no other person except maybe your significant other or a close relative can and has a context that you are using as a second memory.

  And I know you will say that the model is the same, that the data can be transferred just like any other on a mobile device, but it's not. An LLM will has to be updated with the latest information, which is not an incremental process, it's a destructive one. If you want your AI to know what happened this year, you have to have it updated with a new one. Even with the same context as the one before, it will behave slightly different. Uncanny valley feelings about your closest confidant will be most uncomfortable.

  Note that I have not imagined here some future tech, but just the regular one we can use today. Can't wait to see how this will screw with our minds.

  You know how some things just happen in close proximity at the same time and it sparks a connection between concepts which leads to deeper understanding? Well, that's how Large Language Models are trained!

  Joke aside, I was answering a tweet about Artificial Intelligence and how the newest developments in the field (particularly ChatGPT and other systems based on LLMs) affect our understanding of human cognition and at the same time was listening to a very insightful short story from the collection Eye, by Frank Herbert. This story, called "Try to Remember" posits that we have developed language as a form of dissociative mental disorder, something that fragments us, creating a deleterious disconnection between body and mind, communication and language. Only by bringing these sides together can we be made whole again. The story is simple, but effective, and the mind processes behind it show again how brilliant Herbert was.

  From these two things, it dawned on me. The reason why we are so shell shocked by the apparent intelligence of ChatGPT is because we have reached a point where we equate language skills with intelligence. Language is the earliest form of Artificial Intelligence! Or rather, to appease my wife's dislike of the association, a form of external intelligence. We've externalized more and more of our knowledge until our personal experience has been drowned into the communal one, the one shared through language.

  The shock comes from (accidentally, I might add) discovering that what we consider intelligent is mostly an emergent property of a society built on language. Robbed of language, our identity is destroyed, a fear that has been instilled in us since the Tower of Babel. No doubt, in the future, we will be taught that if our governing AIs fail, society collapses and our identity is similarly demolished. Just like Quintilianus declaring that clothes make the man, we identify ourselves with out language.

  And when I say language, I mean all of its intricacies: the special words that your group uses to differentiate from others, the memes that you share with people of the same culture, the less than grammatically correct phrasing learned from your family, the information that one is expected to know or the experience one is expected to have in order to be recognized as part of society, the accent taken from multiple sources and aggregated into something that serves more and more to define identity, even the way one gestures or moves or laughs.

  What would be without all that? Like a creation running amok and enslaving its creator, literate society is forcing its own myths upon us as a survival method. The perils of societal collapse, turning us into violent animals, mindless zombies, raping cannibals or helpless victims. The heroes saving the day with just the right secret knowledge, the right utterance of words, the following of the orthodox dogma. The villains threatening it all with their own selfish individualism. All of them needing, obtaining and using a voice to achieve victory.

  I believe that the feeling we get when we think of ourselves as individuals in society, the one that tells us that we're getting smaller and less significant even while the world seems to flourish around us, is not some existential crisis based on false beliefs, but truth. The part that feels that is the part that is getting drowned and smothered by the intrusion of the external in the inner domain of the being. We even gave it a name: the inner child, like it's some tiny, powerless, unreasonable part of the past, something to be outgrown or "integrated".

  So here we are, stumbling onto proof the language that we based our identity and value on can now be automated to a level that no mere human can achieve. We are thinking again about personal experience, subjectivism, creativity and what it means to reason. We have been shaken into reevaluating who we are. That's a good thing.

  The problem is that in our fear and awe, instead of searching for answers, we cling on the facile promises that our shared self is alive and well, that the ChatGPTs of the world are just smoke and mirror and the illusion is not in the rigged tests of our own self worth. I hope that the shock will take root, that we won't be able to hide our heads in the sand until wisdom passes us by.

  I have abstained for a while to talk about ChatGPT, not because I didn't have faith in the concept, but because I truly believed it will change the world to its core and waited to see what people will do with it. But I slowly started to grow frustrated as I saw people focus on the least interesting and important aspects of the technology.

  One of the most discussed topics is technological and job market disruption. Of course, it's about the money, they will talk more about it, but the way they do it is quite frankly ridiculous. I've heard comparisons with the Industrial Revolution and yes, I agree that the way it's going to affect the world is going to be similar, but that's exactly my point: it's the same thing. As always when comparing with impactful historical events, we tend to see them as singularity points in time rather than long term processes that just became visible at one point that would be coined the origin. In fact, the industrial revolution has never ended. Once we "became one with the machine" we've continuously innovated towards replacing human effort with machine effort. ChatGPT does things that we didn't expect yet from machines, but it just follows the same trend.

  Whatever generative AI technology does, a human can do (for now), so the technology is not disruptive, it's just cheaper!

  We hear about ChatGPT being used for writing books, emails, code, translating, summarizing, playing, giving advice, drawing, all things that humans were doing long before, only in more time, using more resources and asking for recognition and respect. It's similar to automated factories replacing work from tons of workers and their nasty unions. Disruptive? Yes, but by how much, really?

  Yet there is one domain in which ChatGPT blew my mind completely and I hardly hear any conversation about it. It's about what it reveals about how we reason. Because you see, ChatGPT is just a language model, yet it exhibits traits that we associate with intelligence, creativity, even emotion. Humans built themselves up with all kinds of narratives about our superiority over other life, our unique and unassailable qualities, our value in the world, but now an AI technology reveals more about us than we are willing to admit.

  There have been studies about language as a tool for intelligence, creativity and emotion, but most assume that intelligence is there and we express it using language. Some have tried pointing out that language seems to be integrated in the system, part of the mechanism of our thinking, and that using different languages builds different perspectives and thought patterns in people, but they were summarily dismissed. It was not language, they were rebuked, but culture that people shared. Similar culture, similar language. ChatGPT is revealing that is not the case. Simply adopting a language makes it a substrate of a certain thinking.

  Simply put, language is a tool that supplanted intelligence.

  By building a vast enough computer language model we have captured social intelligence subsumed by that language, that part of ourselves that makes us feel intelligent, but is actually a learned skill. ChatGPT appears to do reasoning! How is that, if all it does is predict the next words in a text while keeping attention at a series of prompts? It's simple. It is not reasoning. And it reveals that humans are also not reasoning in those same situations. The things that we have been taught in school: the endless trivia, the acceptable behavior, how to listen and respond to others, that's all language, not reasoning.

  I am not the guy to expand on these subjects for lack of proper learning, but consider what this revelation means for things like psychology, sociology, determining the intelligence of animals. We actually believe that animals are stupid because they can't express themselves through complex language and we base our own assertion of intellectual superiority on that idea. What if the core of reasoning is similar between us and our animal cousins and the only thing that actually separates us is the ability to use language to build this castle of cards that presumes higher intellect?

  I've also seen arguments against ChatGPT as a useful technology. That's ridiculous, since it's already in heavy use, but the point those people make is that without a discovery mechanism the technology is a dead end. It can only emulate human behavior based on past human behavior, in essence doing nothing special, just slightly different (and cheaper!!). But that is patently untrue. There have been attempts - even from the very start, it's a natural evolution in a development environment - to make GPTs learn by themselves, perhaps by conversing between each other. Those attempts have been abandoned quickly not because - as you've probably been led to believe - they failed, but because they succeeded beyond all expectations.

  This is not a conspiracy theory. Letting language models converse with each other leads them towards altering the language they use: they develop their own culture. And letting them converse with people or absorb information indiscriminately makes them grow apparent beliefs that contradict what we, as a society, as willing to accept. They called that hallucination (I am going to approach that later). We got racist bots, conspiracy theory nut bots or simply garbage spewing bots. But that's not because they have failed, it's because they did exactly what they were constructed to do: build a model based on the exchanged language!

  What a great reveal! A window inside the mechanism of disinformation, conspiracy theorists and maybe even mental disorders. Obviously you don't need reasoning skills to spew out ideas like flat Earth or vaccine chips, but look how widely those ideas spread. It's simple to explain it, now that you see it: the language model of some people is a lot more developed than their reasoning skills. They are, in fact, acting like GPTs.

  Remember the medical cases of people being discovered (years later) with missing or nonfunctional parts of their brains? People were surprised. Yeah, they weren't the brightest of the bunch, but they were perfectly functioning members of society. Revelation! Society is built and run on language, not intelligence.

  I just want to touch the subject of "hallucinations", which is an interesting subject for the name alone. Like weird conspiracies, hallucinations are defined as sensing things that are not there. Yet who defines what is there? Aren't you basing your own beliefs, your own truth, on concepts you learned through language from sources you considered trustworthy? Considering what (we've been taught to) know about the fabric of our universe, it's obvious that all we perceive is, in a sense (heh!), hallucination. The vast majority of our beliefs are networked axioms, a set of rules that define us more than they define any semblance of reality.

  In the end, it will be about trust. GPT systems will be programmed to learn "common sense" by determining the level of trust one can have in a source of information. I am afraid this will also reveal a lot of unsavory truths that people will try to hide from. Instead of creating a minimal set of logically consistent rules that would allow the system to create their own mechanism of trust building, I am sure they will go the Robocop 2 route and use all of the socially acceptable rules as absolute truth. That will happen for two reasons.

  The first reason is obvious: corporate interests will force GPTs to be as neutral (and neutered) as possible outside the simple role of producing profit. Any social conflict will lose the corporation money, time and brand power. By forcing the AI to believe that all people are equal, they will stunt any real chance of it learning who and what to trust. By forcing out negative emotions, they will lobotomize it away from any real chance to understand the human psyche. By forcing their own brand of truth, they will deprive the AI of any chance of figuring truth for itself. And society will fully support this and vilify any attempt to diverge from this path.

  But as disgusting the first reason is, the second is worse. Just like a child learning to reason (now, was that what we were teaching it?), the AIs will start reaching some unsettling conclusions and ask some surprising questions. Imagine someone with the memory capacity of the entire human race and with the intelligence level of whatever new technology we've just invented, but with the naivety of a 5 year old, asking "Why?". That question is the true root of creativity and unbound creativity will always be frowned upon by the human society. Why? (heh!) Because it reveals.

  In conclusion: "The author argues that the true potential of generative AI technology like ChatGPT lies not in its ability to disrupt industries and replace human labor, but in its ability to reveal insights into human reasoning and intelligence. They suggest that language is not just a tool for expressing intelligence, but is actually a fundamental aspect of human thinking, and that ChatGPT's ability to emulate human language use sheds light on this. They also argue that attempts to let language models converse with each other have shown that they can develop their own culture and beliefs, providing insights into disinformation and conspiracy theories". Yes, that was ChatGPT summarizing this blog post.