#434 — Can We Survive AI?

0:28Welcome to the Making Sense Podcast. what we're doing here, please consider becoming one.

0:36I am here with Eliezer Yudkowsky and Nate Soares. Eliezer, Nate, it's great to see you guys again. It's been a while. Good to see you, Sam. Been a long time. So you were, Eliezer, you were among the first people to make me concerned about AI, which is going to be the topic of today's conversation. I think many people who are concerned about AI can say that. First, I should say you guys are releasing a book, which will be available, I'm sure, the moment this drops. If Anyone Builds It, Everyone Dies, Why Superhuman AI Would Kill Us All. I mean, the book, its message is fully condensed in that title.

1:15I mean, we're going to explore just how uncompromising a thesis that is and how worried you are and how you're worried you think we all should be here. But before we jump into the issue, maybe tell the audience how each of you got into this topic. How is it that you came to be so concerned about the prospect of developing superhuman AI? Well, in my case, I guess I was sort of raised in a house with enough science books and enough science fiction books that thoughts like these were always in the background. Werner Vingy is the one where there was a key click moment of observation, Vingy pointed out that at the point where our models of the future predict building anything smarter than us, then said Vingy at the time, our crystal ball explodes past that point.

2:06It is very hard, said Vingy, to project what happens if there's things running around that are smarter than you, which in some senses, you can see it as a sort of central thesis, not in the sense that I have believed it the entire time, but then in the sense that some parts that I believe and some parts that I react against and say like, no, maybe we can say the following thing under the following circumstances. Initially, I was young. I made some metaphysical errors of the sort that young people do. I thought that if you built something very smart, it would automatically be nice because, hey, over the course of human history, we had gotten a bit smarter.

2:41We'd gotten a bit more powerful. We'd gotten a bit nicer. I thought these things were intrinsically tied together and correlated it in a very solid and reliable way. I grew up, I read more books, I realized that was mistaken. And 2001 is where the first tiny fringe of concern touched my mind. It was clearly a very important issue, even if I thought there was just a little tiny remote chance that maybe something would go wrong. So I studied harder, I looked into it more, I asked how would I solve this problem? Okay, what would go wrong with that solution? And around 2003 is the point at which I realized like, this is actually a big deal.

3:17Nate? And as for my part, yeah, I was 13 in 2003, so I didn't get into this quite as early as Eliezer. But in 2013, I read some arguments by this guy called Eliezer Yudkowsky, who sort of laid out the reasons why AI was going to be a big deal and why we had some work to do to do the job right. And I was persuaded. And one thing led to another. And next thing you knew, I was running the Machine Intelligence Research Institute, which Eliezer co-founded. And then, you know, fast forward 10 years after that, here I am writing a book. Yeah, so you mentioned Miri. Maybe tell people what the mandate of that organization is and maybe how it's changed.

4:03I think you indicated in your book that your priorities have shifted as we cross the final yards into the end zone of some AI apocalypse. Yeah, so the mission of the org is to ensure that the development of machine intelligence is beneficial. And, you know, Eliezer can speak to more of the history than me because he co-founded it and I joined, you know. Well, initially, it seemed like the best way to do that was to run out there and solve alignment. And there was a series of, shall we say, like sad, series of bits of sad news about how possible that was going to be, how much progress was being made in that field relative to the field of AI capabilities.

4:48And at some point, it became clear that these lines were not going to cross. and then we shifted to taking the knowledge that we'd accumulated over the course of trying to solve alignment and trying to tell the world, this is not solved. This is not on track to be solved in time. It is not realistic that small changes to the world can get us to where this will be solved on time. Maybe so we don't lose anyone. I would think 90 % of the audience knows what the phrase solve alignment means, but just talk about the alignment problem briefly. So the alignment problem is how to make an AI a very powerful AI.

5:24Well, the superintelligence alignment problem is how to make a very powerful AI that steers the world sort of where the programmers, builders, growers, creators wanted the world, wanted the AI to steer the world. It's not necessarily what the programmers selfishly want. The programmers can have wanted the AI to steer it in nice places. But if you can make an AI that is trying to do things that the program, When you build a chess machine, you define what counts as a winning state of the board. And then the chess machine goes off and it steers the chessboard into that part of reality. So the ability to say, to what part of reality does an AI steer is alignment.

6:05On the smaller scale today, though it's a rather different topic, it's about getting an AI whose output and behavior is something like what the programmers had in mind. If your AI is talking people into committing suicide and that's not what the programmers wanted, that's a failure of alignment. If an AI is talking people into suicide and people who should not have committed suicide, but AI talks them into it, and the programmers didn't want that, that's what they tried to do on purpose. This may be a failure of niceness. It may be a failure of beneficialness, but it's a success of alignment. The programmers got the AI to do what they wanted it to do.

6:42Right. But I think more generally, correct me if I'm wrong, when we talk about the alignment problem, we're talking about the problem of keeping superintelligent machines aligned with our interests, even as we explore the space of all possible interests and as our interests evolve. So that, I mean, the dream is to build superintelligence that is always corrigible, that is always trying to best approximate what is going to increase human flourishing. That is never going to form any interests of its own that are incompatible with our well-being. Is that a fair summary? I mean, there's three different goals you could be trying to pursue on a technical level here.

7:25There's the superintelligence that shuts up, does what you ordered, has that play out the way you expected it. no side effects you didn't expect. There's superintelligence that is trying to run the whole galaxy according to nice benevolent principles, and everybody lives happily ever afterward, but not necessarily because any particular humans are in charge of that. They're still giving it orders. And third, there's superintelligence that is itself having fun and cares about other superintelligences and is a nice person and leads a life well-lived and is a good citizen of the galaxy. And these are three different goals.

8:02They're all important goals, but you don't necessarily want to pursue all three of them at the same time, and especially not when you're just starting out. Yeah. And depending on what's entailed by superintelligent fun, I'm not so sure I would sign up for the third possibility. I mean, I would say that, you know, the problem of like, what exactly is fun and how do you keep humans, like, how do you have whatever the superintelligence tries to do that's fun, you know keep in touch with moral progress and have flexibility and like what even what do you point it towards that could be a good outcome all of that those are problems i would love to have those are you know right right now just you know creating an ai that does what the operators intended creating an ai that like you've pointed in some direction at all rather than pointed off into some like weird squirrely direction that's kind of vaguely like where you tried to point it in the training environment and then really diverges after the training environment.

8:56We're not in a world where we sort of get to bicker about where exactly to point the superintelligence, and maybe some of them aren't quite good. We're in a world where no one is anywhere near close to pointing these things in the slightest in a way that'll be robust to an AI maturing into a superintelligence. Right. Okay. So, Eliezer, I think I derailed you. You were going to say how the mandate or mission of MIRI has changed in recent years. I asked you to define alignment. Yeah, so originally, well, our mandate has always been make sure everything goes well for the galaxy. And originally, we pursued that mandate by trying to go off and solve alignment because nobody else was trying to do that.

9:35Solve the technical problems that would be associated with any of these three classes of long-term goal. And progress was not made on that, neither by ourselves nor by others. Some people went around claiming to have made great progress. We think they're very mistaken, and knowably so. And at some point, it was like, okay, we're not going to make it in time. AI is going too fast. Alignment is going too slow. Now it is time for the people that, you know, all we can do with the knowledge that we have accumulated here is try to warn the world that we are on course for a drastic failure and crash here, where by that I mean everybody dying.

10:13Okay, so before we jump into the problem, which is deep and perplexing, and we're going spend a lot of time trying to diagnose why people's intuitions are so bad, or at least seem so bad from your point of view around this. But before we get there, let's talk about the current progress such as it is in AI. What has surprised you guys over the last, I don't know, decade or seven or so years? What has happened that you were expecting or weren't expecting? I mean, I can tell you what has surprised me, but I'd love to hear just how this has unfolded in ways that you didn't expect? I mean, one surprise that led to the book was, you know, there was the chat GPT moment where a lot of people, you know, for one thing, LLMs were created and they sort of do a qualitatively more general range of tasks than previous AIs at a qualitatively higher skill level than previous AIs.

11:08And, you know, chat GPT was, I think, the fastest growing consumer app of all time. The way that this impinged upon my actions was, you know, I'd spent a long time talking to people in Silicon Valley about the issues here and would get lots of different types of pushback. You know, there's a saying, it's hard to convince a man of a thing when a seller depends on not believing it. And then after the chat to UPT moment, a lot more people wanted to talk about this issue, including policymakers, you know, people around the world. suddenly AI was on their radar in a way it wasn't before. And one thing that surprised me is how much easier it was to have this conversation with people outside of the field who didn't have a salary depending on not believing the arguments.

11:55I would go to meetings with policymakers where I'd have a ton of argumentation prepared and I'd sort of lay out the very simple case of like, hey, people are trying to build machines that are smarter than us. The chatbots are a stepping stone towards superintelligence. Superintelligence would radically transform the world because intelligence is this power that, you know, let humans radically change the world. And if we manage to automate it and it goes 10 ,000 times as fast and doesn't need to sleep and doesn't need to eat, then, you know, it'll by default go poorly. And then the policymakers would be like, oh yeah, that makes sense.

12:25And it'd be like, what? You know, I have a whole book worth of other arguments about how it makes sense and why all of the various, you know, misconceptions people might have don't actually fly or all of the hopes and dreams don't actually fly. But, you know, outside of the Silicon Valley world is just, it's not that hard an argument to make. A lot of people see it, which surprised me. I mean, maybe that's not the developments per se and the surprises there, but it was a surprise strategically for me. Development-wise, you know, I would not have guessed that we would hang around in AIs that can talk and that can write some code, but that aren't already in the, you know, able to do AI research zone.

13:01I wasn't expecting in my visualizations this to last quite this long, but also, you know, my, my advanced visualizations, you know, one thing we say in the book is, um, the trick to trying to predict the future is to predict the questions that are easy, predict the, the facts that are, that are easy to call. And, you know, exactly how AI goes, that's never been an easy call. That's never been something where I've said, you know, I can, I can guess exactly the path we'll take. The thing I can predict is the end point. The path, I mean, there sure have been some zigs and zags in the pathway. I would say that the thing I've maybe been most surprised by is how well the AI companies managed to nail Hollywood stereotypes that I thought were completely ridiculous, which is sort of a surface take on an underlying technical surprise.

13:49But even as late as 2015, which from my perspective is pretty late in the game. Like if you've been like, so Eliezer, what's the chance that in the future we're going to have computer security that will yield to Captain Kirk style gaslighting using confusing English sentences that get the computer to do what you want. And I would have been like, this is, you know, a trope that exists for obvious Hollywood reasons. You know, you can see why the script writers think this is plausible, but why would real life ever go like that? And then real life went like that. And the sort of underlying technical surprise there is the reversal of what used to be called Moravec's paradox.

14:31For several decades in artificial intelligence, Moravec's paradox was that things which are easy for humans are hard for computers. Things which are hard for humans are easy for computers. For a human, you know, multiplying two 20-digit numbers in your head. That's a big deal. For a computer, trivial. And similarly, not just me, but I think the sort of conventional wisdom even was that games like chess and Go, problems with very solid factual natures like math and even surrounding math, the more open problems of science, the notion that we were going to get things that... So the current AIs are good at stuff that five-year-olds can do and 12-year-olds can do.

15:18They can talk in English. They can compose, you know, kind of bull crap essays such as high school teachers will demand of you. But they're not all that good at math and science just yet. They can, you know, solve some classes of math problems, but they're not doing original brilliant math research. And I think not just I, but like a pretty large sector of the whole field thought that it was going to be easier to tackle the math and science stuff and harder to tackle the English essays carry on a conversation stuff. That was the way things had gone up in AI until that point. And we were proud of ourselves for knowing how contrary to average people's intuitions, like really, it's much harder to write a crap essay in high school in English that really understands, you know, that even keeps rough track of what's going on in the topic and so on compared to, you know, how that's really in some sense much more difficult than doing original math research.

16:11Yeah, or counting the number of R's in a word like strawberry, right? They make errors that are counterintuitive. If you can write a coherent essay but can't count letters, I don't think they're making that error any longer. Yeah, I mean, that one goes back to a technical way in which they don't really see the letters. But I mean, there's plenty of other embarrassing mistakes. You can tell a version of the joke with the joke of a child and their dad are in a car crash, and they go to see the doctor, and the doctor says, I can't operate as my child. what's going on, where it's like a riddle where the answer is like, well, the doctor's his mom.

16:48You can tell a version of that that doesn't have the inversion where you know. Where you're like the kid and his mom are in a car crash and they go to the hospital and the doctor says, I can't operate on this child. He's my son. And the AI is like, well, yeah, the surgeon is his mom. He just like said that the mom was in the car crash. But there's some sense in which the rails have been established hard enough that the standard answer gets spit back out. And it sure is interesting that they're, you know, getting an IMO gold medal, like International Math Olympiad gold medal, while also still sometimes falling down on these sorts of things.

17:23It's definitely an interesting skill distribution. You can fool humans the same way a lot of the time. Like there's all kinds of repeatable errors, humorous errors that humans make. You got to put yourselves in the shoes of the AI and imagine what sort of paper would the AI write about humans failing to solve problems that are easy for an AI. So I'll tell you what surprised me just from the safety point of view, Eliezer. You spent a lot of time cooking up thought experiments around what it's going to be like for anyone, any lab designing the most powerful AI to decide whether or not to let it out into the wild.

17:58You imagine this genie in a box or an oracle in a box, and you're talking to it, and you're trying to determine whether or not it's safe, whether it's lying to you, and you famously positive that you couldn't even talk to it, really, because it would be a master of manipulation. And I mean, it's going to be able to find a way through any conversation and be let out into the wild. But this was presupposing that all of these labs would be so alert to the problem of super intelligence getting out that everything would be air gapped from the Internet and nothing would be connected to anything else.

18:32And they would be they would have we would have this moment of decision. It seems like that's not happening. I mean, maybe the most powerful models are locked in a box, but it seems that the moment they get anything plausibly useful, it's out in the wild and millions of people are using it. And, you know, we find out that Grok is a proud Nazi when, you know, after millions of people begin asking questions. I mean, do I have that right? Are you surprised that that framing that you spent so much time on seems to be something that is, it was just in some counterfactual part of the universe that, you know, is not one we're experiencing?

19:13I mean, if you put yourself back in the shoes of little baby Eliezer back in the day, people are telling Eliezer, like, why is superintelligence possibly a threat? We can put it in a fortress on the moon. And, you know, if anything goes wrong, blow up the fortress. So imagine young Eliezer trying to respond to them by saying, actually, in the future, AIs will be trained on boxes that are connected to the internet from the moment, you know, like from the moment they start training. So like the hardware they're on has like a standard line to the internet, even if it's not supposed to be directly accessible to the AI before there's any safety testing because they're still in the process of being trained and who safety tests something while it's still being trained.

20:00So imagine Eliezer trying to say this. What are the people around at the time going to say? Like, no, that's ridiculous. We'll put it in a fortress on the moon. It's cheap for them to say that. For all they know, they're telling the truth. They're not the ones who have to spend the money to build the moon fortress. And from my perspective, there's an argument that still goes through, which is a thing you can see even if you are way too optimistic about the state of society in the future, which is if it's in a fortress in the moon, but it's talking to humans, are the humans secure? Is the human brain secure software?

20:35Is it the case that human beings never come to believe in valid things in any way that's repeatable between different humans? You know, is it the case that humans make no predictable errors for other minds to exploit? And this should have been a winning argument. Of course, they reject it anyways. But the thing to sort of understand about the way this earlier argument played out is that if you tell people the future companies are going to be careless, how does anyone know that for sure? So instead, I tried to make the technical case. Even if the future companies are not careless, this still kills them.

21:09In reality, yes. In reality, the future companies are just careless. Did it surprise you at all that the Turing test turned out not to really be a thing? I mean, we anticipated this moment from Turing's original paper where we would be confronted by the interesting psychological and social moment of not being able to tell whether we're in dialogue with a person or with an AI. And that somehow this landmark technologically would be important, rattling to our sense of our place in the world, et cetera. But it seems to me that if that lasted, it lasted for like five seconds and then it became just obvious that you're, you know, you're talking to an LLM because it's in many respects better than a human could possibly be.

22:01So it's failing the Turing test by passing it so spectacularly. And also it's making these other weird errors that no human would make. But it just seems like the Turing test was never even a thing. Yeah, that happened. I mean, it's just like, it's so, I mean, that was a, one of the, the great pieces of, you intellectual kit we had in framing this discussion for the last, whatever it was, 70 years. And yet, the moment your AI can complete English sentences, it's doing that on some level at a superhuman ability. It's essentially like the calculator in your phone doing superhuman arithmetic, right?

22:42It was never going to do just merely human arithmetic, and so it is with everything else that it's producing. All right, let's talk about the core of your thesis. Maybe you can just state it plainly. What is the problem in building superhuman AI, the intrinsic problem, and why doesn't it matter who builds it, what their intentions are, etc.? In some sense, I mean, you can come at it from various different angles, but in one sense, the issue is modern AIs are grown rather than crafted. It's, you know, people aren't putting in every line of code knowing what it means, like in traditional software.

23:25It's a little bit more like growing an organism. And when you grow an AI, you take some huge amount of computing power, some huge amount of data. People understand the process that shapes the computing power in light of the data, but they don't understand what comes out the end. And what comes out the end is this strange thing that does things no one asked for, that does things no one wanted. You know, we have these cases of, you know, chat GPT, someone will come to it with some somewhat psychotic ideas about, you know, that they think are going to revolutionize physics or whatever, and they're clearly showing some signs of mania.

23:58And, you know, chat GPT, instead of telling them maybe they should get some sleep, if it's in a long conversational context, it'll tell them that, you know, these ideas are revolutionary, and they're the chosen one, and everyone needs to see them and other things that sort of inflame the psychosis. This is despite open AI trying to have it not do that. This is despite, you know, direct instructions in the prompt to stop flattering people so much. These are cases where when people grow an AI, what comes out doesn't do quite what they wanted. It doesn't do quite what they asked for. They're sort of training it to do one thing, and it winds up doing another thing.

24:35They don't get what they trained for. this is in some sense the seed of the issue from one perspective where if you keep on pushing these things to be smarter and smarter and smarter and they don't care about what you wanted them to do they pursue some other weird stuff instead super intelligent pursuit of strange objectives kills us as a side effect not because the ai hates us but because it's transforming the world towards its own alien ends. And, you know, humans don't hate the ants and the other surrounding animals when we build a skyscraper. It's just, we transform the world and other things die as a result.

25:13So that's one angle. You know, we could talk other angles, but... A quick thing I would add to that, just trying to sort of like potentially read the future, although that's hard, is possibly in six months or two years, if we're all still around, people will be boasting about how their large language models are now like apparently doing the right thing when they're being observed and you know like answering the right way on the ethics tests and the thing to remember there is that for example the uh mandarin imperial system in ancient china imperial examination system in ancient china they would see they would give people essay questions about Confucianism and only promote people high in the bureaucracy if they, you know, could write these convincing essays about ethics.

25:59But what this tests for is people who can figure out what the examiners want to hear. It doesn't mean they actually abide by Confucian ethics. So possibly at some point in the future, we may see a point where the AIs have become capable enough to understand what humans want to hear, what humans want to see, this will not be the same as those things being the AI's own true motivations for basically the same reason that the imperial China exam system did not reliably promote ethical good people to run their government. Just being able to answer the right way on the test or even fake behaviors while you're being observed is not the same as the internal motivations lining up.

26:42Okay, so you're talking about things like forming an intention to pass a test in some way that amounts to cheating, right? So you just use the phrase fake behavior. I think a lot of people, I mean, certainly historically this was true. I don't know how much their convictions have changed in the meantime, but many, many people who were not at all concerned about the alignment problem and they really thought it was a spurious idea, would stake their claim to this particular piece of real estate, which is that there's no reason to think that these systems would form preferences or goals or drives independent of those that have been programmed into them.