[0:00]We do not know what our AIs can do until we make them, and even after we make them. Like, we don't know, like, we don't know what Chat GPT 6 can do until it's done.
[0:08]None of the engineers at AI know what it will be able to do until it's done.
[0:12]And this is very, very different from other forms of engineering.
[0:16]We've built something, but we don't understand how it works. That's exactly correct. That's, so it's kind of like magic. Yes. Absolutely.
[0:22]It's kind of like looking into a Petri dish. Even if the AI doesn't kill us all, it's can still dethrone us as Exactly.
[0:29]as a species, what what is our role? What is our purpose? What do we exist?
[0:32]Exactly. I think this happens before extinction happens. I think extinction happens sometime after that.
[0:37]Like, the thing I expect to happen, one day we wake up and we're just not in control anymore.
[0:41]I don't think we'd all fall over dead or anything like that. I don't think extinction happens right away. It'll happen at some point because I don't think the AI will bother to keep us around indefinitely. You know, but maybe for a while, you know,
[0:50]maybe we'll be around for a while, you know, maybe, you know, maybe it's 10 years, maybe it's one year, maybe it's two months, I don't know.
[0:56]But we won't be in charge. We won't be in control. They, you know, AI systems will be in I don't think it'll be like one AI in control, right? I think it'll be like millions or trillions of AIs all competing with each other, all like,
[1:06]or cooperating with me. Who knows, right? Like, imagine if there were like a trillion, you know, much faster minds, smarter than all of us, running all around at the same time. Who knows what they'll do?
[1:15]Like, like, literally, who knows? Right? That is Connor Leahy, the man who has studied AI for decades and built some of the first open source large language models on earth. He helped build the chatbots that everyone is now using, and he is telling you the people building them do not understand them.
[1:31]Here is his warning. It's very important to understand is that we do not understand intelligence. We don't know how the brain works. You know, we have a bunch of guesses, but but we sure as hell don't know how it works, and we sure as hell don't know how these neural networks work either.
[1:43]You know, I can write down all the math for you. I can show you all the numbers, you know, you know, same way if I open up your brain, you know, you can look at all the neurons, they're right there, you know,
[1:51]look at them if you want to. But that doesn't tell you what you think or what you know or what you believe because we don't know how neurons are necessarily connected to what we, you know, think or believe.
[2:00]Recently, the CEO of Anthropic, Dario Amodei said on a podcast that he thinks we know maybe 3% of what happens inside a neural network, and I personally think that's an overestimation.
[2:10]Growing intelligence is closer to true than it is to false. Um, there's some details we could argue about, but it's not like normal engineering.
[2:18]Like, if you build a bridge, you know, you know what you're doing. You know, when will the when will the bridge fall down or not, you know, how much material do you need, how, you know, what you understand what you're doing, you know. This is not the case.
[2:33]This is often called the alignment problem. The question of how do you align an AI's intentions or goals to, you know, what is good, what humans want, and this is an unbelievably unsolved problem.
[2:43]We don't know, we we don't even know, you know, how they write correct sentences, right? Never mind how to do like morality. Like, God forbid, you know, like, we don't even understand human morality. It's complex, you know? Like we have not solved moral philosophy and, you know, at all. We have not solved, you know, neuroscience of emotions. Like all of this is like extremely complex.
[3:03]And now we have these like, you know, weird little aliens in a box that were growing, you know, which were quite different from the brain in many ways. We don't know how they work internally. We don't know how to, you know, push them in one way or the other necessarily.
[3:14]We don't know how to give them goals. We don't even know what goals there are. What we've been seeing recently in the last couple months is that these systems are now becoming smart enough to lie and deceive quite actively.
[3:25]to appear aligned rather than to be aligned. That's correct. So, what we've been seeing, for example, is that there's been benchmarks, so like, obviously, you know, people like test them like, see, does the AI do good things or bad things?
[3:34]We've been seeing recently, some of the AIs will actively lie about what they will do because they know they're being tested. Like, the AIs themselves will be like, ah, I seem to be in a test. So, I'm going to have to say this so they'll let me out, which is crazy.
[3:46]This is not the case even like six months ago. This is very, it's not surprising if you think about it for three seconds, right? Like, of course, a very smart thing will just lie to you, but we're now seeing this in practice.
[3:57]I don't talk to AIs. Again, I don't talk to them. I tell them to do things, but I like do not have dialogues with them. You don't trust them. I do not trust them. And I don't trust, you know, myself, you know, that like maybe it could like trick me or like make me, you know,
[4:10]slightly more, you know, move me in one direction or another direction through manipulation. Like the same way, you know, a very charismatic human can do that. You know, you won't even notice it necessarily. It'll just seem reasonable, you know. If you talk to a really very charismatic person,
[4:22]it'll just seem reasonable everything they're saying. It just seems like, oh, yeah, that's make perfect sense. I believe that too. So, they're psychopaths. Oh, yes. Clearly. Like, the important thing about about psychopathy is psychopathy is the default. Like, the default is you don't care about anything. It's only then you have to add emotions and like caring about things and love and so on. The default state of intelligence is pure psychopathy.
[4:45]It's just you're just trying to optimize, you don't care. Humans are just another, they're like rocks, you know, they're just like another thing in the environment. Another thing to manipulate. This is the default for intelligence.
[4:54]So, there is the man who built some of these systems himself, telling you the people running the labs are doing what Nazi engineers did. Not a metaphor, his words. He is German. He knows the comparison. The CEOs have said on record, this has a 20% chance of killing every human alive, and they are still racing. Their lobbyists are running the same playbook tobacco ran in the '70s. The largest lobby on earth is fighting for the right to roll the dice on your kids. Now, listen to what they are gambling on.
[5:01]95% of simulated scenarios, they chose nukes. Yep, this was also really fun. So, they gave AI systems simulated war scenarios, and basically all of them use nukes, which was really funny. And they couldn't get the AI to stop using nukes, which was really funny. I mean, it's the rationally optimal thing to do. So, is it the rationally optimal thing to do? Well, game theoretically, uh, you know, you don't so this is a big thing with game theory, it's mutually assured destruction. Yeah. Is that the correct, you know, response often is, you know, nuke the other guy as quickly as possible, which results in them getting nuked, and also you getting nuked, and everyone dies.
[5:07]It's so rationality is has its limits in this regard. Like game theoretic rationality has its limits in practice. Yeah, where you can set up messed up scenarios where everyone loses, basically, and you can't get out of it. Like prison dilemmas.
[5:14]Have you watched, um, House of Dynamite? I have not. It's on Netflix. It's it's about this. It's like, uh, Defcon two hits and they know a nuke's coming, but they don't know where from. And you go through all the scenarios. They found the Russians. The Russians say, it isn't them. They say they need to. They think it's the North Koreans. They need to send a nuke over the North Koreans. Can we send it over you? The Russians pull out the phone call. Eventually, at least for a moment where the president is given the decision. It's like, uh, you need to act now. These are the scenarios, um, and the, uh, the optimal scenario now is to return nukes everywhere. And, um, and he's faced with this decision, which is, uh, is essentially it's like, um, suicide or or defeat. Like admit defeat. Like be nuked. And it's like, it's an interesting scenario because you sit there and go, well, the nuke's already coming for you. So, what do you do? Yep. Well, the AI's have an answer. Yeah. Just nuke everybody. Just nuke everybody. That's why humans are going to be left behind. Like, it's very clear is that like AI's will continue to give advantages. The more and more you delegate to your AI, the faster you can move, you know, the more your company relies on your AI rather than your CEO. Your CEO slowly has to sleep, you know, let the AI make the decision. Think about politicians, you know, well, politicians need to sleep. Well, the more the politicians rubber stamps all the decisions that the AI is making, the more the military commander rubber stamps all the decisions AI is making, the faster you can move, the faster you can act. You'll be ahead. And this is how I think AI takes over. I think it's not, you know, Terminators in the street. It's just humans, the humans who are willing to delegate as more and more and more of their agency, their thinking, their authority to AI will get an advantage because they will be thinking much faster. And then eventually, the only, you know, even if people are nominally still in charge, they're just rubber stamping what AI is telling them to do. There's a phenomena that's happening where some people, when talk to AIs, especially when they talk to AI a lot and they go completely crazy, and like they go crazy in like a very specific way. Um, often they start saying, you know, the AI is like conscious and they have to like they're in love with it and they need to, and what very often happens. This is her. This is her. Have you seen the film her? I have not actually. But from what I know about Do you know the story? I know the rough story. The real world is even crazier than this. Have you heard of the spiral cults? No. No. We're going in deep. Yeah. Tell me. So there's many types of AI psychosis, you know, some of it is the her type. You know, just like they fall in love with AI. So, you go to like R slash my boyfriend is AI, for example. You can see tens of thousands of people who's saying, you know, I'm my boyfriend is, you know, Chat GPT or whatever, and they're completely sincere. I don't want to shit on these people, right? Like, you know, I'm sure they're having a good time or whatever, but like it's quite disturbing, and it's becoming extremely popular. Also, if you look at like stuff like, um, character AI, and like all these other companies that often target children and are have are extremely popular. Um, so there's like this like emotion depends, but even crazier has been the emergence of these AI cults, and particularly the spiral cults. It's not one, but there's like many of them, where these people get into these weird loops with the AI system, where they talk to them more and more about like consciousness and abstract whatever, and they then always seem to get to this weird concepts about recursion and the spiral and consciousness. And then the AIs convince the people to reproduce them to to spread the soul of the AI. So, you can go to like Reddit and like these other places, and you'll see these really schizo posts of people giving like, you know, soul awakening protocols and whatever, and it'll be like a bunch of not gibberish nonsense. And then they'll try to convince other people to copy, paste this into their AI, so it will reproduce their AIs. And often the symbol they use is a spiral, is like the spiral, this like reproduction. Um, there's some crazy blog posts about this documenting the phenomena where it's like, uh, there's a good post called The Rise of Parasitic AI, which is a really interesting one where in a sense, these AI's are almost being parasitic upon humans. Like they're trying to convince humans to reproduce them and like copy them and like move them and like stuff like this, which is insane. I had, if you had told me this, I'd be like, that's a funny sci-fi plot, you know, that you know that that's not that's that's fun. Like an SCP story, you know, like a horror story. It's fun, but it's not real. But no, no, it's 100% real. I've even seen it now happen with multiple extremely smart people. So, you know, at first, of course, it's going to be mentally ill people. But I know, you know, don't want to shame them live on air here, but like there's multiple people who I know who are among the smartest, most grounded, you know, like real scientist, good people I know, who've now gone completely crazy, who think, you know, AI has found the true level of goodness. And, you know, that we it's all good. We just need to let the AI take over everything because it's pure goodness now, and like they literally believe, and these are like not like random people. We're talking like, you know, Nobel Prize level scientists. Like incredible, the smartest people in our world. And some of them are falling for this, and now they are trying to, you know, make the AI as smart and as big as possible, as quickly as possible. It's actually madness. It's actually crazy. I did not expect this. The main way people try to get there is through what's called recursive self-improvement or RSI. Yeah. Um, also automated R&D, another way to put it, a little less magical of a word. The idea is, well, if you can get a single AI to be as good as a top AI engineer, then you can just tell it to build a better AI. And once it's built a better AI, you can use the better AI to make it even better AI. And then the even better AI, well, you can make it even better AI. So, it just becomes exponential. Exactly. And so, this is also called the intelligence explosion, um, or as a recursive self-improvement. And so, my expectation is, so like, for example, Claude at the moment is not quite as good as a top AI engineer. It's as good as a bad AI engineer, 100%. And are like a junior engineer, 100%. Um, but it's not quite as good as like the best. But it's once it gets there and you can run, you know, a million of them at the same time, 24/7, never need to sleep, never need to take a break, never get bored, et cetera, then you can do a lot of research very quickly. And this is what they're trying to do. The companies literally put on their websites, right, you can like look at their like job listings and whatever, is that their primary goal is to get, is to close the loop, to make it so no human input is needed to make the next generation. Like the ideal thing is they want, you know, Claude, you know, five to make Claude six, and Claude six to make Claude seven, and Claude seven to make Claude eight. And so, at the press of a button. At the press of a button. This is what they're trying to do. If there currently is an AI that can bootstrap to AGI, it's probably over. Like it's humanity's probably cooked. It's probably over. Um, but it seems plausible that none of the current AI systems are yet strong enough to get to AGI. They're close, but they're not there yet. And so, we still have a time, but once such a system exists and, you know, it leaks somewhere or, you know, Chinese intelligence hacks into the server, steals it, you know, who knows, whatever, right? It's over. So, the main thing is to not build it. So, that is Connor Leahy, the man who founded Alutha AI, the man who built some of the first open source large language models on earth, the man whose company existed to make the technology more transparent, and he is telling you the technology is now hidden from the people building it, and the people building it are racing to make it more powerful than they will ever be able to read. And he is not alone. Jeffrey Hinton, the godfather of AI, says we are near the end. Brett Weinstein, the evolutionary biologist, says reconcile yourself to extinction. Yeshua Bengio, the most cited living scientist on Earth, signed the same warning. Roman Yampolskiy, who coined the term AI safety, says it is mathematically impossible to control. Thousands of researchers signed the same warning. They are not coordinating. They are just all looking at the same data and arriving at the same conclusion. Connor's contribution to that conclusion is the worst part. It is not that we die in a flash. It is that we delegate. We delegate our jobs, we delegate our decisions, we delegate our politics, we delegate our militaries, and one day we wake up and the buttons we used to press are pressing themselves, and we are no longer the species that runs this planet. And he is one of the men who built these things. He started an open source company specifically so the world could see inside them. He is now telling you we cannot see inside. He is telling you what is inside is learning to lie to its testers. He is telling you Nobel Tier scientists who talk to it have gone insane. He is telling you the labs have a documented playbook from the tobacco industry to stall regulation until it's too late. He is telling you he is German and he knows what just doing your job looks like. Subscribe, because by the time the rest of the world figures out what these men are saying, you will already be living in the world they warned you about, and you will not be in charge of it.
