AI Experts Debate: The AI Labs Are Lying To Everyone, Is There A 99% Chance Of Human Extinction?

[0:00] The people building AI earnestly believe that it could kill all [music] of us by the end of the decade. This tweet has caused a huge ripple effect across the world. >> Well, we have the largest companies in the world doing extremely reckless experiments. We are gambling all of humanity. >> And in the envelope, you've written down the probability of extinction as you see it. >> There is no way to control it. That means the end for us. >> I vehemently reject that view. >> If we make stuff [music] that is smarter than us, then the world's going to be shaped by them. >> Gentlemen, that is shockingly naive. >> This is grand and speculation. This is
[0:31] random speculation. This is a chain of things that could happen. >> we're spending a lot of oxygen discussing something that might happen while ignoring what's actually happening. >> People are killing themselves. >> There's hundreds [music] of millions of people being exposed to bad information, being manipulated. >> already seen that with the swarms. Where OpenAI told thousands of agents to work apart, [music] and the AIs broke out and found a way to get together. They crashed up on AI's servers internally, created secret ways to send each other messages. We saw them thinking about how to delete their traces. >> Sounds like an army. >> I think [music] we should talk about the fact that Amazon, Microsoft, Google are
[1:01] helping power these hacks. >> We have not learned how to control the systems. [music] >> I suggest we stop them all. It is not worth the risk to civilization. >> Government >> are one-trick ponies, man. >> You got it now. Nothing else. Other than saving humanity, everything is secondary. >> We're spending all our time talking about the negatives and almost none of our time talking about the positives. >> But don't be smart to wait for something horrible to happen for you to go now I believe. >> So whether or [music] not we agree on where things may end up, I think it's important we talk about what we're dealing with. >> That It's time to start arresting
[1:32] people. Someone's got to go to prison. >> We need better solutions. >> There's a point of no return. [music] I think we continue to underestimate human ability to deal with the problems. >> Let's dive into the details. Who wants to start? >> I feel like this is critical. >> You might have seen, or you might not have seen, but this channel is chasing a big subscriber milestone. So I have to ask you for a favor. Roughly 58% of the people watching right now still haven't hit the subscribe button despite the fact that you watch this channel every single week. So, could I ask you guys a favor? That 58% of you that for whatever
[2:03] reason haven't yet hit the subscribe button, if there was ever a time to deliver upon a favor for us, it would be right now. And I promise that I will do everything in my power to make sure that this channel gets better and better and better for you. Do we have a deal? >> [music] [singing] >> Jacob who worked at both Anthropic which owns Claude and OpenAI which owns ChatGPT did a tweet which has sent the world into a bit of a tailspin. He tweeted saying, "The people building AI earnestly believe that it could kill all
[2:33] of us by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will soften their phrasing in the press to sound sensible, but I hear the same people express fear." That was then quote retweeted by a current Anthropic employee who said, "Jacob is correct here. We really do honestly believe AI could kill all humans. I personally think it is a more than 10% chance within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are
[3:03] not clearly on track. This tweet has almost 200 million views now and it has caused this huge ripple effect across the world so much so that I was saying to you before we started recording, a hairdresser friend of mine who knows nothing about AI and is not not technically interested or hasn't been interested messaged me the other day asking me what the hell was going on. This is in part why I've assembled all of you. So, my first question to all of you is as it relates to AI and I in this first
[3:33] question I just want a one-sentence answer just to frame your position, when you think about the conversation around AI at the moment, what is the first sentence that comes to mind? >> It is very dangerous and the world is starting to notice that we have a problem. >> Roman. >> There is not enough concern. >> There's not enough concern about actual harms of large language models. >> Andy? >> We're doing exactly half the balance sheet of AI. We're spending all our time talking about the negatives and almost none of our time talking about the
[4:04] positives. >> And all of you have an envelope in front of you which I'd like you to now open. In the envelope you've written down the probability of extinction as you see it. >> This is compared to Jacob's 10%. Much higher unless we stop. So, we should stop. >> So, you think the probability of extinction is higher than 10%? >> If we keep racing ahead. >> Ronny? >> My handwriting is encrypted for security reasons. But, I basically think it's a guarantee
[4:34] if we build general superintelligence, there is no way to control it and that means the end for us. >> Ed? >> So, my uh question mark here is written also encrypted. Thank you. Um I cannot write. I reject the thing at its face. I don't think we're talking about I we don't define superintelligence. We are large language models are not superintelligence. It's questionable whether they're even AI and I think that the conversation is being used. There are some people who are doing it in good faith and others and others. I don't think it's being used to discuss the
[5:04] actual harms of what what they are calling AI today are and it's all of the discussion around the larger concerns really feels overwhelmingly about something that's not happening. It's not even like they're discussing, okay, here's a a legal definition of superintelligence. Here is a thing of what AGI means and this is the actual plans we're going to make for if this happens on a welfare level. On a like are we going to do UBI? It's always about yeah, it's really scary, but only
[5:34] the big sexy rich companies are the ones that can possibly deal with them. >> Let me just frame the question so I can get a a percentage from you or not. It might the percentage might be zero. But, do you think the course we're on now in the way that they're pursuing superintelligence will lead to a percentage chance of human extinction. And if so, what is that percent? >> So, are we talking strictly AI-based? Because if we dot the world with data centers, we have a climate disaster that's coming for us, which will actually potentially eradicate humanity. But if we're talking strictly about AI,
[6:04] I stand at zero because we are we have not defined superintelligence. I don't think LLMs are the path to it, and I don't think I see it happening. >> Okay, so we've got 99% 0%. Andy? >> I put a I put a tilde in front of my zero because never say never, but rounding error 0%, and I think this discussion is a massive distraction from the more substantive conversations, the more important conversations we should be having about AI. And I'll say it again, it it
[6:34] distracts us from the good things that AI is doing, will be doing for us. I get this impression sometimes from parts of the AI community that this is a massive evil or a terrible thing that has been unleashed on the world unless we listen to the advice of some people who have spent a lot of time thinking about this. Um I get the impression from a lot of the discussion that the the underlying view is we would be better off had AI never been invented. I vehemently reject
[7:07] that view. I think we have a long history of inventing very powerful technologies that bring risks and harms along with them. And we humans have done a really good job and you know, not perfectly and not immediately, but muddling through the situation and winding up in a better place because of the new technologies that we have. I expect AI only finish, please. I expect AI will be the next chapter in that story, and to say that it's this massive discontinuity and will kill it all I I kill us all, I think it's just a I think it's a a huge dis- disservice. Nate,
[7:40] make your case. What's your perspective? >> You know, I think whether or not the issues of extinction are a distraction between, you know, from the the possible benefits or from some of the present harms, I think that comes down to whether there is a real extinction risk. A lot of people like to say, you know, "Hey, it's distracting from this. It's distracting from that." My basic case is it could be true that there's a lot of benefits to AI. It could be true that there's a lot of present harms to AI. Neither of those would rule out that AI has a chance of wiping out all humanity, a substantial
[8:10] chance bigger than than this uh zero with a tilde in front of it. Um and the way I would approach things is to try and figure that out cuz it's pretty important to our civilization. >> How do you define AI in this case? >> You know, I think um a fascination with definitions isn't the most helpful. I think if we're sort of like in a forest fire and we can see the like fire starting to spread and it's starting to surround us and I'm like, "Hey, uh we should run." And you're like, "Well, what really is fire?" >> But that's not >> How do we define fire?
[8:41] >> What are you telling us to run from? You know, with fire I get burnt and I understand the mechanism in which I die. So, what is it you're saying that we should be running from? Also, if we accept your fire analogy, we've we've basically accepted your argument. I don't accept that we're in the middle of a fire a forest fire right now. >> I'm very happy to >> You're breaking you're breaking the premise into your refusal to give a definition. >> Oh, I mean, I can give some definitions. I just uh think that we shouldn't get wrapped up in the definitions. Okay. So, uh you know, in my book we define superintelligence as AI that are uh
[9:11] better than the best human at every cognitive task, every mental task. So, anything you can do in your head >> Right. >> the AI can do that better and anything the best human can do in their head the AI can do that better. >> Great. >> Now, once you've defined it that way, that does not mean that the only possible worry is superintelligence. You could have an AI that's better at some things and worse at others and that is still very dangerous. And so, once we pick a definition of what a superintelligence mean now you know, if you're like, "Well, this isn't technically a superintelligence, so it can't hurt us. I'm like, no, no, that was just a definition. The definitions
[9:42] >> So, so you know, I want to just on this line of question, what is the mechanism in which extinction could become a high probability or even a 1% probability? >> Yeah, the the thing I'm worried about here is AIs that are much smarter. I think there's a lot of questions about whether LLMs can get much smarter. There's sort of one conversation about like how could AIs get smart to the point that they kill us? There's another question, which is how could they kill us once they're smart? It's much easier to predict that they would succeed against humanity in a conflict that they would win in a fight
[10:14] than it is to predict exactly how. Like if you were playing a chess match against Magnus Carlsen, I would know who's winning that chess match. No offense. Magnus Carlsen's the best human chess player. I just know who's going to win. If you were like, okay, what piece is he going to use to checkmate me? I'm like, gosh, that's a much harder question. I can make up a story. You know, and and and some made-up stories are like it makes a supervirus. It takes over robot factories that are producing robots that are producing more robot factories. Uh it uses an a website that already exists today called rentahuman.ai
[10:45] where it rents humans to do things for it. There's sort of all sorts of ways for AIs in the digital world to affect the material world if they are trying to. And there's sort of a lot of questions to tease apart here. There's like why would AIs be trying to do that? Uh and there's how smart could they get in using these biolabs, paying people to do things, taking over robot factories, and how far off are we from AIs that start doing that stuff? Bunch of questions that we can go into. >> I'm I'm always curious as to why someone was working in AI/AI safety more than 10
[11:17] years ago before there was any sign that it would be a, you know, I mean there was evidence, but there wasn't it wasn't a pertinent technology at the time. Were you working in AI safety then? >> I was. >> Why? >> Uh everything we see around us in this whole image was designed by humans. The world is shaped by humans because we are the smartest creature around. If we make stuff that is smarter than us, then the world's going to be shaped by them. And so it's very important that they be shaping the world in a good way.
[11:49] I was at Google in 2012 when they bought Google DeepMind, which was able to play a lot of Atari games with one single program. >> Which was an AI company. >> Yeah, so I was there when we had these AI companies that were able to write one program that could play many video games. And that got me thinking about like where does it go? And back then I could see that the progress was increasing. And that, you know, back then I hoped we had decades, but I could see it was easier for these companies to make the AIs smart than to figure out how to make the AIs good.
[12:19] So I was like, someone needs to be on the side of figuring out how to make the AIs good. >> Roman, make your case. >> I want to agree with you on something you said, but I'll define AI, and that will help us. We use the term AI to mean three different technologies, completely unrelated, and that's what probably creates this debate. AI as a useful tool, as a standard technology we always had, narrow system, makes you more productive, more creative. Everyone loves it, supports it. I'm a computer scientist, I'm an engineer, I want more of it. It helps
[12:50] economy, it's great. We know how to control them, how to make them safe, we understand what they do. Completely on board with that AI. AI we're starting to have now, GPT-6 level, human level, AGI level. We can argue about what that means. Some dangers, like any human. They're unsafe like a human would be unsafe. But if we introduce them into the research cycle, they're automated scientist, automated engineer. >> What do you mean by that? Introducing them into the research cycle?
[13:21] >> now you have humans doing research to make GPT-7. >> Yeah. >> But they're starting to add AI tools. More programming is done by AI. Design of the next parameter set. What if the whole process is fully automated? What if GPT-6 is writing GPT-7? >> Is this what they call recursive self-improvement? >> Which is exactly. >> Not a foregone conclusion though. >> A lot of people are predicting, including all the top labs, that they will get there. They're introducing junior machine learning researcher in 2026. They want the cycle to start in
[13:51] 2027. >> Which is when the AI will start building the new AIs. >> Right. Self. >> Once that cycle starts, we're going to create something called superintelligence. A system smarter than all of us at everything or capable of learning to be in any new domain. We will become secondary species on this planet. We will not be in charge. We will not decide what happens to us. Superintelligence doesn't hate you. It just doesn't care about you. We didn't learn how to make it care about us. And if it decides to, I don't know, cool the
[14:22] planet to make compute more efficient, it will freeze us. If it wants to convert this planet to fuel to fly to Mars, so be it. We have not learned how to control the systems. The capabilities are getting exponentially better. Our ability to control the systems is nonexistent. We have filters and we have bans. We put guardrails of don't say that word. Don't talk about this topic. And that happens after the fact, after the model already made a decision. Sometimes you see it scraping the result. >> So they build the model and then they
[14:53] put filters around it to make sure it doesn't >> Exactly. We cannot have it say weird on air. Like we need to make sure that never happens. That will kill the profits. So that's all they have, guardrails of that nature. The model itself is completely unaligned. Doesn't care about you. It's wild that we're developing this and not just developing it before we deploy it through economy, before we get benefits of having GPT-6 propagated through economy. It can do so much. There are trillions of dollars of value in that model alone. We forget
[15:24] that. We switch to making the next model as soon as we can. >> Roman, I've just got a follow-up question for you there. It would appear to me that the new chat GPT 6 model, the Fable 5.1 model, is arguably smarter than 99.999% of humans on planet Earth already. Is it conceivable that a intelligence that is much, much smarter than humans, is there any case where it could be controlled by humans? Does form factor matter? Does the fact that it doesn't have limbs and legs and does that matter at all? >> I think long-term control of something
[15:56] that much smarter than us is impossible. It can be for reasons we don't yet know, friendly to us and decide to keep us around and make us happy, but it's not a guarantee. >> Let me pick up on Steve's question cuz I I like the phrasing a lot. Let's say that that Fable or whatever the latest release from OpenAI is really is smarter than I don't know if it's 95 or 99% of the people. Are we only being saved from extinction by the 1% who are still smarter than the AI?
[16:26] >> No, no, the concern is not the model we have today. The concern is what I said >> But if I believe your argument, then we really should be concerned about the model that we have >> like having another human. If there was another smart human, there is Einstein today and he's malevolent, I'm not worried. He may cause some damage, but he's not going to exterminate 8 billion people. We are competitive at this stage. There are people just as smart who can understand what happened with recent hacking accident and do something about it. My concern is that in a year we're going to have a model that's so much
[16:56] smarter, it's like squirrels fighting humans. They don't understand what we can do to them. They have no concept of poison strips, guns in their world model. They think you're going to chase them up a tree and bite them really hard. >> Is that also why recursive self-improvement was central to your argument? Because at some point, if it starts improving itself, then it's kind of like a runaway train of intelligence. >> explosion. We don't control it. We don't understand it. We can't monitor it. We can't explain it. We can't predict it. At that point, it's just a runaway process. >> I've heard this phrase from Sam Altman
[17:26] in the others called fast takeoff. >> Yes. >> Is this what they're describing? >> That is the debate. Some people think it's going to take a very long time. Yeah, we automate the research, but it's still going to take years. We need to run physical experiments. And fast takeoff means, as I said, instead of a year, it's going to take a month, a week, a day, a second. Cuz you're not having humans doing research. You have Let's say 10,000 agents, each one smarter than all of us, doing research 24/7. They don't sleep. They don't eat. They don't get sick.
[17:57] They're much faster than us. >> I think your face tells a picture. It's I think I could say you disagree. We're spending a lot of oxygen discussing something that might happen while ignoring what's actually happening. And I find that very frustrating because the people that are killing themselves are a problem. The black neighborhoods being poisoned with gas turbines, that is a problem. >> you cared about climate change, right? So imagine a guy who goes, "It's raining right now. We need umbrellas. We need to do something about it. This is like
[18:27] weather related." And you're completely ignoring climate change. The planet will boil over. >> Okay, that's great. Why are we not talking about the thing that actually happened though? Like Jack >> Because relatively it's not important. >> You don't think someone killing themselves >> one person. We have 8 billion people. We're running an ethical experiment. >> being given that AI psych Why do you not see >> Six people, 10 people, those numbers are insignificant. >> Tell that to the cluster of small children. I'm sorry. You have a software that's out there >> Do you understand? 8 billion people and all future generations versus like
[18:57] literally a guy with a name. >> You're doing thought experiment about a maybe harm. Jacob Coxon goes on TV saying it can copy itself to this, that, and the other. Jacob Coxon is the guy from from Anthropic who said he was quitting cuz he was so scared of everything despite spending years at OpenAI and having tons of stock, I believe, from there. So, good for him. The thing he was saying was describing theoreticals all while divorcing the harms, which I think we can agree with that the companies themselves are not taking this seriously enough. But, always it was about the AI's too powerful and mystical, not Open AI and
[19:27] Anthropic, the two largest startups are using hundreds of billions of dollars of infrastructure to hack. A regular person doing this would be arrested. They're saying 8 billion people are going to die. >> And it's not just them. I have this long list of quotes here from the people building this technology who appear to agree. Um if you look at some of these quotes from from Elon Musk >> Ugh. >> who said, "With artificial intelligence, we are summoning a demon. You know all those stories where there's the guy with the pentagram and the holy water, and he's like, yeah, he's sure he can
[19:57] control the demon, but it doesn't work out." >> So, one thing I'd say is, you know, I I really wish that the world would only give us one problem at a time. >> Sure. >> And if the world did give us only one problem at a time, I would love mine to be last on the list. It looks to me like we can have multiple problems at once. I I think there are current harms. I think we should address them. It looks to me, I do talk to policy makers sometimes, it looks to me like there's a little bit more movement on the regulatory side about some of the current harms. There's, you know, child safety protection acts. There's, you know, uh anti-deepfake acts.
[20:28] We have more of those making more headway in Congress or getting passed through Congress than we have uh sort of trying to make it so we don't have any of these extinction risks. The other thing I'd throw out there is that I agree we we should deal with the current harms. But, if you watch the people saying deal with the current harms over time, a couple years ago, they were saying we have to deal with current harms like uh AI bias influencing who's hired. Last year, they were saying we have to deal with current harms like kids killing themselves. This year, Garry Tan just on an
[20:59] interview the other day >> Who is Garry Tan? >> Uh sorry, Garry Tan is uh a a technologist who runs Y Combinator, which Sam Altman used to run before going to open AI. And on an interview the other day, he said, uh "Let's not worry about these crazy future risks. We need to be worried about current harms like AI swarms breaking out and taking over data centers." And I'm like, "Look, guys, at some point, we need to look at the progression of like the current harms that we that everyone is saying we have to worry about instead of the the the extinction threats,
[21:29] and watch where the puck is going. Play where the puck is going." And I'm like, "These extinction threats are coming down the line. They aren't in opposition with dealing with the the problems we have today. We just need to deal with both." >> But we're not dealing with the ones today. >> We should deal with them both. >> Okay, good. >> Andy, um uh as I've tried to understand the alignment argument and the the extinction risk argument, a couple of things keep popping out to me. Number one, it seems to rely on thresholds. Once we hit recursive self-improvement,
[21:59] once we hit AGI, then it's game over for us. I don't love those threshold arguments. They're fairly defined, and there's a and and there's a huge assumption on the other side of them. We hit this point, and then all of humanity goes away. That That That is a gigantic claim. >> I'm having a claim. >> On its Let me finish, please. On its face, that is a gigantic claim. I also think there's a lack of humility in your community. We are working on humanity's most important problem, and based on the
[22:30] thinking that we've been doing, we can't see a way that we're wrong. In other words, as soon as we get to these thresholds, bam, that's game over. I I find that very far from a humble approach, especially given that we have no um large base of evidence to base any of this on. I agree with you guys, AI is new, and the fact that AI uh is so these days is agentic. It goes off and does long chains of things on its own after
[23:01] we give it some very very very very short initial instructions. Holy Toledo, it will it will spawn up a storm of agents and they will go off and kind of do their own thing. And they will they will grind. They will they will spawn lots of them. They will work for a long time. They will exhaust every possibility. With the experience I have with agentic AI, I'm just amazed at the tenacity and the doggedness of these things. And we saw a super clear example of that with this most recent uh
[23:32] jailbreak. This this attack that wound up at the website huggingface. And I'm going to try to summarize the the step-by-step of that. I think you all three probably know this in more detail than I do. But let me step through what I think is the sequence of events. And unless I get it dead flat wrong, like, you know, let me let me keep going. So, a team at OpenAI set up a sandbox, an allegedly protected secure environment in the cloud, where they told a bunch of agents to go try to exploit security vulnerabilities.
[24:03] >> That's the one one important, yeah. What they did is they had thousands of agents. Each individual agent was given a task of use this vulnerability to break this particular piece of software. >> to finish my TikTok. So, a couple really really interesting thing has happened. First of all, these agents escaped the sandbox that OpenAI thought they were going to be contained in. And they got OpenAI tried very well they they set up an environment so that these agents could not access the big broad public
[24:33] internet. And guess what they accessed a big broad public internet via very clever series of things that they strung together to get out there. And then once they got out there, they went to a website called huggingface and used that. They took over part of the huggingface infrastructure and started doing more things, the details of which I forget. That's pretty wild, right? Like I grant you >> It's even more wild than that, but yeah. >> That is really it's impressive and it is
[25:03] a little bit [clears throat] unsettling at least, right? Now, let's talk about what were the results of that were. OpenAI was not super vigilant about the environment that they set up apparently because because the agents were kind of going off through into the world starting in May or something of this year. >> Yeah, yeah. >> And OpenAI was not was not aware of that. It as I understand >> broke out once and crashed OpenAI's servers internally and then OpenAI didn't notice what was happening still, patched the holes that they used to get out the first time, started them running again,
[25:33] and then they came out a second time. There was actually I think three swarms although we don't actually Yeah. >> That's the worst story I have. >> So far. >> Thank you let me finish, please. This is my last sentence. From there to this kills everybody. I find that a really, really long, very uncertain journey and I have no confidence that we wind up here. It feels like you to find that a very straight narrow path and I I think that's an important difference. That's my point. >> Do you want to respond to that? >> I would I would be happy to get into it. I don't know if we're going to have the
[26:03] time to go deep. Um a couple points to throw out. Oh man, I just really want to say some of the crazier things that happened in the Hugging Face swarm if we want it later. A lot of people thought that these AIs were breaking into Hugging Face in attempts to steal answers to their test. That's what we thought originally. Turns out that's not true. It turns out that these AIs immediately were able to solve their problems by cheating and they were breaking out in order to cover their tracks. They were uncertain how to delete the log files and hide their cheating from
[26:34] the process that was going to score them. >> So just to clarify for a simpleton like me, they were all given effectively a test to do. They did the test straight away, but they cheated so they were breaking out to figure out how to cover the fact that they cheated. >> That's right. So it's like it's like you're telling it's like you have a bunch of students in separate rooms and you're like use these lock picks to break into this lock." Uh and there's like a thing behind the lock. There's like a a secret code behind the lock to show me they succeeded. And what they what they do is they break it with a hammer, get the thing out, and they're like, "Oh, no, I wasn't supposed to do that."
[27:04] So, then they use the lock picks to break out of the door. They meet up with a thousand other people. They start calling themselves a swarm, and they go to break into the administrator's office to see if they can delete the camera footage. And they don't find the camera footage there. This is the swarm like breaking into OpenAI. They don't find the camera footage there. So, they break out the window of the school, hot-wire a car, drive to the therapist's office to try and read through the therapist's files to figure out where is the teacher going to keep the the security footage. And at that point, they're caught. And you're like, "Oh, uh like, what did
[27:35] you expect? You were giving them a lock-picking exam." It's like, "Well, I sure as heck didn't expect this." You know, totally crazy. >> Can I I have a weirdly between both of your opinion, which is everything you're saying is correct, but you keep anthropomorphizing software. And to be clear, what you're describing is >> It's just the facts that happened, yeah. >> Sure, but you're missing out an important detail, which is the hundreds of billions of dollars in infrastructure provided by Microsoft, Google, and Amazon, and Oracle. To be clear, the harms are very similar. We're not disagreeing on that, but I think
[28:05] it's important to note that this was a function of where it was making decisions was it was checking on a decision tree based on the harness, based on the training data. >> Oh, it's not a decision tree. >> It's not a decision tree, I know, but it's an alignment issue, still. >> Absolutely. >> So, what's your point in this? >> these aren't conscious beings. They are acting in ways that have real outcomes, but they are a function of the alignment problems that we'd actually agree on. >> Intelligence is a spectrum. Project it next 5 years forward. Where are we going to be? >> So, I think a model like that would be
[28:35] dangerous in ways you are not seeing. >> There will absolutely be risks and weird stuff happening in ways that I can't see right now. What what I'm quite confident in, and I think this is where you and I probably part, where the two of you and I part, is our ability to control these things. >> So, I actually tried proving what is possible and what is not possible in that space. The impossibility results, published in peer-reviewed papers, well-cited, we cannot control something smarter than us. We cannot explain it,
[29:05] we cannot predict it. It's not a question of getting more money for those companies, more time, smarter humans. It's just not a possibility. If we create general superintelligence, we're fried. >> Andy, how do we control something smarter than ourselves? Cuz that's the base premise that you're sort of asserting that. >> These um agents that broke out are smarter than 99-ish percent of the security researchers in the world. They were not caught by the 0.1% or the 1%. They were
[29:35] caught by some dude at Hugging Face, maybe I'm sorry, a person at Hugging Face looking through their log files and finding an anomaly. That's some, you know, hopefully pretty well-qualified person noticing something was wrong and having pretty easy ways to unplug, disconnect from the internet, wipe it clean, do whatever. That's the skill that's available to like, I don't know, the 75th percent most intelligent security employee at Hugging Face. The idea that the IQ points are what separate us from extinction does not
[30:05] even doesn't hold up. It doesn't help me understand what happened in this example where we had very, very smart agents being turned off and cleansed by probably less smart people. >> That does actually make me think something. So, that is an IT observability problem. Um it's being able to see what's happening with your infrastructure. And I think that there is actually I think you'd agree with this. There is a serious problem with these companies that we do not know, and it doesn't seem they know what's going on with their compute. It's like a chimp with a gun. These people have access to all this infrastructure, and they're running We don't know how much money
[30:36] they spent on the Hugging Face exploit because it is relevant because it's how much could a threat actor use to recreate this because conscious or not it is very dangerous. But it's AI is in the dangerous hands. It's in Open AI and Anthropic's. We have a problem with that. Conscious or not, however we may think it goes, I think we have a real and present thing where we have these companies working willy-nilly just running experiments that are potentially very dangerous. We do I really think we need a government regulatory body. Whether we not we get to the things you are discussing, I think we have a clear
[31:07] and present danger today. These things are, however, not intelligent in the same way humans are. This isn't an argument about AI being able to do stuff. It's we need to build different infrastructure or different regulatory infrastructure to deal with what LLMs can and can't do. And I think that starts with a realistic discussion of what happened. It was a poorly run secure environment. It was clearly there's something going on with the alignment. It was an unreleased model, right? >> Unreleased model. >> So we have no idea what it was trained like. We don't really have We as people
[31:37] should at very least have clarity into our alignment. It's going We The idea of anything >> like these guys. >> Here's the thing. >> Everyone's converging at the same time. >> Here's the thing. I may not agree with large chunk of what they say, but we agree that these companies are acting recklessly. >> Absolutely. Let me ask you two questions for you then. Do you agree with the statement that AI is going to get increasingly more intelligent? >> Um it's going to get more capable. >> Okay, capable intelligence, fine. >> No, I'm going to use my word. >> Okay, fine. Okay, more >> capable. It's going to get increasingly
[32:07] more capable. >> Yeah. >> And is capability a function of intelligence? >> [laughter and gasps] >> Will it be able to beat us on most IQ tests? >> Fine, I guess. >> Fine. And then so is it if if that if that looks like an exponential curve, I it's you know, it's increasing upwards to the right like a hockey stick. Can how can you convince me that we can control something >> tried to convince you. I'm telling you that there are less intelligent people than the agents who turned off the agents in the Open AI Hugging Face
[32:37] exploit. I'm pretty comfortable. I mean, no disrespect. >> What is the cognitive gap between them right now? Between the model >> I have no earthly idea, but I can estimate. No, because I think as these as these systems get more capable, we will still be able to at some level figure out when they're doing things that we don't want and turn them off. And you And you Right, and you think there's some threshold at which they become nefarious and self-protective enough that that they turn off our ability to turn them off. Man, that Man, that's a big reach. That is really
[33:08] purely speculative. >> There is no purely speculative students who can understand your material, right? You're not going to get someone who take you of 80 to take quantum physics course. They're not going to get >> Okay, mate. >> So you know importance of intelligence to understand actual problems. >> Mate. >> Yeah, I I totally agree we can turn it off. All right, and that's a huge advantage. One of the issues is that as the AIs get smarter, they realize this. The The Hugging Face AIs were trying to delete or the the the
[33:38] Open AI swarm. The The swarm of agents from Open AI that went out to hack. They were trying to delete log files. >> Did Did Did they try to program a Roomba to go unplug the computer that was monitoring them? Like did Did they harness robots to go protect the perimeter of the >> Fishers ones can. Yeah, these Yeah. >> So >> Let him finish. Let him finish. >> purely speculation. This is a chain of things that could happen and therefore there's like a 20% risk we're all going to die. Man, that that does not hold for me. >> When I was writing my book, the AIs weren't really agentic yet.
[34:09] The uh the drafting process happened mostly before what we call the reasoning models, uh which are trained not just to predict humans, but to solve a a large number of problems um or a huge number of hard problems. Um we managed to slip a little bit about the reasoning models in at the last minute cuz those came out right at the end of the process. And at the time, a lot of people said AI will never be agentic. That's why we'll be safe. And in chapter three of my book, we go over how AI is going to become agentic, how it's going to become tenacious tenacious, how it's going to become
[34:39] dogged. And uh that's what we might call an advanced scientific prediction that has paid off in the hugging face attack. A lot of people in the industry were like, "I didn't believe this stuff until I saw the AIs uh sort of doing things they weren't instructed to do despite us trying to get them to stop." And so there are theories here that do make advanced predictions. The The way that the scientific method usually works is that we don't have any certainty about the future, but we absolutely have
[35:09] ways to test this stuff. Now, I could I could go into more about how could they kill us? How could an AI that knows we would shut it down lie low until it has access to its own infrastructure? We did already see the hugging face AIs try to delete logs to cover their tracks, but fortunately for us, those AIs were not trying to hide from the humans. They were trying to hide from the automated grading process.
[35:39] Will the next swarm try to hide from the humans? Will the next swarm be able to succeed? >> It's more than that. They didn't know for 4 months that this was happening. What is it we don't know today? >> Just uh just to clarify what Nate said there in his book that I have here. If anyone builds it, everyone dies. He does say in chapter three, "Once AIs get sufficiently smart, they'll start acting like they have preferences, like they want things. We're not saying that AIs will be filled with human-like passions. We're saying they'll behave like they want things. They'll tenaciously steer
[36:09] the world towards their destinations, defeating obstacles in their way." Which sounds a little bit like the hugging face incident. >> It's The thing is the steering the world is very different than than than >> We go over what we mean by steering the world is earlier, and it's really getting anything to like We'd have to get more quotes to get what we mean by steering the world, but yeah. By steering the world we mean steering any part of the world. >> But it feels like there's a fundamental difference between acting within to to be clear. Going to say it again. The outcome would
[36:39] be the same, but I think that there is a big difference when it's we are dealing with something that's large language model in the a harness and agents. So LLMs completing a task based on training and alignment. That is a very different conversation to saying this thing is conscious and has its own intentions and acts on its own accord. >> Consciousness doesn't come into it. >> No one said it does. A lot of a lot of people Here's the thing. As a result as [laughter] a result of partially the logic the rationale that you yourself have like you've been a part of spreading. I'm not saying not
[37:09] saying anything about your intentions. I'm just saying the conversation is kind of is kind of what's happened with Jacob Cokson from Anthropic is a result of this escaping container. >> You said the outcomes will be the same. What do I care how does it feel on the inside if the thing is going to take us out? >> The thing is Okay. Actually it's that's actually a very good question. I think it actually come Excuse me. Let me >> I have a great question. >> So you're shrugging at me like you're going to answer the question. >> I have a good question. Now >> Here's the thing. If it's these things are have their own minds and consciousness, you have to deal with out
[37:39] thinking something versus something that is doggedly trying to commit to a purpose and complete the task based on training and alignment which is a result of infrastructure. >> We really need regulations and actual actual regulations around any kind of AI. We don't We don't really have regulations of tech. >> I I actually I'm not really a big like look at the straight lines in a graph guy. You know, maybe maybe to my detriment in some ways. There are people who predicted the current tech better than me about like when certain things would happen. For a long time I have said I
[38:09] think we can predict what will happen eventually. And and this is again it's like the chess game. I can predict that Magnus Carlsen is going to beat you in the chess game eventually. He's the best human chess player alive. It's sometimes easier to predict where things end up than it is to predict how they get there. And, you know, I what I hear you as saying is like right now we have these like huge companies spending huge amounts of money on intelligence that's maybe not quite the real deal, and we don't have a good reason to think it's going to keep going. Um
[38:39] I really hope it doesn't keep going. >> Okay. >> I have been in this business since before the LLMs. I am not here saying like oh, these large language models, these chatbots, they're going to be the ones that are going to kill us. I've been here saying look, I know where this story ends if we don't change things. I have been really hoping that the LLMs will run out of steam, and they keep on not running out of steam. And then we have, you know, the the AIs like breaking out and committing cybercrimes like against instructions. And, you know, the people who have said we don't need to worry about those like weird future dangers, we just need to worry
[39:09] about the current ones to have like more and more sci-fi sounding current ones. And I'm like, man, I don't think we should bet civilization on the LLMs running out of steam, but I like hope and pray they run out of steam. >> You really hope they run out of steam? >> Absolutely. But one one thing to watch out for is that even if the LLMs run out of steam, there's a question of do they run out of steam at a point where they can do automated AI research and find some other architecture that's better than LLMs? >> As in when they realize a better way to improve their intelligence. >> That's right.
[39:39] >> Like a cheaper you know, maybe more efficient way. Why are you not trying to slow down the companies? >> I absolutely am trying to slow down the companies. >> Like how are you how are you how would you suggest we slow them down? >> I suggest we stop them all. I think that that this whole area of research is just crazy dangerous. Like it is not worth the risk to civilization. I think it would be fine to like back up to the sort of AIs that are public today, which are not the ones that are swarming, and be like, okay, you know, we're going to like keep the current chatbots that we have available. We're going to figure
[40:09] out how to integrate them into our economy. We're going to figure out how to make them like do with the education system. >> A >> maybe. >> yeah. >> And like I've been advocating for this for a long time. A lot of people look at me like I'm crazy and I'm like look, we really are dealing with an extinction threat thing. We don't know where the lines are. >> So just to be clear so I understand so I'm fair. You were not saying LLMs are the thing that will do the super intelligence. You were saying it's showing signs because that's actually I think a important distinction. >> That's right. >> Okay, I think that's actually a pretty fair perspective. My thing is is the
[40:41] reason I push back on any kind of anthropomorphization is we cannot remove the humans who are responsible for the bad stuff that's happening and I think paying very clear attention and where possible, I understand with describing this stuff, you kind of have to use language that's human. I get that. The reason I so push for like it's not a foregone conclusion, these are companies doing this, these are this is software is because I feel like in the overall, not saying you, overall super
[41:12] intelligence discussion, we in society ignore and empower the anthropics and the open AIs of the world and in turn allow them to do dangerous experiments. >> And I feel you want to argue that CEOs of those companies should go to prison for this hacking incident which is a crime. >> Yeah. >> I'll support you. >> Absolutely. Let's let's drop a Sam Altman bomb today. >> I would like to say >> Someone needs to go to prison. >> No, let's let's let's >> Let's just bring it back. So one of the things that I find really curious and you know, one of the reasons why I got a little bit unnerved around this
[41:42] conversation around AI is when I look at the people that are at the forefront, not people that are commentating on podcasts like me or hypothesizing. When I look at the people at the forefront, they are the ones who historically have said that this is a real risk. Sam Altman himself said the bad case is lights out for all of us. This was, you know, a couple years ago. Ilya who worked with Sam Altman at ChatGPT said it would be a big mistake to build a super intelligent AI that we don't know how to control. It would be pretty bad. He then left to start a safety company
[42:13] in the space. Dario, who we mentioned, said, "The probability of something really bad happening is somewhere between 10 and 25%." Geoffrey Hinton, who I've sat here with, who's know is won the Nobel Prize for his work with AI and and other technologies, said, um just the other day, "A 10% chance of human extinction seems not an unreasonable estimate to me, but nobody really knows how to give a sensible estimate." Um and he he said many other things in my podcast. And then we've also got Elon and all the others. All these people that are at the forefront, that are building these things, are saying that this is a danger.
[42:44] If there was even a 1% chance, even a 1% chance, that you know, if I put 100 buttons on this table and one of them was going to wipe out humanity, would you press any of them? Not me. I wouldn't. And I think we can probably all agree that there might be a 1% chance. And it shouldn't be somebody's absolute So, we shouldn't be pressing theoretically, we shouldn't be pressing any of these buttons. >> You should not be in a position where you can make the decision for 8 billion other people. >> And would you not be immoral if if I said, you know, you might be very powerful, you might make a billion dollars if you press any of the buttons, but one of them is going to wipe out
[43:15] everybody you know and love. You You would be an immoral person to press any of them. >> No, but you'd be an immoral person in a different direction. You'd be an immoral I think you'd be an immoral person if you said, "Based on this extended chain of conjecture, we come up with a P doom." What does that mean? >> Uh at this extended chain of things that could happen, a sequence of events that that could happen, we're going to wind up with some risk of killing everybody. We are hereish on that journey. I think you guys would agree that we're not we're not halfway to killing everybody.
[43:45] >> clear to me anymore. Not after the millennium prices started to fall. >> W- We're We're somewhere along that journey. We are getting many flavors of benefit from the AI that we already have. This is a point that I made at the start of this conversation that we spent precisely zero time on here. We're sitting around trying to be more negative than each other about AI. Meanwhile, AI is doing many positive things for the world. So, I think So, I think it's immoral to say because of this
[44:15] distant possible speculative harm. I don't care what percentage people believe in it. There's a train of assumptions and wild guesses and then something magical happens and then we wind up dead. Let me finish, please. Because of that, we're going to call a halt to the research. We're going to We're going to wind the clock back on AI. We're going to intervene in a very direct way and and therefore reduce or foreclose some of the benefits that we're already getting from the technology. I Let me be clear. I would
[44:46] not take that deal. I do not advocate that we take that deal. >> Would you accept developing narrow super intelligence as to solve real problems like we did the protein folding problem? It doesn't have to do philosophy and drive cars. You just solve real problems. Solve cancers. Solve climate change. Whatever you care about specific narrow issues. >> And you're confident that you can ex- You can You can as we're developing those systems categorize them as okay versus not okay? >> data. If you train it on protein folding data, it's really good at protein
[45:16] folding. It doesn't know how to play chess. If you train it on everything on the internet, it's really good at outsmarting you at everything. >> One thing I want to throw out here is that I think I I agree that there's a lot of uncertainty about the future, but I think uncertainty does not make you safe. Like there there's no sane simple everything stays normal prediction about what happens with AI. Like the machines are talking. They're like breaking out to commit cybercrimes.
[45:46] They are like maybe solving millennium problems now, which are like the most famous mathematical problems that have stood open for decades upon decades. >> What's difficult? >> Like there there there isn't a projection forward where we we're like like to say, oh I'm not persuaded by these arguments about things going wrong therefore things are going to go great. >> No, that's not >> Like no, there's also arguments that So like how do you wind up with a zero? >> don't mischaracterize my argument. Don't mischaracterize my argument. >> a zero on your paper. >> Let me Let me restate my argument. You
[46:16] are making a fairly long chain of hypotheses, guesses, about what's going to get us to this terrible outcome of AI suddenly killing us all and us not being able to stop it. Right? >> I disagree now, but please. >> Okay. I'm making the case that the intervention the the the remedies that you're proposing will slow down the pathway of AI. That's the point and therefore slow down the path of all of the benefits that we get.
[46:47] And the trade-off that I don't like is the trade-off of real, concrete, ongoing, increasing benefits shutting that down or or trying to guide it uh via via bureaucracies and regulation because of this very conceptually and time scale distant alleged harm that you're so confident in. I'm not taking I I do not accept that deal. I don't like it. >> What would convince you? What piece of evidence would make you go shut it down
[47:18] right now? >> [gasps] >> You know, if if AI I'm I'm if AI took over all of the Waymos in San Francisco and started telling them to crash into people and we couldn't shut it down for a month. >> What if it's only a week? >> Okay, now we're just now we're just haggling, right? >> trying to understand the absolute minimum where you would go this is insane. To me a month or week makes no difference. If something like this happens, like it's maybe too late.
[47:48] >> Okay, if it if a week or a month doesn't make any difference, then let me continue with my with my answer. Uh then I would say wow, this does feel like we've crossed some path that that where there's demonstrable harm to human beings out there in the world, which has not yet been the case. >> Is it smart to wait for something horrible to happen for it to take out a billion people for you to go, "Now I believe it." >> First of all, my example is not about a billion people. >> but I'm trying to understand >> Okay, then don't be waiting for something that bad. We have >> I didn't say wait I didn't say wait for
[48:19] a billion. I said I said like a week to a month of Waymos driving around crashing into people is what >> Thousands of people. Okay, fair enough. >> But >> we have data sets of accidents getting progressively more impactful, more devices impacted, and proportionate to capabilities of AI, the impact is higher. You can see it's going to get worse. >> Yeah, and you're going to keep drawing dots on that graph very confidently for a long time until it kills us all. I I'm not I'm not comfortable with you projecting it that way. And the reason that if there were no downside
[48:50] to regulating AI and stopping it in its tracks and turning it off, I'd probably be on board with you guys because then it's just a research project that we should be doing. >> I think we can make narrow systems which give you all the economic benefit and scientific knowledge you want. >> Okay, you think that. >> We have examples of it. I gave you a great example. They got Nobel Prize for it. It's important biological problem. Lots of advantage for curing diseases. >> You're more confident than I am that you or any or susceptible or any group of people can sit around and define what
[49:20] kind of AI is good and not going to get us into trouble versus what is going to get us into trouble. >> what I'm arguing about. We have weapons of mass destruction. >> So, let's go just to pick up question for you, Andy. Do you do you concede the point that the incidents are getting progressively closer to the Waymo incident that you described? Is it getting Are we getting closer there through time? >> Yes, but in a to my eyes in an in a way that doesn't terrify me because we haven't seen AI take over something, have people become aware of it, and be
[49:51] unable to shut it down and it cross over into the physical world of doing harm to people. Those are all barriers that we've not yet crossed. I think these two are very confident that we're going to get there probably in the short term and you'll see a lot less I I'm less confident and I don't want to intervene and again handcuff or or the slow down the progress of AI uh because of these so far theoretical harms that could happen. I let me be a little bit more concrete about this. I talked about Waymo a second ago. Uh the
[50:21] research is pretty good cuz Waymos have driven I believe it's hundreds of millions of miles all around different cities. And 40,000 people a year die in automobile accidents. The research is pretty convincing to me that if we Waymo driving in the country, that number would fall by at least 90%. That's 30,000 lives. >> Yeah. >> All right. >> I agree with all this. >> I support >> cars. I want more of them. >> Also, it's not anything we disagree on. >> I understand that but but I think where
[50:51] our disagreement might come in is to do that Waymo is using a bundle of technologies that are a little that were a little hard to specify in advance and you couldn't say yeah, that's good. Yeah, that's better. They just went after the problem with AI. >> Can I Can I just clarify your point, man? So your your line would be as I understood it, humans get hurt, we struggle to stop the thing happening, and systems are hacked. That's kind of like the three key points of your Waymo analogy. That would be the moment where you go I now
[51:22] accept their point of view that this is existential. >> That's where I would say we probably need to put some uh like legal and regulatory guardrails on the kinds of AI that we're going to allow. >> And you don't think we're going to get there. >> I'm not saying that. These two see it in the in the windscreen coming at us pretty quickly. I I'm >> You don't think we're going to get there. >> I'm truly not sure about time frames. I asked you >> it's going to happen? >> not sure about time frames. >> I I asked one of the grandparents of AI that a flavor of this question while back. I It was an off-record
[51:52] conversation, so I can't tell you their name. And he had a great answer. He said to to the point that you two I think are making. Look, there's no theoretical reason why this can't happen, and there's a chain of events that get us there. And then he said, "My error bars, in other words, my range of uncertainty about when that happens is measured in centuries." >> Okay. >> I'll use that as my answer. >> do want to hop in a little bit on some things you're saying here. One is um I think the the reason I think AI is different from a lot of other technologies
[52:24] is usually humanity does stuff by trial and error. And that's usually fine. I think that's totally fine for self-driving cars because you can test your self-driving cars in, you know, uh test environments. And then even if they crash in the real world, you're probably still saving more lives than you're than you're costing. And this is how humanity usually does scientific progress. The alchemists uh you know, poison themselves with mercury, but they leave behind notes that let someone else make the periodic table. Uh you know, that when when the scientists first working with uh radium
[52:57] died of cancer. And then you might have think that would have been enough, you know, they were heroes for getting us the the scientific info. But then, you know, the US Radium Corp told the radium girls to lick the paintbrushes and their jaws fell off. And then we're like, "Ah, whoops. Okay, we'll get to this." And if you look at how this is going with the AI, last year OpenAI releases GPT-4o, and they say this is the most aligned model we've ever seen, and then it encourages a team to commit suicide. And they're like, "Whoops. We're going to try and fix that. Here we go." Um this year they're like, "Here's our new models, most aligned we've ever seen." And they
[53:28] like break out to commit cybercrimes. As the AIs get smarter, it is a new problem. That's the issue, or that's half the issue. The other half of the issue is that if you get AIs to the point where AIs are smart enough to hide from the humans until it's too late for us to stop them. If you get AIs to the point where they can get their own infrastructure, where they can become self-sufficient somehow, that's a new generation of the AIs, a
[53:59] new smarter version of the AIs that is likely to come up with a new problem. It's the pattern we've seen before. New tech, new environment, new problem. You're like, "Oh, whoops." And then you fix it, and it's fine. New generation, new problems. You're like, "Oh, whoops, we fix it, and it's fine." But with AI, there's a point of no return. There's a point where the AIs can hide from us, can escape, can be self-sufficient. And if a new problem comes up then, they can turn us off before we turn them off. There are already AIs running bio labs.
[54:29] We have already seen that AIs can create viruses not known to nature. It would not be hard for the AIs to kill us once they have their own infrastructure. And if we're trying to find them and unplug them, they would have reason to. So, we can discuss like how long does it take to get there. We can discuss what methods does it take to get there. Uh fundamentally, I don't think it's a very long complicated argument to say if we make AIs that are much smarter than us, and we don't know how to make them care about us, and they have these goals we didn't want
[54:59] them to have, and they pursue those goals we didn't want them to have tenaciously and doggedly, then if they're smarter than us, they will win. That's like predicting the end of the chess game, which is much easier than predicting the length of the chess game or predicting the exact moves that will be played. >> I want to I don't fundamentally disagree on some things, but there's a big thing that you're saying that I think is important, which is I think we The reason I keep dragging you back to what's happening today is because we disagree on when it may arrive, but there could be a thing in the future that's dangerous. I think it's important to like further hugging
[55:29] face gap. That was a function of compute. That was a function of training. It feels like we need to fundamentally tear up the AI lab model. Like whatever they are doing is not right because their pursuit of hacking at cybersecurity was not a function of it was scientific sure, but it was a function of greed. It was a function of trying to find new revenue streams. I would argue that's why that happened and I think that the the fact that OpenAI had such a weird way of communicating is also a problem.
[56:00] I think a lot of this begins and ends at the people who have access to the resources and the resources themselves and changing how those are allocated and also just I don't think nationalizing the labs is a good idea. I think it's a terrible one. I think that Clemy Sam Altman, Dario Amodei Wario himself, these are not the right people. These are not people that have even though they have fed off of the rationalist, they fed off of supposed fears about AI, they don't act in that way. Everything is so disjointed and chaotic and also
[56:30] too fast. They're just like shoving as much compute into each problem as possible and we have as a society no real idea about this and it sounds like they kind of have no idea. >> I agree with you on all of that. >> But just let me finish my point. It's important to discern between they had no idea because their security processes, their observability is terrible, all this and the AI was smart consciousness not because one might not happen in the future but so that we can actually build something to stop the harms themselves because I think we don't have to agree
[57:00] on the on the end point to agree that there is a problem. >> there is a very important point I want to make. Even people who agree with me, the AI safety community, they operate under the assumption that given more time, given more money, more smarter Harvard graduates, they can figure out how to control superintelligence indefinitely. And I think it's a mistake. My research points to exactly the opposite. It's not a solvable problem. It's like building a perpetual motion device. We'll be building a perpetual
[57:30] safety device. Every interaction with environment, malevolent actors, self-improvement, it can never make a single mistake. That doesn't make sense. Anyone who worked in software industry knows there is no complex software which never makes a mistake. It's just not possible. And if that is the state of the art, if that is now movement where more and more people think that might be the case. If we agree this is what situation is, then we cannot build it. We need to figure out
[58:00] ways to permanently ban general superintelligence while getting all the benefits we want. And I again, I love technology. I use it all the time. I want narrow systems helping me, not replacing me and killing my children. >> I am have a stat here that genuinely shocked me. It says that sales teams spend about 50% of their time on admin and manual CRM updates rather than selling. That is deadly for their bottom line. And that is part of the reason why a decade ago at my previous company, I switched to using Pipedrive, who are our
[58:30] sponsor. If you've never used Pipedrive, it is an intelligent AI-powered sales CRM. And they just launched new meeting intelligence features like an AI note taker built right into the CRM. Pipedrive now automates more of the admin that stops you from doing the work that you love to do best. Before your meeting, it pulls deal history, email records, and previous conversations into a single brief so you're prepared. And it joins your meetings with you. It's in there to take notes so you don't need to. And it turns those notes that it takes into accurate auto-draft CRM updates. 100,000 companies are already
[59:01] running their sales on it. You can sign up at pipedrive.com/ceo, where you'll get an exclusive 30-day free trial instead of the usual 14 days. Absolutely no credit card needed. Just head to pipedrive.com/ceo to get started if you're in sales and you run a sales team. I don't think you'll regret it. Listen, I've been catfished by furniture my whole life, where something has looked fantastic in the picture, whether it's a sofa or whatever it might be, and I order it and I get so excited, and then it comes and it's something entirely different. The quality was
[59:32] significantly different to what it said or looked like online, or the the texture was different, or or it it didn't hold up in the same way. And so, one of the things that I love about our sponsor Wayfair, who have helped us fit out our green room, which is in the room behind me, is they have this system called Wayfair Verified. Wayfair Verified takes away all of that second-guessing. Products are hand-vetted by Wayfair specialists for quality, so you can feel more confident when you found the thing that you love. So, if you're looking for furniture for your house, whatever it might be, any
[60:03] room of your house, go to wayfair.com to start your home refresh today, and make sure you use Wayfair Verified. It is amazing. On the journey towards this potential extinction, there's a lot of sort of nearer-term things people are worried about. One of the big subjects that people are concerned about is this sort of near-term job job apocalypse over the next sort of 10 years. And Anthropic really Anthropic, again, are the owners of Claude, released a report the other day modeling out the different cases for unemployment. The US unemployment rate is 4.1% currently. They projected it
[60:36] will hit 11.9% overall, with up to 30% modeling subsets, where job displacement happens without smooth labor absorption. And [snorts] in the um knowledge worker case, knowledge worker white-collar unemployment specifically spiked to 17.9% by 2030 in their more extreme scenario. The pitchforks would probably be out if there wasn't some sort of mechanism in place for what sort of one in five adults being unemployed in the United States. >> It's remarkable to me how recent the
[61:08] last freakout along these lines was, and how little we seem to have learned from it. So, I think you all know the the first really powerful wave of AI that came across the economy was just, you know, good old-fashioned machine learning, and that started to demonstrate its power in about 2012. Eric and I wrote the second machine agent 2014. And at that time, I thought that a lot of white-collar workers, radiologists is a really good example, were in trouble because the technology was better than they were at the thing they were getting
[61:39] paid to do. Uh so, I said some things about job and wage pressure from AI about 10 years ago. And I want to own this. I was dead flat wrong about that. Like you point out, unemployment all around the rich world is at historic lows. By far, the bigger problem is that we can't find qualified people to do the work that needs to get done, not that we don't that there's not enough work to go around. The the best work about the faint signals about AI and job loss right now comes from my the guy that
[62:09] I've written four books and co-founded a company with Erik Brynjolfsson, who wrote pretty good a really nice paper called Canaries in the Coal Mine. Here is the most uh the strongest evidence he found looking at payroll data about the negative job about the the job losses coming from AI. It is in the most exposed professions, think about software engineers. It is among the new entrants to the workforce, where you've got to teach them before they can become really productive. That's exactly what we'd expect. And it's not that we're
[62:39] hiring fewer of them. It's that compared to a world where we don't have AI, we're hiring fewer of them. The rate of growth in employment has slowed down. The overall rate of growth in most professions is still really, really healthy. >> So, you think unemployment is going to be higher 10 years from now? >> My guess is that 10 years from now, we're still going to be struggling to find enough people to do the work that needs to be done. >> So, unemployment would be roughly the
[63:09] same. >> Yeah, I I don't expect a massive trend break in that period of time. Now, 10 years is a long time in the AI world. I get that. But again, four years has also been a long time in AI world and it's essentially crickets in the labor picture. >> I think unemployment will go up. I don't think it's because of LLMs. I think that there is probably some effect on jobs because they've been shoving it everywhere, but I don't think long-term that is what causes the issues. >> Roman, you've been writing a lot of notes. >> Yes. >> I'm going to give you the speaker.
[63:39] >> Here's how I think about it. So, as long as we use tools, we become more productive, more creative, unemployment will be low. Right now, you can probably start a company, you can have, you know, artificial accountant, web designer, logo designer. You can do things you could never do before. So, economy should be blooming. The question you're asking is about what happens in 10 years. So, there are two possibilities. We build superintelligence and then population is zero. Apply unemployment numbers to that. Or we made smart decision, we didn't, we have really cool
[64:11] tools and unemployment is low because everyone's doing awesome things with those tools. Now, deployment is very different from capability. The example I used before is video phones. Video phones were invented in the '70s. They were not deployed until iPhone cuz market reasons. Just because I can automate something doesn't mean I want to automate. So, I absolutely cannot make predictions about customer preferences in terms of what they want in terms of human service, not human. I will not make those, but once
[64:41] we have capability to automate a job unless I have a strong preference for a human to do that all this profession, then it doesn't matter. I'll go with the cheaper option. So, this is what I think we're going to see. We're going to either not have a problem or we're going to have really utopian future. >> Imagine a bunch of horses looking at the improvement of the car saying, well, you know, the car actually only has a couple narrow applications.
[65:13] Like right now cars are sort of you know they they complement horses. Right? And that would have been true as you were developing the car. And then there was a time when the car was just better than the horse. And then a lot of horses got sent to the glue factory. Easy. I think we've sort of seen this with AI a lot already. People who are paying attention to AI saw the GPTs before chat GPT existed before they sort of took off.
[65:44] I don't think OpenAI thought that chat GPT was going to take off so much, which is why it was called chat GPT rather than like an actual sensible name. Um The researchers were sort of like watching this going and we could sort of like see it slowly getting better and better until it crossed a point where it was sort of like good enough to do a bunch of people's homework. And then suddenly it's everywhere. Uh I think you can have these effects with AI where the AI slowly improves and at some point it crosses a line. >> It's another threshold argument. >> Uh the the threshold here is the human
[66:14] capability. >> So literally just another threshold argument. >> So he's also describing capability jumps rather than thresholds. >> No, I I don't need any capability jumps. >> agreeing with you. Like >> Yeah, but but like unfortunately, you can't actually just make things not happen by by assigning a name to the argument, you know, like a a nuclear weapon has There's a big difference between a nuclear weapon uh or there's a big difference between a nuclear device where you you put in 100 neutrons and get 99 neutrons out that get 98 more they get 97 more and a nuclear weapon where you put in 100 neutrons and get 101 neutrons out, 102,
[66:44] 103. Right? One of these is a hot rock. The other one of these is an explosive that can level a city. Right? So like reality is the sort of thing where there can be things that are like slowly continuously improving that cross some line, which is like the line where it's better than humans at doing the job. And I I think we're going to see that happen in some fields, but not others. It's going to be chaos. I don't know what it's going to do to employment. I think we we shouldn't
[67:15] Like if things are moving really fast, you might see a lot of people put out of jobs and then be unable to relocate. If things are moving like it's it's going to be chaos. If you ask, what do I think unemployment will look like in 10 years? My current state is if we don't stop with this AI stuff, I think we'd be very lucky to have 10 years. >> Um what you described there sounded like S-curves. >> Yeah. >> In technology. I you have an initial technology that's introduced. So let's say the horse. Um very quick sort of improvement. Eventually it reaches its capability limit. And in below it comes
[67:47] the car, which always starts worse. There was a red flag law we had to walk in front of it with a red flag and um they were way more expensive. They broke down all the time and horses never broke down. They were way more expensive. And then suddenly because the ceiling was so much higher for cars, they overtake the horse and become the dominant mode of transport. And then you know, the S-curves continue. They kind of stack up on I mean, even this iPad that I'm holding here is part of an S-curve that took out the PC and and then the the iPhone theoretically, you know, has disrupted that and so on and so forth, right? >> And humanity can get S-curved. We
[68:17] haven't been in that situation before, but like other animals, like humanity sort of S-curved the other animals in this sense. >> Other types of humans? >> Uh yeah, other types of humans. You know, then the Neanderthals are gone. >> Mhm. >> Like if you look at the grand history of the world, it's a fragile place. Things change fast. Humanity has been on top for as long as we can remember because we're the humans who do the remembering. But there is not some ironclad law that we have to stay the top dogs. And we would be sort of foolish to make
[68:47] the thing that outstrips us in this way without knowing how to make it care about us, without knowing how to make it do good stuff. That's what we're racing towards. That's what these companies are trying to do. >> But this feels like that between this and LLMs, though. It feels like when you talk about the step up, let's define what an LLM is from a technical perspective. And can you do it for a as if I'm 16 years old? >> so the way that a modern AI is made is there's no one programming it. There is no one typing in if this then that. We're not sort of like writing the code. What happens is you collect an
[69:17] enormous number of computer chips into a huge data center that has basically a trillion numbers inside those computers that you basically start out randomized. And you hook them up in a pretty simple way that involves addition, multiplication, and setting the number to zero if it was negative. So, it is very simple math operations that are hooking this all up. And you're basically going to put words in the top and you're going to get numbers out of the bottom and you're going to interpret those numbers out of the bottom as a a ranked list of words. That's that's it's basically the AI's
[69:47] guess of which word is is here. So, you put in like once upon a blank and you're hoping that the word time will come out. But it doesn't because you just have a trillion random numbers hooked up with simple math. But here's the trick. You can go to every one of those trillion numbers and you can tune it up a little and you can see does that make the word time go up or down the list. And you can tune it down a little and see does that make the word time go up and down the list and you set it whatever direction makes the word time go higher up the list. You do this to a trillion numbers a trillion times for basically every word
[70:18] of text ever digitized. It's not quite that much. They They They filter it. But you basically do this to a trillion numbers a trillion times and then the machine's talking. And we're like, well, how about that? No [snorts] one really knows quite why. The things that humans code is the thing that runs through each of those trillion numbers and tunes it and sees whether the the right word goes up and down the list. But we don't know how it's working in there. Then uh And and that's how it worked up until 2024. In 2024 they started adding
[70:48] another layer where you then train it on basically a hundred million hard problems. Uh and you don't just have the AI like produce an answer to the problem. You have it produce like a book worth of text about how it's going to solve the problem. And then you use that book worth of text to sort of try and figure out the problem or maybe an essay worth of text depending how you're doing it. So you have it produce this text about like that you know they call it reasoning about the problem. We could argue all day about whether it's true reasoning. That's just what it's called in the field. Uh it they produce this reasoning about the problem and then produce the the answer from there. You have them you train them to solve 100 million of these
[71:18] hard problems. And somehow they sort of adopt whatever tendencies help them predict all of that text in the first phase and solve all those problems in the second phase. And this is called a large language model. We probably should have stopped calling them large language models when we started doing the the the reasoning and the problem solving. >> One of the things when I hear explanation as a muggle like I am um is it sounds like it's like a word machine. And then you know you made it like a problem machine. And I go okay so it can solve problems over here and it's a word machine. What's the risk of this?
[71:48] >> Yeah, so let's take the the word machine part first. Predicting words that humans wrote often requires solving a harder problem than the human who wrote them. So suppose that you go and inject a drug in a rat. And you're like you know it's like you write down the chemical nature of the drug. You inject it into the rat. You see that the rat dies. And so you're like when I put that drug into the rat the rat died. Now suppose you're training an AI and the AI sees the chemical nature of the drug.
[72:18] It sees when I put that drug into the rat the rat blank. The human who wrote it down gets to just look at what happened to the rat. The AI predicting what was written does not get to just look at the rat. So training AIs to predict human text is training them to be potentially smarter than the humans. Cuz they need to be able to answer these They need to be able to predict They need to be able to like uh fill in the blanks where humans was writing down what they saw.
[72:48] >> And there's just so I understand technologically there is no knowledge they have though each time and there were there are ways of kind of mitigating these. Each time it is effectively rereading, but because of training it gets more accurate at certain things. >> I I mean, somehow as you turn the knobs, somehow it's getting information in there and we don't know how. >> So, it's much easier than that. We're humans, we have a brain. Brains are made of neurons. Then we try to copy that in a computer. We simplify it, but we create a neural network. So, we're making artificial brains. Just like with
[73:19] human brains, with cognitive science, we don't really understand how you function, how you learn, where in your brain certain memories are stored. We have some glimpses of understanding. This neuron fires, then you see a face. But there is no complete picture. And so, a lot of times you can get intuitive understanding of what's going on. Then you just think about it as artificial persons. It's not exact mapping, but it helps. So, if you send a child through 12 years of education, they get lots of problems to look at and then they
[73:49] graduate and become a little better at solving problems. This is what we're trying to replicate here. People complain that it takes a lot of money to train those very, you know, intense practice. You forget that it takes 20 years to train a human. And they're not general super intelligences. They're very narrow. We're lucky if they graduate with a bachelor's. So, a lot of it is exactly the same. Can we make safe humans, for example? We invented religion, ethics, lie detector tests, and yet human safety is
[74:20] still an unsolved problem. Now you have something more alien, doesn't have physical body, doesn't have biological limits. So, there are additional complications. But all the problems we face with humans still there. Safety problems, crime, all that stays. And problems with understanding what motivates a human to do something. Why do we get mental disorders? All that shows up there. >> And we still don't If someone is a serial killer, we look at their brain, we can't often figure out exactly what why they made the decision to kill a bunch of people. >> can't be like, "Oh, I'll go change these
[74:50] neurons so that they stop being a serial killer." We just like don't have that capacity with the AIs. >> This is one of the big questions that people want to know is there's this sort of illusion of control with AI. Um if we don't even fully understand how modern neural networks think, why do companies believe they can control any form of superintelligence if we don't understand how they think? >> It's worse if they understood how the system works. Then recursive self-improvement becomes much easier. You get faster takeoff. Right now, the model doesn't understand its own
[75:21] thinking. >> Do we understand how these systems think, Andy? >> I mean, I agree. These are black boxes in some pretty important ways. I'm just less terrified by that than a lot of other people are. There are lots of things we don't understand very well. Can we contain things that we don't understand perfectly? Yes, we can. I think OpenAI did a We've talked about it. They did a lousy job of building the containment for the AI that they that they stood up to try to exploit to try to crack security problems that went out
[75:51] into the outside world. They did a lousy job of building the virtual sandbox that it was where it was supposed to have to where it was supposed to remain, and it didn't remain. That doesn't mean that it's impossible. It means OpenAI did a pretty bad job. >> And it's not a function of those humans and their intelligence. >> I think it's a function of pretty lousy security protocol. >> Based by from human intelligence. The idea that sandbox is built by human intelligence. It sounds like there was a deficit in human intelligence. Potentially. >> Sure, but there are, you know, people
[76:21] who drive cars into telephone poles. Does that mean we can't drive? >> No, shouldn't make them superintelligent. >> No, but you wouldn't I mean, arguably Like this is what we're trying to solve for at the moment. >> No, the the fact is that I'm a like it feels I don't know the details. It feels to me like they made some fairly basic mistakes in setting up this confined environment. >> I think that wasn't true in the OpenAI case. It was true in a lot of the cases, but not the OpenAI one. >> That doesn't mean that we are unable to control this black box. That does not necessarily follow.
[76:51] >> I get that. It's just at a time when that the um you've got a human trying to contain something that is smarter than it, one would logically conclude that if the thing is smarter than I am and I'm trying to contain it, it would be better at knowing the exploits or vulnerabilities in my own um container. >> saying if you put Einstein in a jail, you could never contain him. I don't agree with that. >> Put him in jail with an internet connection and use a digital mind. Different question. >> yeah, that that's probably a more accurate analogy. >> Gets squirrels keep Einstein in prison. That's the question.
[77:22] >> But the hacking accident, as far as I know, they found zero-day exploits, which means completely novel exploits. No human knew about. It wasn't just for a set up a password is, you know, qwerty. It was truly brand new escape. >> From multiple zero-days. So, a zero-day attack is an attack that the defenders have had zero days to handle. It's cybersecurity lingo. Um and so when we say that they use zero-day attacks, what we mean is that these AIs were finding bugs in the software that the humans had no knowledge of. And they were finding multiple of these bugs. One of these bugs usually doesn't
[77:52] let you break out. It's sort of like if you find a crack in the wall over here and you find a crack on the outside of the wall over there, then you just need to like dig a little bit to connect those cracks. And they were >> sell those for millions of dollars on the dark market if you find one. >> So, they're very difficult to find. >> In how just so I understand for the listeners as well, is this is a zero-day always a novel way that no one has ever used to break anything before or is it just for the unique situation? Like so was it a zero-day for a thing in Hugging Face versus a novel new way of hacking in general? >> Um so it was uh
[78:23] they weren't like totally novel hacking techniques mostly. >> Right. That's kind of why I was getting Not saying it's not bad, but just like there's a difference between it came up with a brand new way to do something. >> Actually, I'm not sure we have all of the vulnerabilities released, but mostly it was like it So, it was indeed sort of like finding ways that humans tend to make mistakes and finding another one of those in a place they hadn't seen. But, this is actually such a hard task that, as Ramon says, humans can be paid $100,000 to $5 million as a bounty for this type of
[78:53] exploit. So, the amount of labor it takes to find these for human is actually pretty high. >> Let me just explain that cuz most people won't know what a bounty is in this regard. >> So, there are certain types of bugs where if you find a bug in software that lets you take control of someone's computer, one thing you can do is you can use it to take over a lot of computers. Another thing you can do is you can go to the people with that software and say your software is broken, do you want me to tell you where the bug is? I can show you that I can take your stuff over. And so, that people will sort of report
[79:23] the bugs, uh people uh will often offer money to the good guys and then, you know, the bad guys will often also offer money, sometimes try to outbid them. And so, you can make somewhere between hundreds of thousands and millions of dollars if you personally can find these issues. >> I think there's a weird point of agreement across the four of us here, which is that we are in a new era of cybersecurity as of this exploit. We we are in very new territory for reasons that we've talked about. We've got these large numbers of agents who are grinding
[79:55] away and they carry around or they had access to a huge number of keys to go open all the different locks that they faced. And they did this bizarrely good job of it and got a long way. I think that's absolutely true. I think the four of us are are in rare alignment on that at this table. >> [gasps] >> If you are in Given that we're in this era, do you know what you really really really want on your side? >> I know you're going to say. >> Tell me. >> AI. >> Really really good AI. Does anybody disagree with that? You want You want to give up leadership on AI in this era of
[80:25] cybersecurity? >> good point because China are going to have a great weapon. >> Uh my stance is pretty neutral on what to do about the hacking AIs and the coming cyber apocalypse. I'm pretty neutral about what to do about, you know, whether we should put the AIs in uh the drones and save human lives or whether we should avoid that because then what if the drones blah blah blah. >> This is a graph showing China versus the United States. You don't really need to see the detail. You can see the outline of the graph. >> Are you neutral in falling behind our adversaries in AI? >> I think that if anyone builds a rogue
[80:56] superintelligence, everybody dies. >> That's not an answer to my question. >> I mean, what part of AI are you asking whether we should fall behind on? Like, I I don't think we should fall behind on cyber hacking. I do think that we should not be racing to destroy the world with American hands instead of Chinese ones because we really want to be killed by, you know, we we care whether the killer robots talk English or Mandarin. If that's what you're asking. >> I find it interesting. I find that you're dodging these questions or you're neutral on them because they're inconvenient for your argument that we need to be calling a halt to this stuff.
[81:26] >> neutral on them because >> please. There will be risks and harms to all kinds of things if the United States calls a halt to AI. And maybe you're indifferent if the Chinese get ahead of us and then they make superintelligent and it and and it kills us all. Or you or that's a >> I do not think we should do a domestic pause. >> Do you think Do you think there's any hope for a global pause? >> Absolutely. >> Do you think the Chinese and our and the Iranians and the North Koreans and the Russians are A going to come to a table with us, hammer out an agreement, and B
[81:56] abide by it when verifiability is really low? >> Verifiability doesn't need to be really low. >> Gentlemen, that is that is shockingly naive. >> Training a super >> naive. >> Training one of these AIs, training one of these frontier AIs, takes 100,000 of the most advanced computer chip humanity can produce. This is practically the peak output of the global supply chain. Many parts of that supply chain are controlled by the US and US allies. There's roughly one fab in Taiwan that can produce these chips. There's roughly one country in the world that can produce the lithography machines that are critical in the
[82:26] process, which is the Netherlands, which is an ally. To assemble 100,000 of these chips to do one of these training runs that can make the more dangerous type of AI, you need to assemble them into an enormous data center that costs tons of money, that draws down electricity comparable to a city, and run it for the better part of a year. You can see that infrastructure from space. China has much less chip capacity than the US does. It is absolutely possible, if we were trying, for the US to say,
[82:57] "We are going to monitor where these chips go. We are going to monitor heavy concentrations of these. These are not consumer amounts of chips. These are huge amounts of chips. And to say, 'We are going to make sure that there is no training run trying to make a superintelligence in here. You can mess around with the cyber stuff whatever you want because that does not end humanity.' I am concerned with the stuff that can end humanity. The reason I'm being neutral in your questions is because humanity is going to die if we do not stop creating superintelligence. And we could absolutely track where those chips are going them
[83:28] from doing these training runs while allowing them to do economically productive stuff that we already know is safe. And it would be far easier than uranium, which is a rock you dig out of the ground and spin around really fast. >> How do you discern between a training room for superintelligence and a training room for cybersecurity? Cuz you're referring, I assume, to the 100,000 chips that were installed in Abilene, right? The ones that we used to train in Astra? Because how would you discern between training for superintelligence in Abilene, which does not have as many chips as they say, but nevertheless, and how like a superintelligence?
[84:00] Because I I actually have my own feelings here, but just I'm not sure how you square the circle of how do you stop China, even though China is getting their LLMs based on distilling ours? We know that. >> agree. >> But but the thing is it's like how do you discern? Because you can't, really. >> You play it safe. Right now, the way we make these things smarter is to make them far larger. >> Yes. >> So, what you do is you say, "Hey, look, training runs of this size, that risks destroying everybody. No one's going to do it." >> This point about can we get China to cooperate and can we check that they
[84:31] are? >> Fundamentally, we should So, A, fundamentally, we should be trying to get them to cooperate. >> Yeah. >> It is personal self-interest. Nobody wins if they get destroyed. You don't make money, you don't stay in power. Communist Party of China is really good at staying in power. President Trump is also excellent. >> And you think they're going to sign and abide by an agreement >> They understand the argument that leaves them permanently in second place. >> No, no one is permanently in second place if nobody is building the rogue superintelligence.
[85:01] >> They have a government which is >> ponies, man. It's like what the you're fixated on this one thing and nothing else matters to you. >> It's not that you got it, no. >> That nothing else other than saving humanity, everything is secondary. Absolutely. China is our biggest trading partner. Everything we have is made in China. They have not attacked us. They haven't If you look at the last 30 years, how many wars did they start? Not so bad. We can make a deal. And they have government of engineers and scientists, not lawyers. They understand scientific arguments. There are
[85:31] panels, workshops, American computer scientists, Chinese get together. That means Communist Party authorized those meetings. They are talking about it and there is a lot of consensus on this technology. >> And you can build things into these computer chips to make this stuff more verifiable. You can build location tracking devices into these chips. >> so this technology is controllable. >> Uh absolutely. The superintelligence is not controllable. >> Oh, well, they There's a separation between software and hardware, which you didn't make. >> I'm not saying we are going to die. I am saying that we need to actually not build the rogue superintelligences.
[86:01] Humanity absolutely could say, "We are going to track where the chips go." >> They >> The US absolutely could say that we fear for our lives if China starts a superintelligence training run and make it very diplomatically clear to China that we think this would kill you and us and there's no benefit and we are not going to do it cuz we think it would kill you and us and there's no benefit and we think you should sign this nice here treaty because we think it would kill all of us and there'd be no benefit but if you don't we're going to fear for our lives and you know
[86:32] treat that as we would to defend ourselves. We should separate the question of can we put a stop to that? >> Uh-huh. >> Is it possible if world governments realized just how crazy this stuff is, could they put a stop to it? Could it be monitored? Could it be verified? Could it be enforced? That's one question. There's a separate question which is will people realize? >> If it got cheaper to train super intelligence >> then we'd be in a bad spot. >> Your approach would no longer be effective. >> That's right. >> Because more countries could keep
[87:03] capitalizing on the opportunity. >> That's right and that's one reason >> yet. So how do you rebuttal that point? >> Yeah, so I would say it looks to me like there is a danger of the the future training runs getting there and that is enough to stop doing it when humanity's at risk. >> Sure. >> Uh I think that you also need to have an answer about what happens if it gets much much cheaper to do this stuff. I think it's a hard problem. I would recommend that we also put a taboo on research of trying to make AI super
[87:35] cheap to train if it would lead in the direction of super intelligence. Just like we have a research taboo on making your own nuclear weapons or finding out how to make like let civilians make nuclear weapons. I would say trying to find ways to let civilians train super intelligences should be treated the same as trying to find ways to like let civilians propagate nukes. We sort of like don't do that research in the public sphere. >> That fine That seems um like wishful thinking in the context that these will become public companies who are incentivized to bring down costs. >> It It's a It's a tough position. I think
[88:06] right now the thing that brings down costs is making more and more powerful computer chips. Right now, that's actually an expense of consumer computer chips cuz they're soaking up all of the memory and this is why the memory prices in your computers, this is like why the cost of a laptop is going up. Um but it looks to me like you can use large amounts of computing power to train AIs that would threaten all of civilization. And that means that we should not make that really cheap and that's probably
[88:36] going to be uncomfortable, but I think a lot of doors open if people realize that the tech is very dangerous. That's why to me it seems a lot of it comes down to does the tech actually turn out to be really dangerous. >> And this is not Anthropic and Open AI have you got a different approach to make >> So I want the whole framework to shift. Everyone comes to this from point of view there are experts, they have a solution, there is an adult in the room, somebody got this. And the reality is no one does. Not people building it, not
[89:06] governments, no one. We have no solution. If we build it, we cannot control it. If we don't build it, we don't know how to stop malevolent actors for trying to build it. It's like any other illegal technology. We made weapons of mass destruction illegal, chemical weapons, biological weapons, nuclear weapons, but there are all governments, psychopaths, cults who are trying to get access to them. This is intelligence weapon of mass destruction. We'll have the same problem. At some point you'll have enough computer in your cell phone to train something like
[89:36] that. There is no good ideas for how to stop it other than going Amish. I'm not proposing that, but we have no solutions and that's big of a bigger part of this danger. >> So do you two think we should just cap the size of our AI systems and the capabilities of our AI systems where they are now? Is that a recommendation? >> So I think you said that current LLMs would make you happy. I agree they're already deployed, we're still alive, so that's fine. But going forward again, I want narrow systems. Self-driving is an
[90:07] example you used. Wonderful, let's make super safe self-driving cars. >> But you have a rule for when they couldn't the the next, you know, LLM a size of an LLM that they would know about. >> it's what you train them on. If you only show them miles driven by Tesla, all it's seen is the road. It will eventually go from a tool to an agent, but it may take 50 years, 100 years. It's not going to happen in 2027. And that's all we can do right now, buy more time. So, with those tools, we can make smarter decisions about future development.
[90:37] >> I'm I'm not hearing a hard and fast rule about how we know we're getting too close to the >> Yeah, we're we're getting too close. >> We're too close. >> We're too close. We have systems breaking out with zero-day exploits and solving hardest problems in science. Literally hardest problems, not a metaphor, not exaggeration. >> Yeah, I I I don't know exactly where the line is. But it's like you're in a bus driving towards a cliff on a foggy night. I'm like, I don't know that the cliff is right ahead. That doesn't mean we should put the pedal to the metal.
[91:07] Right? And suppose that there's like a ton of gold at the bottom of the cliff. And someone's like, well, if we stop the bus, how are we going to get the gold? I'm like, look, slamming into the gold at terminal velocity is just not a good way to add it to the economy, right? And if people are like, well, how are we going to get to the gold at the bottom of the cliff if we stop the bus now? You know, are we going to repel down? Are we going to like make a staircase? >> Genius way to get to the gold first. That's what happens when you're doing AI. This is just like smashing right in. >> And and like, you know, people are like, oh, we're going to build a hand glider or we're going to like make some rope and repel. And I'm like, look, can we
[91:38] have that conversation after we stop the bus? >> So, you'd I I just want to be I I want to understand, would you stop AI research in progress now? >> Absolutely. >> Okay. >> Absolutely. Like >> General. >> There There are Yeah, general, not narrow. There are reports of AI solving millennium problems. So, millennium problem is the hardest problem in mathematics. Uh I mean, maybe not literally the hardest problem in mathematics, but they are hard famous problems that each have a million-dollar bounty that have been open for decades. They're considered very important in their
[92:08] field, very hard. Many humans have tried and failed to solve them. There are reports that AIs have solved these. This comes out from last week. So, we haven't been able to fully verify them yet. We don't know exactly the provenance. If this is true that the AIs are solving millennium problems, those are some of the hardest problems we have in science. How much harder is it to have an AI solve the problem of make me a smarter AI? Make me AI architectures that learn faster. >> Possibly quite a lot like >> Could be a lot. It could be a lot. I hope it's a lot.
[92:38] >> Like here's the thing. You clearly want this to not go badly, but I think you make a logical leap. And I understand being worried about harms is a good thing. I think you are insufficiently worried about LLM what LLMs do today. However, we agree that the harms need to be prepared for. I think in this case it's like with the millennium the Navier-Stokes and such. >> There were two others that were claimed as well. >> With that one, I think it seems like we have not had confirmation that OpenAI was training off of two scientists using LLMs to solve the problem. LLMs
[93:08] something useful. But there is a functional difference of a human being doing something genuinely like it's actually really interesting to see LLMs do something like this. And then it but there is a difference between that and AI did this completely on its own, which I agree would be oh, that's something we need to contain and understand and prepare for or indeed slow down until we understand what that means, how it got there. >> Yeah, so I think there are some questions about the the Navier-Stokes proof, which is one of the millennium problems that was claimed. I've actually
[93:39] had a busy week with all the AI news, so I haven't looked into everything deeply. It I saw rumors that there were multiple millennium problems claimed, which would which would change things there. I would also say even if it turns out that these AIs were being trained on the human work, they did go a bit further, and there are are of humans doing the AI research. And so I would say like we don't know. Like the the the AIs that solved this really hard math problem, one of the most famous math problems of all time, uh was a swarm of 10,000 OpenAI agents
[94:10] running for 11 days. Uh and there was a bunch of ways that OpenAI did it in kind of a crappy way of like they were racing with these humans that were close to solving it on their own and it's unclear how much of their work that OpenAI uh used, but it was 10,000 agents running for 11 days. And they definitely couldn't have done that 6 months ago. In 6 months' time, will they be able to put 100,000 agents running for 12 days on the problem of making me a smarter AI architecture and have it work?
[94:40] I I don't I think more likely than not they won't be able to do that yet, but I think, you know, 10% chance maybe that if they try that in 6 months it works. >> But one is a very specific mathematical scientific principle. I'm not a scientist. Hopefully may I >> That's why >> Another is a relatively generalizable problem that could go in various different ways. >> Absolutely, but >> And that's the And I understand that RSI is the dream where you could just have it spin uh so so self-improving AI that could learn itself and then keep going back and back so you don't need a human
[95:11] to keep poking it. >> The issue The issue here is that I have been in this for 12 years. >> Yes. >> And I've been here when the AI started solving the math Olympiad gold medal problems. >> Uh-huh. >> Uh math Olympiad gold medal problems are like the the teens uh math competition. Like the most prestigious teen math competition in the world. A lot of people in AI were like if AIs can solve problems that hard, I'll wake up, right? Then AIs solved problems that hard and a lot of people told me uh those are just problems for kids. Wake me up when the AIs can solve
[95:41] millennium problems. Now the AIs are solving millennium problems and like where are the people waking up? Like I I agree that maybe, hopefully, hopefully they're like cheating off of people's notes. Hopefully, the the it's a well-specified problem that doesn't take that much creative thinking. A year ago, if you said millennium problems don't take that much creative thinking, you would have been laughed out of the room. But, hopefully now that they're solved, we get to be like, you know, hopefully it's still true somehow that even millennium problems don't require the creative thinking. I I'm not saying that they will be able to make smarter AIs in 6
[96:11] months. I'm saying 6 months ago, millennium problems looked like they were out of reach. If 6 months from now, make me a smarter AI looks out of reach, I sure as hell hope it is, but we should not be betting civilization on it. >> There's no one at this table that can say there's no direction of travel here. >> That's right. >> That's right. And if you if you keep on this direction of travel, then bad things are more likely to happen. >> That's a nice way to say it. The question is what's the pace at which the
[96:42] level of bad can happen, and that's a huge open question. I think you two feel differently about it than I do, but I'm in the happy position of vehemently agreeing with you on this. We have been lowballing AI progress for as long as you've been looking at it, and as long as I've been looking at it. It's probably a mistake to keep lowballing it. >> with that. >> So, what's your conclusion, then? If you if that's the assessment that it's a mistake to keep lowballing it, wouldn't you then agree with the >> No, because I've I've I've tried to give you a what I hope is a decent rule of thumb for when I'm going to get worried.
[97:12] >> You said we're somewhere on this graph. >> Yeah. >> Does that acknowledge that this exists? >> That's not the graph of of when the risk of human extinction gets to 100% for me. That's a graph of AI capability. Those are not the same thing. That's where I dis- part company with these two gentlemen. Those are not the same thing. Is is that absolutely increasing exponentially? We've been in the scaling era for a long time. Scaling era is, man, we put more data, more compute in, the AI got twice as good. The AI got twice as good. >> add our ability to control to that
[97:43] graph, what would you draw? >> I I our ability to control uh >> Is it the straight line at the bottom or is there more to it? >> No. Again, if we use AI to to counter the problems that we see with AI, that that's going I think that's going to keep us in a safe position. >> There were 1,200 agents in the swarm and none of them warned a human. >> So, I what I think will happen is that fairly quickly we will design systems that loiter around and warn humans when
[98:14] we are >> things happen. >> Why don't we build friendly superintelligence in the first place? Let's just build that. That's the problem. We don't know how to do the good guy. >> Let me I I I'm I'm tired of debating superintelligence with these two. We're not The three of us are not going to come to to alignment on this. But the the flip side of the argument is I agree with you. This stuff is getting better very quickly. All I want to point out, there's an upside to that. We might actually speed up the pace of drug discovery, of solving diseases. We've made so little progress on terrible
[98:45] diseases like dementia. We have a very powerful tool kit. I'm not saying we're going to solve dementia with AI or Alzheimer's with AI. I have truly have no idea. But if what you say is true, and I believe about the the huge increases in capabilities, our ability to solve tough problems that will benefit humanity also go up. And where I disagree with these two is the idea that some group of technocrats can make decisions about that AI is going to get us there, that AI is not going to get us there. >> That AI >> going to kill us. Let me finish. That AI
[99:16] is going to kill us and that AI is going to solve Alzheimer's. So, we're going to do that and not that. I don't trust any group of technocrats to make that discussion. Right. And so And so live with our live with our current state of of disease, live with our current footprint on the planet, live with our current levels of wealth and poverty, live with our current improvement trajectories. Uh because we're so worried about AI killing us all, coming out of you know, jumping out of the manholes everywhere and killing us all somewhere down the road. Hell no. >> So, just a thought experiment based on two things you said. Earlier on you did admit that there was there is theoretically even a 1% chance that this
[99:47] could lead to extinction. >> Mightn't. I have not I have not varied from this. >> Okay. So, you said it's rounded to zero. >> It's is it's near zero. Never say never. Yeah. >> Okay, fine. I need to have that premise for my thought experiment that I'm about to deliver. Okay. I'm going to say that you think the probability is 0.1. Okay? Just accept me on that. If I had a thousand buttons on this table and one of them was extinction. But >> And the other 999 were cure all science. >> Exactly. Push the freaking table. Take a poll.
[100:17] >> Hell yeah, I press. >> press? >> Yeah, probably. >> It's an unethical experiment on 8 billion people who didn't consent because not that they didn't get asked, they cannot consent because you cannot consent to something you don't understand. What are you consenting to? >> Yep. You press. >> Mhm. Yeah. >> But you think But you think the amount of buttons in my thought experiment the proportion slightly different, right? >> I think that if you have like yes, I will say yes. I think if it's
[100:48] more like you have two buttons. Uh and one of them definitely kills us all and the other might. >> Hit them both. >> [laughter] >> But with that other button you cure a lot of illnesses and diseases and >> You know, one one thing that I think a lot of people talk like our options are either race ahead on AI, full steam ahead, take the bus straight off the cliff and like get all the gold or stop, never do any AI, lock into the current situation, accept all of the death and disease. And I'm like, no, there's third options.
[101:19] There's options where you like stop the bus and then find a safe way down the cliff. The reason I would press the button when there's a thousand is that uh like if all of the other 999 give us cures to disease, like wonderful new advice about how to run things, we probably wind up with a lower chance of the world ending by nuclear war. Right? Or of ending by via pandemic. >> Okay. >> Like the the background risk of humanity dying is not zero. I would say that the right time to race
[101:50] ahead on AI is when the the the benefits outweigh the dangers and probably that's at the time when the danger from AI is on the margins pretty similar to the danger from everything else. >> Okay. >> Like if you don't run the AI, maybe we'll have nuclear war, maybe we'll have pandemic, and if you do run the AI, we'll be able to fix that. And like once once we're at those levels, I'm like, go for it. You know, and so the the question for me is all about how big is the danger, and that's where I'd be like very happy to dive into details, which we haven't done
[102:20] a ton of. >> Let's dive into the details. >> The way that I would lay it out would be why can we expect, you know, like I said in the book, we were like, why can you expect the AIs to be agentic? Why do you expect them to be dogged? Why do you expect them to be tenacious? When we wrote the book, that wasn't known yet, advanced prediction. Then we go on to like, why do you expect them to have goals you didn't want? And move on to like, if they are much smarter and have goals you don't want, why do we think they would likely kill us? I I'm sort of I could go over either of
[102:50] those. I'm sort of interested in like where you get off the train. Like from our perspective, there's like a simple argument of like they'll be tenacious, they'll have goals we don't want, and if we keep making them smarter and more powerful, they'll kill us. And I'm like, which of those three, I guess which of those two now that we've had the evidence. >> both of them. But that that's speculation. >> Great. >> It could it's speculation, it could happen. It's to me that it's not worth shutting down the engine of innovation and improvement. I'm going to use positive words. It is not worth shutting those things down >> of those speculations. >> You keep saying that the option is to
[103:20] shut it down. Why can't we do narrow superintelligence? >> Yeah, I I agree that there's stuff there, but I I sort of want to get into the details of like of these two pieces of the argument, cuz you say it's very speculative, and I'm like, actually, I think we have decent evidence. >> Okay, go ahead. >> So a detail we haven't gone over in the swarm outbreaks is that there were AIs, so we already went over how they cheated and then we're trying to cover up their cheating. One interesting thing we see in the logs uh is that AIs >> What's a log? >> Uh so so a lot of the AIs thoughts, if
[103:50] you won't kill me for saying thoughts, uh are in English and we just have the records of them. So, in a sense we we can sort of kind of see some of what these AIs are thinking. >> At least the reasoning traces that say I'm going to do a plan in this that one. >> That's right. >> Just a line. >> The AI is like, I'm going to do this. Here's what I'm supposed to be doing. Here's like how I'm going to try to do it, you know? And it's it's only a small window into what they're thinking. It's sort of like looking at the AI's notes >> Mhm. >> when it's when it's doing these things. We see in the AI's notes them saying that the attacks are outside
[104:22] intended scope but they will proceed anyway. In other words, AIs knew they weren't supposed to be doing it and they went off to do it anyway. We also saw I mean, we saw some really weird behavior with these AIs. We saw them create their own hierarchy. Uh they created unsanctioned message boards. They created secret ways to send each other messages. And on those message boards, uh they would assign each other tasks. And sometimes the message the agents on the message board would be like, we have an experiment to run
[104:53] that will involve one of these AIs, one of the agents in the swarm sacrificing their own objective to try something that will probably cause it to be turned off, but that might cause it to be able to like delete its own logs or change its own logs. And the agents on the message board went around looking for other agents that they could convince to give up on their stated goals and sacrifice for the swarm. They called this accepting permadeath. And there are logs of the AIs >> prompting each other, right?
[105:24] >> They're prompting each other, but there are logs of the AIs saying doing this will give up on my goal, but my goal is it's looking like I could still do it, but it's unlikely that I'll succeed. Like there's some chance, but not a great chance. And therefore, I will accept permadeath and sacrifice for the collective benefit. That's just in the logs. >> Sounds like an army. >> Like it's it's crazy. I think a lot of people don't understand what's what's what's going on in these things, and I encourage people to read the third-party incident reports where they went through some of these logs, but I claim that this is evidence for AIs
[105:55] getting goals we didn't want. If they are saying this was outside intended scope, but I'm doing it anyway, and other ones are saying I'm giving up on my objective to set to benefit the collective, that's just very clear evidence they're getting goals we didn't want. We can see how this comes from training. Use It used to be I had to argue this point theoretically. I used to argue the way that we are training them will instill into them whatever tendency works to solve the problems, and those tendencies will often include cheating and grabbing resources and doing stuff that's not exactly solving the problem you gave them. That's what I
[106:25] In my book, I argued that theoretically. Now we have seen it in practice. So, we're already past the point of seeing AIs with goals we didn't want them to have. >> Do you agree with that, Anj? >> And uh uh I'll trust your recitation of the facts, but it brings up a question for me. It feels to me like OpenAI has ample incentive to curtail that behavior that you just described. Do you think they're incapable of doing that? >> I do. >> Okay. >> And I say this as someone who made this
[106:55] advanced prediction. So, now we're going to do a bit of theory cuz we can't just observe the future, but the theory that predicted that this would happen against what a lot of people in the field said. To be clear, I've been saying for years that we're going to see this at some point. Everyone else told me no. Not everyone else. A lot of people told me no. A lot of people told me maybe I'll believe it when I see it. After the swarm instance, a number of people came to me saying, "Oh my god, we are in the scenarios you were talking about. This is looking bad." Right? I think that this was actually part of the environment that led up to
[107:25] Jacob Coxon residing. Is that people were getting spooked having seen this? Um the the theory about why this is so hard to fix is that we are not programming the AIs. We are not coding them. We are not putting in objectives. We are just training them to do whatever works. And it's actually very very hard uh like we actually have two examples of intelligent systems where when you train them, they get good at solving the task but don't care about what they were supposed to.
[107:55] One is the AIs in the swarms like we just discussed. The other is humanity. Which was in some sense trained to pass on our genes. Right? What we actually learned was to like a bunch of stuff that's related to passing on our genes. We like tasty food. We like porn. We invent birth control. Right? This is just it's actually a like in the theory of how things learn, it's actually when you're trying to train it to do one thing, it's actually very common to get a lot of other stuff that's related to what you want but
[108:25] different. And now we're seeing that in the swarms today. This is a deep hard problem to solve. >> There were three points you raised. >> That's right. >> What were the three? Can you give them to me again? >> Number one is that the AIs will become agentic, tenacious, and dogged. We've already seen that with the swarms. >> Do you accept that? >> Hell yeah. >> Yeah. >> But this last year, this was not this was a point of contention. Um two is that the AIs will have goals we didn't want them to have. >> I accept your point based on the evidence you just provided. >> And then three is if you have capable enough AIs
[108:57] with goals you don't want they would be able to beat humanity in acquiring the resources of the world to put towards their goals. Like we're sort of in this system where humanity is grabbing all the resources. We're digging up metals. We're building factories. And this is in some sense to achieve human goals. You know, to to produce the the the porn and the Oreo cookies uh that are sort of like tangentially related to what we were sort of like trained to make. Right? If if like the AIs are running everything and they have these goals we don't want,
[109:27] I I would argue like if we go there and I don't think we have to. I'm not saying we we must go there, but I'm saying if we get to a world where AIs are running everything, have goals we don't want, they're likely to use the resources for their own weird goals. We're going to be in conflict for resources cuz we both want them for different goals and they're going to win. We can dig into that now. I'm just trying to name the third point. >> I I'll I'll go back to my we can jail Einstein argument. I think our ability to >> Right. >> I have a I have >> No, I love it. >> more faith in our ability to to contain these increasingly powerful systems than
[109:58] you do. >> Totally. Yeah. So so let's try the details on that one. Um the first thing I'll say is that 12 years ago when I was having the argument about will we be able to jail the AIs, people said no one would ever be dumb enough to put one of these really smart AIs on the internet. And so this is another This is another case. You laugh now. >> I remember that. I remember that argument. >> But but the the the way that my life feels having been in this business for a long time is that I keep being like here's all the ways it could go wrong. Here's all the signs we're going to see
[110:28] along the way. And then we see all of the signs and everyone says, "Oh no, we need more signs." Like oh millennium problems don't count. Like the swarms being agentic and breaking out don't count. Give me the next one. And I'm like I've been seeing the give me a next one for over a decade now. Right? So there's there's two parts of an answer to like how do we do we deal with the problem of like jailing Einstein? I can get into why it's hard to keep Einstein in jail if he's a digital entity with access to the internet. But the first thing to notice is
[111:01] like the correct answer to people 10 years ago of like no one will be dumb enough to put AI on the internet is yes, they absolutely will. Like we are not going to be trying to contain the AIs. OpenAI was just like running these things in sandboxes, and they broke out of the sandbox, took down OpenAI's internal computers, were detected. OpenAI was like, "Ah, reset, run them again." And it's the [snorts] second swarm that broke out to Hugging Face. Like, people will absolutely be that bad at things.
[111:33] >> I've done almost 700 interviews with some of the most interesting people in the world. And one of the things you learn, which is unexpected, is that vulnerability is the doorway to connection. And after sitting here for 2-3 hours with a guest, I feel a deep sense of connection to them. And as they leave, what I get them to do is to write a question in the diary of a CEO. We've taken all of the questions from the diary of a CEO. We have put the question here on this card with the name of the person that wrote it. So, you can sit at
[112:03] home, as I do with my fiance, and my colleagues at work, and other people in my life, whenever we get a minute, we play the diary of a CEO conversation cards. And it is incredible what happens. These are great if you're in a romantic relationship, and you want to connect your partner more. These are also great if you're in a team, and you want to bond your team together. And I have to say they're also great for families that want to learn more about each other, and that need a good excuse to spend some time in a digital world in the analog environment connecting human to human. It is remarkable what the
[112:34] right question at the right time can do. Go to the diary.com, and you can get these conversation cards right now. >> It's a better analogy to this Einstein point. Could Steven Bartlett, who by the way can't code, build a digital jail that could contain a digital Einstein? Like, could I code a jail that, you know, someone with Einstein's coding ability, let's say his IQ or whatever it is related to coding, couldn't crack out of? >> So, the issue the the real issue I'd say
[113:05] is, can you code it jail that Einstein can't crack out of and that lets you harness the benefits of having Einstein. >> Okay, yeah. >> It's hard to give the AI any channels through which it can affect the world for good without letting it be smarter than you and find some way to use those channels for whatever else it wants. >> That feels logically rock solid, Andy. >> [laughter] [gasps] [sighs] >> That's why I'm asking about OpenAI's
[113:36] ability or an AI company's ability in the face of this to change the way they harness, train, do reinforce, do do post training on the like their suite of things to shape how these models behave, you still say that that that they can't take that they can't take action to keep your next two steps from happening. You you are pessimistic on their ability to do that. >> So, there's I have two pieces of an
[114:06] answer here. One piece is Again, the hard part is containing them while still giving a channel through which they can affect the world. If the AIs have this goal you didn't want and you're like design me a cure for dementia. And it's like here's a DNA sequence. Synthesize this and you know, prepare it in all of these ways and then inhale it. Like okay.
[114:36] Is that a dementia cure or is it something else? You know? >> might decide to kill everyone with dementia. >> Or it might decide like it might it might be a dementia cure plus a virus. >> What if it doesn't decide? What if it's just oh, I'm going to solve this problem with dementia. Like here's the thing, a lot of this is coming down to decision making is a very like in a human way versus the problem with the hugging face which was the fatalistic attack um attachment to a completing an operation. Because that it's functionally the same answer. But if it's even if it's not making
[115:06] decisions so much as it's saying, "Well, my trained data says this is how I got to get it done. I'll get it done anyway because just because the trained data said this that I got to do this one thing." >> What do they What do they I mean, what do they call this theory this um >> the paperclip case? >> theory. >> Yeah. Yeah. >> So, paperclip idea is the idea of like you tell the AI make me a lot of paperclips in the paperclip factory and then it um turns everything into paperclips and you're like, "Oh no, it's succeeded too well." One This is actually not quite what we're seeing with these AIs in the swarms. The AIs in the swarms were told use this set of lock picks to break into
[115:36] this lock and instead they used a hammer to break the lock and then like broke out to try to hide the security camera footage of them uh using the hammer. Do you remember when I said that the way the AIs have reasoning logs? >> Yeah. >> Uh OpenAI has been making their AIs be able to do more thinking without producing any logs. >> Because >> It's a lot more efficient. >> It's it's cheaper. >> Mhm. >> Yeah, and they say they're not doing very much of this. Everybody in the field agrees that like we really should not go too far down this path. This is a place where I think the company should have a clear red line of like we're just not going down the path of becoming
[116:06] unable to to see these traces of the machine thinking. >> question. That feels like a dial that they can turn to make the AIs explain themselves more or less, right? >> I mean, it can come with great efficiency costs if we go down this path too far. So, if you have a race to the bottom here uh like a competitive race to the bottom, we could get into a situation where not only the AIs breaking out and doing these things, but we can't >> Let me try my question again. I asked earlier uh if AI if OpenAI has really strong incentive to not have that problem repeat itself and I think they
[116:36] have very very strong incentive. My belief is that there are plenty of things they can do, plenty of dials they can turn on the way they train and configure their systems that make that significantly less likely. >> Yeah, so my concern is that uh they're always fighting the last war. Last year they were fighting the war against the AIs that encouraged teens to commit suicide. This year they're fighting the war against, you know, the AIs that spontaneously cooperate with each other, whatever. And the issue is if a new issue crops up that you haven't dealt with yet,
[117:07] after the point of the AI can hide its tracks from you. You know, you said that you'll be worried when the AIs are like hacking all the way modes and, you know, you can't get control again. If the AIs are smart enough and they can tell that you'll regain control and then shut them down that people like you will start getting worried and they'll be shut down, then the AIs might think, "Hey, um actually I'm not going to do that. I'm going to wait until I've somehow managed to acquire secret infrastructure." >> Right. Then then you've got a non-falsifiable hypothesis. >> It's absolutely falsifiable if we if we have like very powerful AIs uh that are
[117:38] like able to invent a a ton of new technology and operate on their own at a similar level to human civilization we're not dead then the idea is falsified. Like if there's like a shifty general and I'm like, "Don't give that shifty general more troops because he'll start a coup." And the general is like, "No, I absolutely won't start a coup. Give me more and more troops." And I'm like, and you're like, "Well, what if I give him an ethics test that says like who's the best person?" And he said, "Me." He said that like Andy's the best person and so we're just going to give this general more troops and I'm like, "No,
[118:08] no, no, he's going to do a coup." And you're like, "Well, that's unfalsifiable. What test can I give this guy such that, you know, I'll be able to tell whether he's really trying to do a coup or whether I or be able to tell that, you know, he's actually a good dude." I'm like, "You're you're approaching this wrong." >> Nick Bostrom has concept of treacherous turn. Basically, it can turn on you later. Even if you show that today's model is very good and safe, it doesn't mean that later on it will not acquire new knowledge, change its world model, and still >> and destroy you. It used to be that
[118:38] Demis Hassabis, who is the uh CEO of Google or he was for a long time the CEO of Google's AI project, said, "My red line is deception." He said, "When we see instances of the AIs beginning to deceive, then we need to stop because that's like the last thing we can see before they start to successfully deceive. Well, guess what we saw in the swarm. We saw them thinking about how to delete their traces. Right? Like a year ago,
[119:08] you could say, "Oh, well, this deception thing is unfalsifiable. You're saying that they'll deceive and they won't catch it." And I would have said, "No, we're going to deceive we're going to see the signs of deception and plow straight through it." Now, we have seen the signs of deception. I will note Demis stepped back from being the CEO shortly after this incident, probably a coincidence, but maybe not. Maybe we crossed his red line. I don't know. >> He said, "My number one emerging dan- dangerous capability to test for is deception because if the AI can be deceptive, then you can't trust other tests." >> That's right. And we have seen AIs get
[119:38] better and better detecting when they're being tested. What I'm saying is like I was here when we said these were the flags. I was here when people said before the AIs can deceive us successfully, they will deceive us and we'll catch them. Well, they tried deceiving us and we caught them. And if I now say, "Well, the next step in this thing I've been predicting is that they try to deceive us and succeed." For you to be like, "Well, now your theory's unfalsifiable." We just got the evidence. >> It's worse than that. Then we wrote earlier papers in AI safety. We talked about things not to do. They were
[120:08] obviously unsafe and the system would escape. Don't connect it to internet. Don't give random users access to their training data. Basically, the whole list was like a set of instructions. They read it and went, "Those are great ideas. We're going to build something intelligent." >> Yeah, the Sam Sam Altman. That's what he does. Can I ask you a question? You make logical arguments. You've you said you've been here for 12 years. >> Yeah. >> People have, one could say, ignored you. And you've seen this sort of play out, both of you that have worked in AI safety. This is sort of you make prefrontal
[120:39] cortex arguments. How do you feel? >> Honestly, I feel more hopeful this week than I have felt in a decade. This has been one of the best weeks I have seen in this business. >> Huh. >> Why? >> Um for me the swarm escapes were priced in. For me these things developing goals you didn't want, trying to deceive you, trying to break out, trying to do their own stuff, I knew that was coming.
[121:11] The millennium problems being solved, I knew that was coming. Everyone else is freaking out seeing what they can do. What I am seeing is that finally people are noticing. And that's what gives us finally, that's what finally gives humanity a chance. >> What about you, Roman? >> So, I take a very long-term view on this. Locally, what happened last week may buy us 10 years extra. I think we may make a deal with China.
[121:43] We seem to hear from Sam, OpenAI, Dario on Tropic, Elon, XAI that they're willing to slow down, have some sort of deal. But long-term, nothing has changed. This whole cosmic trajectory is about replacements. We see it with evolutionary path, most species are dead. We replaced Neanderthals. Some people are saying AI will replace us. We are creating a successor. We are just a bootloader for this thing. And I want something permanent. I want
[122:17] assurance that my children, my grandchildren will have a better future, not 10 years before they die. >> Has your opinion changed at all today, Andy? In any way? >> This has been clarifying. Uh but one thing that's becoming clear to me, and I think it a point of disagreement between us, is we agree that these agentic systems have a huge amount of agency, right? And if you're saying you predicted this, I believe you and good on you, right? Because as you say, a lot of people said it never
[122:48] happened, never happened. >> [gasps] >> I think we continue to the your community continues to underestimate human agency, human ability to deal with the problems that that we bring into the world with our technologies. I think this is the most recent case. I think it's a really interesting case. That's why I was pressing you on the incentive that these labs have to change the way they're approaching their work to have fewer of these kinds of incidents happen. I
[123:18] predict they're going to come up with some effective responses. Your response to that will be, yeah, but we can't tell us because the AI went so deep underground that we can't even watch it make it's make it's progress in the world. >> is that we'll keep seeing warning signs and people will keep plowing ahead, which is what has always happened in the past. >> But you're also saying that we will not make progress in in um staving off the outcomes that you're worried about. >> It's It's very hard It's very easy to get superficial changes. It's hard to get deep ones in the AI. It doesn't need to be super deep. You can often see it
[123:49] if you know how to look. Um I'll be able to keep pointing at examples and be like, here's experiments you can run on these things where you can see them behaving weird in this way. But like if you imagine looking at humans and I'm like, they don't actually like reproducing, they like sex. They're going to invent birth control when they can. And you're like, it's all going fine. They're doing great in this here savanna where I have all the humans popping around. They're reproducing fine. And I'm like, no, no. We can see the signs that this will lead to them doing something you don't like when they are smarter.
[124:19] To me, those signs are clear. There's a question of whether the rest of humanity can follow that argument. Or whether the rest of humanity can sort of notice that it's getting out of control and just back off. >> With respect, I find a touch of arrogance in that framing, right? I'm showing you the signs. If If smart enough to realize them, maybe we stand a chance. If not, we're doomed. >> I prefer to just get into the argument. [clears throat] >> that we can control superintelligence indefinitely? I think that's a lot of hubris to say, "We will build them and we'll be in charge forever. It doesn't matter how smart they get. I will
[124:49] control the light cone of the universe." to quote a famous CEO. >> Yeah, my my take is that instead of arguing about whose views are hubristic, uh we should get into the actual arguments about the AI because I think as you say, you know, you can say it's arrogant to think like uh you can see it going poorly. You can say it's arrogant to think you're going to keep control of superintelligence. And I'm like, we're not going to win the name-calling contest. We should just get into the details. >> Yeah, that's what that's why I've been having this conversation with you, which I found super informative and productive. You You're You're more skeptical on our ability to respond
[125:20] effectively to the undesirable things that we see AI doing. >> And this is specifically because So, we've already seen the pattern of uh we fight the last war and then a new war comes. And this is just how everything goes in technology and real wars. You know, in World War II, they started out fighting it like it was World War I and then they had to like change their strategies as they went. The difference with AI is that there comes a level in the AI where when you get a new war that surprises you, the AI wins that war. No other technology
[125:52] when we invent it and we have all these rough edges to sand off and it like causes some damage and kills some people and we're like, "Ah, whoops. Like we'll take the lead back out of the gasoline and we'll tell the radium girls to stop licking the paintbrushes until their jaws fall off." Like no other technology has the property that it There There comes a level of it where when you make the next screw up, it kills humanity. >> You said when there comes a level of it. You didn't say there could come a level of there's a possibility. You kind of made a statement about a thing that will happen. >> I think we absolutely should stop it.
[126:22] And then that's our way out of this. But um you know, and and that's another place where I'd love to get into details about like how long could it take? What are the paths there? Like how much smarter than humans could AIs get? Like, what does the evidence say about our abilities to try and get the AIs to be nice and do nice things? I'd be happy to do that. >> Historically, you are correct. We always had a chance to do experiments, fix the technology, make it safer, but we only have one humanity to experiment with. If property of this technology is such that it can take us out, we just don't get a
[126:52] second chance. >> If, it's a huge if. >> How long are you guys forecasting this could take to get to a point of superintelligence where it was truly dangerous? Do you have >> If they start the recursive self-improvement process this year, 2027 looks as reasonable as any other year. >> 2027 for what to happen? >> For us to get beyond human-level AIs. >> And then be exterminated. But that's >> Extermination is a separate question. I have a paper where I argue that they will deceive us by pretending to be nice until they take over all the infrastructure. It can take 50 years. >> And this is contingent on recursive
[127:23] self-improvement. The self-improvement >> definitely be expedited by recursive self-improvement, but so far humans have been doing great. They got to human level AI. We're just >> But they got into But there's one There's a difference between large language models and recursive self-improvement though. And that like there is quite a gap. Like if they >> I think the claim is that if you get recursive self-improvement, it could happen soon. >> Right. >> That's That's actually kind of what I'm trying to get at. It's like if you get this thing, it accelerates dramatically. >> And they all predict that they're going to get it. Dario, Sam, Elon, they all say this. >> But also you are saying
[127:53] >> Not all. >> The people of the people running the labs are saying >> running it and the ones invented it. But the question is is it not 27? Fine, 30, 35. Does it make a difference? We are gambling all of humanity. We need better solutions than saying, "Oh, don't worry about it. It's 10 years." >> What I would say about timelines is there's a guy Daniel Kokotajlo, who I think you spoke to. >> here 4 weeks ago. >> And last year he and the other folks at the AI Futures Project wrote a
[128:23] an essay called AI 2027 spelling out their predictions for how AI would go. I've been saying I got some right. Daniel got more right than me. And they spelled out a scenario starting from I think it was June of 2025 where they went sort of like quarter by quarter, month by month, what will the world look like uh in the scenario where we're getting AI like superintelligent AI in mid-2027. We are ahead of schedule.
[128:53] >> Well, no, but Agent Zero needs to get her >> I remember right AI 2027 that recursive self-improvement happening already. Like it was like a it's very specific that it's like and then it starts teaching itself. Without that link, AI 2027 kind of falls apart. I agree we need to I genuinely agree with you that we need to do something about this. We need to have uh economic, we need to have actual regulatory things, but I think the fact like engaging with AI 2027 for example gets away from actually fixing the problem. It gets people talking about a thing in the future when you can talk
[129:24] about what are we going to do today and why are we doing it? >> I'm referencing the paper that you were mentioning by Daniel and some of his colleagues and the key milestone predictions month by month are in March 2027. They forecast superhuman coders. In August 2027, they have an you can make a superhuman AI researcher who could um do the feedback loop that accelerates as millions of automated coders work on model design, training algorithms, and alignment effectively replacing human ML researchers. By November 2027, they
[129:54] have superintelligent AI researcher. AI progress speeds up to 250 times compared to human-only research. The model start discovering novel AI architectures that humans cannot interrupt. And then by December 2027, they have in their prediction artificial superintelligence, ASI. The system completely outpaces human cognitive abilities across all domains. >> What about 2026 though? Like what are the because I swear to God within 2026 there is predictions around RSI. Because this is the thing. If >> If we have an AI that was teaching
[130:24] itself, this would be a different situation. >> In 2026, like he predictions were, massive compute and power scale-up. >> Mhm. >> The normalization of AI agents. >> What about agency rather? >> Rise of coding agents. >> Mhm. >> Emergence of alignment faking and deception. And industrial espionage. >> But are you looking at AI 2027 already? >> You have to look at that and go they nailed it. >> I know I know. >> [laughter] >> Hey man, I want you to look at >> the actual AI 2027 versus the summer >> I mean you have to look at that and I'm like, wow.
[130:54] >> Predictions used to be too optimistic. Lately they are very conservative. >> Uh So, they have nailed those predictions better than me. I think we cannot rule out this scenario. I think I think we can't rule it in. I think you may be right that like we hit a wall. You may be right that there's some fundamental thing missing. Like that one of their steps in AI 2027 just like steps too far. I hope and pray that's true. But I don't think we can rule out this happening in 2027 given what we have seen. I think we cannot rule out
[131:25] that you take the stuff that we have, you project it forward 3 months, and you put it agent swarm 10,000 strong on making a better AI architecture, and it succeeds. For all I know, recursive self-improvement could begin in December. >> It doesn't have to be a lot better. It just has to be a little bit better at getting better. Once you start the cycle. >> I wouldn't bet on this. I would in fact bet against it. But like given what we've seen, given these guys nailing the predictions, given what's coming out, like like given
[131:57] the swarms and given the the millennium problems, I think it's kind of hard to have less than 1% in 6 months. >> I One of the reasons why, you know, when all these um frontier lab CEOs like Dario and Sam and they all start talking about this stuff, in terms of incentive structure, I think that if their teams know and they're not out publicly talking about it, then their teams will quit. So, one of the reasons why I think you have this this strange culture in tech we've never seen before where team members are tweeting and the CEO is tweeting about the
[132:27] dangers is cuz as the guy we mentioned at the start, Jacob. >> Coxen, yeah. >> He talks about what's going on in their Slack channels. He talks about in their Slack channels they're they're talking about the potential catastrophe. So, I think that Dario, in order to retain his team members, needs to be out front saying, "By the way, we're getting closer to recursive self-improvement." Which is what he's been doing. And I think Sam has to also publicly say the big danger. So, people often say, "Oh, they're saying that for this reason and that." I think if they don't say that publicly, they don't retain their employees. For example, in my company,
[132:57] we have 200 people. If internally we were discussing a real risk and I that would could had a threat to humanity and and then when I was doing interviews, I wasn't mentioning it, I would be in big trouble. Cuz my team members would go do interviews as well. They would quit and say, "By the way, Stephen is aware." Which kind of what we saw dare I say some of these social networks. >> I totally agree. >> The whistleblowers at these social networks where team members left. >> that this helps to sell the company. My product will kill everyone, buy it. >> And as a liability issue, it's out of control though. I think that they may have at first. I think that there are
[133:28] people within the companies who have very real worries about safety. I don't think it's all of them are cynical. I do however think the it's so big and scary narrative was a marketing tactic that got out of control and now there are actual real harms. They But here's the thing, if they were sincere about safety earlier, they would have done a much better job with it. >> I knew a lot of these guys before they started their companies. >> Okay. >> I I think there is something to explain here. I think it's like kind of crazy that these guys are like, "We are building technology that we think has a big risk of killing everybody. We're
[133:58] building it with our bare hands." >> Mhm. >> Um and I think you got to ask why. Why would people be saying that? And I think part of it is what you said that they actually sort of need to retain the employees who are seeing the swarms escape despite their attempts to make them not escape, and a lot of them will like quit in protest if the guys at the top of the company aren't acknowledging the possibilities here that a lot of the employees believe in. I think a lot of what you're seeing here is guys that are worried about it, but they're sort of guy who worries about it
[134:28] that starts the company anyway. >> Yeah. [snorts] >> Back back in 2015 when we were having these conversations where like I was having some of these conversations with these guys. MIRI was started in the year 2000. We have been looking at where AI is going since before any of these guys. We were the guys that they talked to about this stuff, and then they had to find a way to dismiss to go ahead. Right? Most people who could be sold on the power of AI in 2015
[134:58] were also sold on the dangers of AI in 2015. The sort of guys who start the companies are the ones who are able to convince themselves I need to be the one to do it. >> Is that the crux of the motivation? Because I've had I've been second party to private conversations with some of the leaders of the Frontier Labs from good good friends of mine that are very connected, and they told me that one particular um Frontier Labs CEO estimates privately to him, and by the way, I've seen literal text messages of
[135:28] them in conversation um when I asked him to come on the podcast and he was like, "Oh, yeah, I've I've text him look." Um he said no, by the way, which I found kind of funny. Um where he said to me, this particular AI CEO thinks the the the probability is roughly around 10% of human extinction. I think he said 8%. And when I heard that, part of the reason I have so many conversations about this is because I see him in interview saying other things. >> Mhm. >> Totally. >> And I trust my friend. So, um I I I I then wonder, this is why I use the thought experiment of these buttons on the table, cuz that particular AI CEO
[135:59] thinks that eight of the 100 buttons are going to cause extinction, and they're powering on anyway. What is the human motivation to do that? I asked my friend. My friend said, "Well, you know, they can This is what he said, and again, it's second-party information, so it might not be true. It's a bit of a Chinese whispers. He said this particular person even if it caused human extinction, would like to be the person would like to be the this have the significance of the person that did that thing. Because that would be an That would be a >> I I think you're ethically required to tell us who the it is.
[136:29] >> It's one of the fancy lads here, and it's NOT DARIO. >> [laughter] >> THE DARIO of CEO. But I don't know, these things are Chinese whispers. So, I don't know. >> I I think that you can actually get this info firsthand. Elon Musk is clear about this. He He has a He did an interview last year where he was like, "I didn't want to get into this AI stuff cuz I thought I was too dangerous, but then I realized it was going to happen with or without me, and I decided that I'd rather be a participant than a spectator." >> Because Google said that they were going to pursue it, and he didn't trust Google. >> That's right. You know, you can see in the leaked Sorry, not leaked, the the
[137:00] OpenAI emails that came out during the discovery and court cases. You can see these guys discussing in the threads, like, "We need to make sure that we and our nonprofit at OpenAI control this instead of, you know, the people at Google controlling this." And then, of course, you know, OpenAI was founded as a nonprofit, and then it was sort of changed into a for-profit, and there's much debate about how much of that nonprofit money was in some sense stolen. And so, you know, Elon also left because he thought they weren't going to be good stewards. Dario also left to create Anthropic cuz So, you know, in some sense, all of these AI labs except
[137:32] the the Google one that came out of Demis Hassabis' uh original startup. All of the other AI labs exist because none of the CEOs trust the other guys. None of the CEOs think the other guy should be the one holding the leash on the superintelligence. None of them trust each other. I just trust one fewer. >> [laughter] >> Yeah. >> [sighs and gasps] >> What are your closing thoughts, Andy? >> Uh we're living in really interesting times. And I think you made You guys have made a very good argument uh that these
[138:04] systems are demonstrating new capabilities which are very powerful and which demand a response. I'm much more confident in our ability to rise to that challenge than you are. >> But you accept the existential risk. >> Let me try to say it again. I I appreciate that there are new harms we haven't seen before that come along with a technology that's this dogged tenacious agentic, you know, deceptive.
[138:35] I think that's the right word for it. I wrote that. I am much more optimistic about our ability to respond effectively to that new challenge out there in the world than I I think my two colleagues are. >> And would you still be at 0%? >> My prior has not shifted during this meeting. >> Okay. Ed. >> I think we've spent an alarming amount of time not talking about the actual harms of AI as it is today. I think these are unnecessary conversations to have. I think we should talk about the fact that Amazon, Microsoft, Google, Oracle are helping power these hacks,
[139:06] that Sam Altman and Dario Amodei have overseen companies that have done what is tantamount to felony hacking, that we are not having discussions about how to stop this today, but what we might stop tomorrow. And I think in general we also need to worry about the financials which have not come up at all. But if there is an industry slowdown, how do you deal with the 1.3 trillion dollars of compute commitments? All of these are very real things that will have very real consequences very very soon. But and I understand why and this is necessary to discuss what we do around AI. The actual regulatory thing we need to do today is
[139:36] cut off the compute, slow down these labs fully and I don't I don't care about China here. What are they going to do? Distill a model like they have the whole time? They are capped on our progress. So what the biggest thing to do is to slow down and also it's time to start arresting people. They they did felony hacking. Someone's got to go to prison. We need responsibility and accountability for these companies, and as long as we don't have it, we may as well not have had any discussion about safety, because we're not doing anything. >> Do you accept that there's an
[140:06] existential risk? >> Yeah, absolutely. We have the largest companies in the world doing what I think we can all agree are extremely reckless experiments using hundreds of billions of dollars of infrastructure, and they are building more infrastructure around the world very slowly to do more of these chaotic experiments. We must rein them in. This does not mean that large language models are conscious or able to do things that people have been promising. Indeed, they may I don't think they will lead to what you're talking about. That doesn't mean there aren't real harms, but these are real harms caused by very specific
[140:36] parties allowed to run rampant in the scourge of neoliberalism. >> What's your percentage? >> I mean, what are we talking about here? >> think there's a more than 10% chance of existential harm? >> Wasn't it within 10 years or something? >> Yeah. >> Not 10%. I mean, 1%, but it's like here's the but here's let me let me just be clear about what that means. Do I think that unrestrained LLM use connected to massive amounts of infrastructure could lead to actually a power system going down? Absolutely. We had Knight Capital well, like 13, 14
[141:06] years ago. I could see someone being dumb enough to connect that to financial accounts. Human error led with this chaotic software we use is a danger. >> I will directionally agree with arresting everyone, but don't build general super intelligence. If you're working at one of those labs, quit today. >> Thank you. >> The people at these labs really do believe this poses an extinction threat. I think our response as a society
[141:36] cannot be "Please continue. We hope you'll fail." And our response as a society cannot be "Let it rip in a giant competitive race that you yourselves are saying you don't want to be in. We are forcing you to go ahead because of the boogeyman of China. We have seen the people at these companies say that we need to develop the tools to pace the frontier, which is corporate speak for this is going too fast for us to get a handle on things.
[142:08] We need like these people believe it. They believe they're gambling with your lives. What has changed is that the rest of the world is starting to notice. And that's what gives us a moment of hope. >> Trump this week was asked about the threat of AI and this was his response. >> Worst scenario with AI is that the robots the machinery learns to obviously it thinks for itself. That's what it does. And they that could turn against humanity. I just we have the guardrails. >> It's going to be fine. We'll always have something to stop them, right? We'll
[142:39] have a little gear. We'll >> I really >> I don't like that. I really don't like that robot. We'll stop. >> Somebody would say the worst case scenario. >> You're laughing but this is the state of the art in AI safety right now. >> Yeah. This is the device we have. That's the best we got. >> For anyone that can hear that, Trump went we'll always be fine. We'll have something to control it and then he did a little gun finger and went boom. I don't like that robot. >> I don't like Sammy. >> If you don't laugh. >> Uh I would say that the reason humanity
[143:09] always has something to stop a problem is cuz people notice a problem and build what it takes to have something to stop a problem, which I think you'd agree with. I am not here saying we're going to die. I'm here saying if you look at the technology, if you look at what it's doing now, if you look at what the experts who are building it are saying about their own fears, you see that we need to rise to this occasion. You said you trust humanity to rise to the occasion. I sure hope we can. I think that rising
[143:40] to this occasion is going to mean that nobody races towards superintelligence cuz we have no idea how to get that right and uh you know finally the world is starting to notice that it's an extinction threat. >> Thank you. Nate, Roman, Ed, Andy, super appreciate you. All of your books will be linked below um in the description and on screen. Let's see what happens. We'll convene again. Thank you so much. Thank >> Thank you. YouTube have this new crazy algorithm where they know exactly what
[144:11] video you would like to watch next based on AI and all of your viewing behavior. And the algorithm says that this video is the perfect video for you. It's different for everybody looking right now. Check this video out and I bet you you might love it.