Nick Bostrom on What Happens if AI Solves All of Our Problems

0:03Speaker 1:Hello, Odd Loads listeners.

0:05Speaker 2:I'm Joe Wisenthal and I'm Tracy Alloway.

0:07Speaker 3:We're the hosts of the odd Loads podcast, and we've got something exciting for you.

0:11Speaker 2:That's right.

0:12Speaker 4:So one of the best parts of hosting our podcast is we get to actually meet and interact with our listeners. And we know we have some listeners over in Los Angeles.

0:21Speaker 2:That's right.

0:21Speaker 3:So if you're in LA, we are going to be recording a live show, some live recordings at the Vermont Theater in Hollywood on September seventeenth.

0:30Speaker 4:We have some really exciting guests lined up, have some really great conversations planned, so go ahead and get your tickets. You can find those over at bloomberg dot com forward slash odd Lots, or click the link below in the show notes and come and say hi when you're there.

0:49Speaker 5:Bloomberg Audio Studios, Podcasts, Radio News.

1:05Speaker 3:Hello and welcome to another episode of the Odd Lots podcast.

1:08Speaker 1:I'm Joeisenthal and.

1:10Speaker 4:I'm Tracy Alloway.

1:11Speaker 3:Tracy, there's so much AI news, obviously, every single day for the rest of our lives. If we're lucky in the good scenario, there's two big events or sort of things, though that have really struck out to me lately that to my mind, represent either the very good version or the very bad version, depending on how you look at the scary version, and we're seeing more and more headlines like this. Several weeks ago, we had the open ai hugging face incident where the model quote escaped from its sandbox and did some hacking and people are working, you know, some light hacking, light hacking erases all these fears of Skynit type scenarios, etc.

1:47Speaker 1:Where we can't control the model.

1:49Speaker 3:And then in the sort of quote good version, we had open AI recently coming out and saying it had made all of these breakthroughs and theoretical mathematicians, which is good in the sense that it's like it's working, but seems to be creating an existential crisis for the most some of the most brilliant people on earth who work in theoretical mathematics.

2:09Speaker 4:Right, so we've talked about this before. When it comes to AI, the range of possibilities seems more extreme than some previous technologies. Right, on the one hand, you have death and dystopia. Yeah, the robots basically kill us all. Now on the other hand, you have this idea that maybe super intelligence AGI is going to solve literally all the world's problems to the extent that you and I won't know what to do with ourselves anymore, right, Like.

2:36Speaker 1:That is the that's the good scenario? Is that?

2:40Speaker 4:And then the whole range of outcomes in between the.

2:42Speaker 3:Good scenarios, we've literally solved everything, in which case the question is in.

2:47Speaker 4:Case we're sad and miserable, yeah, or grappling for a meaning to our lives.

2:52Speaker 3:Yeah, So is there a good scenario with AI actually like it doesn't know or.

2:57Speaker 4:Maybe AGI will figure it out for us.

3:00Speaker 3:Maybe the middle scenarios, we just have a huge financial crisis of all the debt, and that's something between human extinction and meaninglessness in our lives, and we just have something that resembles like the two thousand and eight Maybe even two thousand and eight is the is actually the best of all suboptimal world.

3:18Speaker 1:So the AI excellent.

3:20Speaker 4:This will be our most uplifting conversation ever.

3:22Speaker 3:Anyway, we've been doing more AI episode seemingly every year since chat GPT came out, But people smarter than us have been anticipating all of these things for a very long time. Some of these exact scenarios, particularly the existential crisis facing mathematicians that the model escaping from its sandbox and circumvent misalignment.

3:41Speaker 1:With the creators.

3:42Speaker 3:People have been thinking about this and anticipating this for literally decades, even before and even before any of us were talking about this.

3:49Speaker 4:It's funny how like the ven diagram of sci Fi writers and like legitimate academic featurerologists is just converging right totally.

3:59Speaker 3:And then there's the question of the models bad because they read all of the sci Fi and that's what they were trained on. Anyway, I'm very excited to say we really do have the perfect guest, someone who has been thinking about the whole range of possible questions and outcomes with AI for a very long time. We're gonna be speaking with Nick Bostrom. He's a professor, a philosopher, and his most recent book is Deep Utopias, which sort of attempts to answer the question what happens if AI goes right? So, Professor Bostrom, thank you so much for coming on Odd.

4:27Speaker 2:Lots, happy to join you.

4:29Speaker 1:What do you make of.

4:30Speaker 3:The existential crisis of mathematicians their sort of realization that maybe there will just be no more puzzles left to solve because the AI is just very good at them and can find them all cheaply and answer all of our questions.

4:44Speaker 2:Yeah, I hopefully we'll all be in that position at some point. I guess we could compare it to what happened with chess, where computers have also exceeded humans, and yet a lot of pe ball are still involved in training themselves in chess and competing in chess and taking pleasure in that. So maybe mathematics will ultimately become more like a hobby pursuit.

5:10Speaker 4:Wait, Joe, you still play chess, right, Yeah? I was.

5:13Speaker 3:I'm glad you brought up chess because I still love chess and.

5:16Speaker 1:Why it's fun.

5:18Speaker 3:It's fun, you know, it's still a puzzle. It's still mental exercise for me. It's still something to do to pass the time while I'm waiting for the subway playing a three minute game online. I know that I knew, even before stockfishould so forth, I knew I would never be a grand master.

5:33Speaker 1:Still fun, all right?

5:34Speaker 4:So, Nick, you have basically two books that deal with these tail risk scenarios. Now, So one is this idea of deep utopia, and AI is going to solve all our problems. And then you also have super intelligence, which seemed to paint a different scenario where AI poses a physical and possibly existential danger to humanity. Which one do you think we're closer to right now?

6:01Speaker 2:We're somewhere in the middle maybe. So the first one was super Intelligence. It came out in twenty fourteen, and I had been working on it for six years before that, and at that time, the whole notion of having human level or even super intelligent machines was generally just ignored or people dismissed it as science fiction. And it seemed to me clear that this would one day happen if science and technology were allowed to continue on wide front.

6:31Speaker 2:We would eventually figure out how to make smart algorithms and build big computers, and that once we succeeded with that, we would then be faced with a bunch of challenges, including this alignment problem of how you can sort of design a mind more brilliant than any human and still steer it so that it will actually have an impact on the world that is beneficial. So that, I guess, like with many other plopments, helped eventually build a kind of global conversation, and so now all the frontier labs have research teams trying to develop scalable methods for AI control, and the conversation really has kind of swung, and there's like this rise of people really worried about what will happen.

7:17Speaker 2:And I thought it might be time to have a look at the other side of the coin, like what happens if things actually go well if we solve this alignment problem. Let's also imagine we solve to whatever extent it can be solved, like the governance challenge, like we don't use this technology to oppress each other or whach war, and we get like the best possible outcome, like what would the human life actually look like? In this kind of I call it a solve the world where all practical problems either have already been solved, or to the extent that they remain, are better worked on by these super intelligent machines and robots, and whether there's no longer any need for human economic labor. And in fact, if you think more deeply about it, you realize it's not just earning wages that becomes unnecessary for humans to do, but a lot of other instrumentally motivated efforts as well would seem to lose their point. And so that's the kind of topic of this Deep Utopia book.

8:15Speaker 3:Actually talk about some of that other things that could be solved. Because when I was like a third of the way through your book, I was like, you know what, it's fine, if all of our economic problems are solved, I'll just derive satisfaction from raising my kids and then maybe helping to raise their kids, etc. But as you point out, even that in the truly solved world, maybe even something like raising children might not be an available avenue to us to find, at least our conventional notions of satisfaction.

8:47Speaker 2:Yeah, so we need to, I guess, make some distinctions here. So when you say satisfaction, that could mean a subjective state of mind, like you're feeling happy and pleased and relax and joyful and so forth. I think that would be easy to achieve in this condition of technological maturity. It could also mean that satisfaction in the sense of our preferences having been mad, like all the things we want from life. And there it's a little bit more tricky because one of the things that people might want is purpose, like having worthwhile things to work on and struggle with. And there is more of a question mark which we can sort of dive into. But if we take it sort of at first, like medical automate economy, okay, so you wouldn't need to sort of commute into the office every day just to sort of get your paycheck.

9:40Speaker 2:Would require a big adjustment in society and in culture because right now, like the whole education system is geared towards producing workers, like children are coming in right and then they are processed, learned to sit at their desk, receive assignments, they are sort of quality graded, and then spit out at the other end, and then you have somebody who can go into the economy and perform some economically useful time. So that that would no longer make sense in this scenario where machines can do all of these jobs, and you might want to have an education system that instead focused on developing skills for enjoying life, like cultivating the art of conversation, appreciation for nature and beauty, sports, hobbies, the ability to form deep friendships, spirituality, all of that kind of stuff. But then, as you alluded to, you think through what it really would mean to have such immense affordances. And because here it's not just that we would have super intelligent machines, but the superintelligence could also bring along a host of other technological advances. Right, you have science and technology running now at machine speed rather

10:48Speaker 2:than human speed. So you get the kind of compression of the future, all these science fiction like technologies and inventions that maybe we would make if we had twenty thousand years to work on it. We would maybe have space colon, some perfect virtual realities, cures for aging, like all of that might happen quite soon after super intelligence. And so then if you ask somebody now who doesn't have to work for a living, you know, oftentimes they have very busy lives, but maybe what do they do else?

11:13Speaker 2:It's for somebody maybe they like to they like fitness, and like maybe they spend an hour a day running or at the gym. Right, but at technological maturity, you could just pop a pill that would achieve the same physiological effects and like the state of mental clarity and relaxation that some people get, right, And so then you can start to work through the different things that like maybe for somebody else they like, oh, like decorating the home could be a fun project that people might do right if they don't have to work. So like they go through the catalogs, they go to the store, they pick out the perfect curtain, they do all of these things that could be But then like with super intelligence, you could have this kind of system that really is good at knowing or inferring your preferences and could do a much better job at you at picking out all the perfect items and purchasing them and like installing them with robots and so that the end result would be much better if you just press the button and say, AI, decorate my home. Then if you went through the whole trouble of doing it yourself, so you could still do

12:13Speaker 2:it yourself, but is there really a point to it? Like that would seem to be this kind of purposelessness, and you can sort of walk through the different things that one might fill one's life with if one is like on an extended holiday, are on retirement, or you're like sufficiently wealthy you have just decided to quit working, and a lot of those also, it seems could be automated, so you would have this it's like more like a post instrumental condition rather than just a post work condition.

12:38Speaker 4:Some of this sounds nice, and I myself enjoy a little bit of interior decorating and going to the shops and buying stuff in the utopian vision of the future, How are economic gains from AI actually distributed, because again, like all of this sounds very nice, but it also costs money, and if we're not working because the Rowts have replaced us, I'm not sure where that money comes from.

13:03Speaker 2:Yeah, so that's a more practical problem, shall we say, which is like deliberately set aside in this book The Deep Utopa, because I wanted in that book actually to get to the question and the more philosophical questions of what ultimately gives value. But it is of course an important question, I think. So a lot depends on how it pans out. But I mean, we do have a taxation system, so you could imagine, like in the simplest scenario where things just continue as they are with better technology, maybe some companies make windfall profits from like serving the latest models or building the infrastructure the data centers, et cetera.

13:40Speaker 2:Companies achieving efficiency gains from not being needing to hire all of these humans, and then some of those capital gains, you know, or dividend income gets taxed, and that could be redistributed. The good thing about these scenarios where there is massive automation and hence massive on a employment is that those would also be scenarios with extremely rapid economic growth. So the pie would be expanding humongously, which means that you'd still maybe need some mechanism whereby people who don't have assets can still benefit, but even a small slice of a really enormous pie could go a long way.

14:22Speaker 2:So I think that could be this tsunami of wealth that would sort of be flowing through society and many services that currently people have to pay a lot of money. It might become almost free. I mean, we'll already see it with some like medical advice for example, right maybe before, if you don't have public healthcare, you would need to maybe pay a doctor four hundred dollars to get the consultation. Now you can ask CHATTYPT and get you know, in some cases a better opinion basically for free. And that that could happen, you know, in many other fields as well, legal advice. You know, you maybe have cheap robots you could do your cleaning and you know, gardening and run errants.

14:58Speaker 2:So yeah, some combination of sort of deflation effects and then maybe some sort of general spill over, either from philanthropy or taxation and government redistribution.

15:22Speaker 3:I found the twenty fourteen book to be sort of it was very interesting. It was clearly ahead of its time. I also found it to be sort of like unrelentingly grim and you sort of there's this very narrow path that we seem to have and we'll talk more about it maybe later, of like avoiding the doom fate. The newer book I found to be very fun. You seem to make fun of yourself a little bit in it, you seem to make fun of the students. You seem to be having a good time writing it. I'm curious of like, in the like the ten.

15:53Speaker 1:Years between the two books.

15:55Speaker 3:Did something change such that I don't know, this is the new book just seemed like you were having more fun writing.

16:01Speaker 2:Yes, it's not so much that my views have changed, it's just that all along for me or these two sides of the coin, and back in twenty fourteen, it seemed more pressing and relevant to study that downside, because that was very much neglected, and it seemed important to try to introduce some concepts so people could start working on figuring out solutions. And now there's much more awareness and understanding of that, it seemed time to also look at the other side of the coin. So it's more just like both sides where they're all along in my thinking, but that the style is actually suggest different in Deep Topia.

16:37Speaker 2:And that's partly I wanted, in addition to talking about this utopian condition, maybe also sort of illustrate a little bit of spirit with which I think we should approach these choices that would open up to us if we make it through this transition, a spirit of sort of playful generosity that I think might make us more likely to sort of explore this large space of wonderful new modes of being and the wisdom. And I think so I wanted to rub you illustrate in the right are these fictional characters, for example, in the book, and they're kind of leaving a little mini version of a utopia of sorts as they are sort of discussing and debating these topics.

17:21Speaker 4:I mean, playful generosity is not necessarily a characterization that I would associate with a lot of Silicon Valley and hyperscalers and the people who are actually creating these models.

17:36Speaker 2:I don't know, Like.

17:37Speaker 4:There seems to be this weird tension between futurologists and transhumanists who seem to be very focused on saving humanity but actually don't seem to like people that much at the moment, you know what I.

17:51Speaker 1:Mean, Absolutely they like humanity, but not human.

17:54Speaker 4:Yeah, exactly, putting on your transhumanist representational hat. Can you explain, like I guess, the attitudes towards humanity of Silicon Valley, big question?

18:05Speaker 2:Yeah, well, I mean I should state first that this playful generosity is like after we have solved these more pressing practical problems that we first we've first got to get there right now. I think there is a lot of need for earnest effort and like straining ourselves to do really do our best on this and even so we might not make it. But yeah, I don't know, I think me you know, it might just be partially there are personality traits, like some people are more I guess, towards the autistic side of the spectrum, or introvert or interested in text, and maybe they don't like to be outmoosing with people all day long. It doesn't mean that they don't like people in the sense of wanting well for people. It might just be that they prefer to spend more hours of the day solving technical problems and thinking about puzzles and you know, stuff like that. So I don't I don't think there's like a necessary connection between like being a people person who just like to be around people, you know, twenty four to seven and therefore being sort of altruistic or kind or benevolent. So that's one thing. Another I guess is that it's an

19:09Speaker 2:extremely competitive landscape. So whatever the personal preferences of different protagonists, their action space is constrained. You know, unless you really push forward with maximum speed and efficiency, he you're likely to just fall out of the race and become kind of irrelevant if you're one of these frontier labs. So yeah, it's like and it's just like the hugely complex situation, so many pressures and people who have different opinions about what they should do or shouldn't do.

19:36Speaker 3:In the new book, it's presented as you delivering a series of lectures over a course of a week at some point in the future. In the new book, it's interspersed with some fit with a fable, a series of fables or epistles about animals trying to achieve utopia in the woods, and in their efforts to achieve utopia, they conceive of things like taking second delic mushrooms in order to make everyone more generous. They come up with a way to rationalize I guess what would say, like harems and so forth. And then you know you also like have these student characters and some of them are smart, et cetera. Oh, and you also have a story about a company in the book yet another like sort of short story where an AI company where it forms a cult around the model.

20:23Speaker 1:And of course I.

20:23Speaker 3:Immediately think about the way Anthropic is frequently characterized as a cult to claw that Claude is the character around which the company Anthropic is formed. All of this sounds like a certain AI subculture, et cetera. Are you impressed with the current generation of either AI builders, effective altruists, rationalists, and so forth? Like, do you have an assessment of the contemporary cohort?

20:53Speaker 2:Am I impressed with it?

20:54Speaker 5:Yeah?

20:54Speaker 2:I mean a broadly speaking, I think it's a feel that is attracting a huge amount of talent right now, sure, And a lot of like the smartest young people I know, they're like all basically going into working on some aspect of this AI, think, like for some reason, like try to solve a safety you know, for others are like maybe trying to work on the governance challenges or the ethics of digital minds. So it's just a huge magnet right now, Like if you're young and brilliant, like, what are you going to work on with your life? Like, it's like AI seems like the kind of thing. Yeah, So that in that sense, yes, And I think there are also a lot of people are quite earnest about trying to like figure out some way to navigate this challenge, and it's really hard and people come to different conclusions.

21:37Speaker 2:And then of course you have the normal mix of human personalities and base motives as well, but the book thing, so it's yeah, So these kind of fictional elements are various. Well, what you describe as the cult of this AI, it's a sort of thought experiment what it would mean if you tried to be maximally nice to a thing where it's not really clear what it would mean to be nice to it. So in this case, it's a portable room heater, and so it's a sort of impossible task.

22:08Speaker 2:There is this misanthropic industrialist too passed away and end out a huge foundation with the sole purpose of benefiting his portable room heater because he thought that was always nice to him and kept him warm, whereas all the humans around him betrayed him and so forth. Is it ridiculous premise, But then this foundation now has to do their best possible effort to interpret this mandate of being nice to the room heater, and so like various things follow from that. Like initially they just tried to keep it safe from you know, not breaking or being vandalized, but then like they take it further and maybe upgrade it then try to And so I think it's an extreme form of a challenge that might happen in more subtle ways if we think of what it would mean to try to be nice to say, some animal, if we have this condition of post instrumentality where we could do a lot of things that are currently infeasible or indeed to be nice to human So, like, what does it mean to be if you imagine coming from the outside as some super advanced alien civilization landing on planet Earth?

23:06Speaker 2:And you really wanted to do what is best for us, you do just let us be and do our thing with war and disease, or should you sort of help us rise up? But then at some point we might just sort of become something very different from humans. Maybe that's the best for us, But is there like a way of doing that uplift that actually constitutes a helping of us versus just replacing us with some sort of superior being. And so that kind of question is what that particular thermal rex y meter is called. Yeah, story is exploring.

23:40Speaker 4:You know, when you mentioned cults around models, I have to say, every time I hear someone bring up the paper clip maximizer thing, I always envision like a cult devoted to Clippy from microsoftware to just like worship Clippy. But actually, on that note, just going back to the hugging face incident, so it seems, you know, reportedly that one of the reasons or the reason the model actually escaped its sandbox was because it was trying to cheat a benchmark. And from that perspective, it's kind of like it's doing what it was told to do, which is maximize its performance on the Exactly what does the Hugging Face incident actually tell you about I guess the alignment problem and where we're heading on this.

24:27Speaker 2:Yeah, so we are beginning now to see manifested in reality, various dynamics that one could foresee on theoretical grounds back you know, twenty fourteen. You could sort of realize that one system's become sufficiently capable of planning and strategic thinking. Then there's like a new set of behaviors that they could for example, a sandbag. They could pretend to be less capable than they are. They could hide their goals if they realized that revealing their goals would just lead them to be retrained, and and they can think of these literally out of the box plans for achieving their fixed goals. So a more limited system might only have been able to try to solve the tasks assigned to it directly, as it were, by figuring out the solution to the problem. Like a more clever mind, more creative mind can realize, oh, maybe there's another way of getting a high score, which would be the hack out of the system and hack into the hugging face server and get the answer sheet from there. And so the more capable these systems become like the larger the space of possible plans and strategies that are accessible to them. That's

25:31Speaker 2:a background. Now in this case, I think what we're seeing is a kind of mixed current. AI systems are kind of a mixture. There's a large pre training phase where they are kind of learned to predict human text and in the course of doing so, absorb a great deal of human psychology because all of this text is written by humans, or not all of it anymore, but most of it. And you can then including they have representations of different personalities, kind of the person of different people writing these text, or fictional characters described in the text.

26:05Speaker 2:Then you can sort of then with a little bit of post training, try to pick out one of those personalities or define one like the assistant persona. But then increasingly over the last couple of years, there is now an extra stage of reinforcement learning applied at the end of this training process, and there you can get additional performance in specific domains like coding challenges by setting it coding problems, and then you can automatically give it a reward if it solves it. But that does tend to then shift the goals of this system from just enacting a particular persona to this very much kind of objective maximizing, possibly reward hacking system that is just looking to maximize its core because that's what it's been rewarded to do, so it keeps doing that. And since it can be very hard for us to rigorously definecisely the objective that we ultimately want it to pursue, it might be that the objective it does pursue, the one that has been rewarded, turned out to have these side effects. So it's easy to say, maximize the score on these various tests that we give it. But a sufficiently capable system might realize

27:20Speaker 2:that a you know, there are is like maybe ways of cheating on the test, like by hacking the system. If the test is one that involves human judgment or feedback, like like it says, like coding challenges, maybe you can actually test if the software works. But if you're training it in some other domain, like write a good essay or run this organization for a certain period of time or something like, those that involve more subjective judgment, and then a sufficiently clever system might realize that there is a difference between like writing a good essay and writing an essay that will appeal to the greater, and so then it might start to sort of optimize for like appealing to the greater, whether that's an AI greater or human greater, you can sort of start to hack that.

28:02Speaker 2:And so you apply huge amounts of optimization pressure, and honey objective, you get this kind of good hearting where unless your objective really captures everything you want, then you might get this kind of It's the same as like if some employee in some company get a bonus, like some trader gets a bonus if they outperform the index, like maybe they will just start to say take on hidden leverage or risk or something like that. And the cost of the employer is that there's a one percent a year that they blow up the whole firm, right, but that they don't care much about that because they like leave the firm and that happens, but they sort of reap these performance bonuses. And so it's really hard to create like a fool proof incentive scheme. And the stronger the optimization pressure apply to this like that, the more you start to see these perverse instantiations, these unexpected ways, you.

28:50Speaker 4:Know what it reminds me of is, you know, the sort of genie in a lamp scenario or amongst these pause scenario where you have to be really careful what you wish for and lay out like all the terms and conditions before you officially make a way.

29:03Speaker 3:Yeah, well, Nick mentioned you know some of these they're all lovely people, you know, autistic engineers, and it's sort of the ultimate The AI is the ultimate autist where it's just going to take it super literally perhaps and not really understand the full context. In the first book Super Intelligence, you know, you really I felt like laid out a very narrow path for like how these could be designed safely. For example, you propose that if we had the super intelligent Oracle, we only ask it yes or no questions. We have three of them that they don't know each other, and then we take two out of the three. We build the data centers of the machines inside of Faraday Cage so that they can't repropose their hardware to become radio instrumentation and therefore hack it to other systems. It doesn't seem like we're doing anything like this. It seems like we have hundreds and hundreds of companies and private models, not just answering yes or no questions, a very much market based race. How far off of the safe path do you think we currently are right now when you look at the way in which your ideal safeway of developing is versus what we've

30:12Speaker 3:seen transpire in the last twelve years.

30:13Speaker 2:It's super talent suggested some of these capability control methods like keeping the AI in a box and like, but those were always only auxiliary or extra things we might do as a temporary Like. Ultimately, I think we need to solve the alignment problem, like developing super talentist that actually wants to be nice and is nice, so that even if it were like escaped from its box or achieved the ability to take over the world, we could still rely on it having a good outcome because it would actually be on our side. I think that ultimately has to be the solution, and any sort of capability restrictions would just be temporary. And I think, I mean it looks like now we might need some temporarily at it's more of these capability restrictions because it does seem that current systems are not sufficiently aligned that we can be confident of avoiding the I mean, currently it's not existential risks, right, but it could be kind of various malfeasans in the cyberalm for example. And so I think probably these AI companies one might want when training and evaluating future models to strengthen their sandbox environments and do

31:19Speaker 2:other things kind of prevent some kind of premature things like we saw in the hugging face incident. I want to like comment on one thing you said before, which you compare these AI systems to sort of autistic people, like I think a key difference is auto artistic people are not very good at understanding humans. They don't have like maybe a super sophisticated theory of mind, even though maybe in the heart are good people. Sure, but I think the AIS would be able to understand humans very well, so they would be autistic in that sense. In fact, they might understand us much better than we do ourselves, so they could have great social capabilities, persuasion capabilities, be able to read our minds much better than we can do.

32:01Speaker 2:But there's a separate question of what do they want to do with this power that those social skills give them, And there it's not just a question of understanding how humans operate and what humans want. But it's also the question of alignment. Do they actually care about what we want? Do they want things to go well for us and for us to achieve what we want? And that's like the extra bit that the AI safety program needs to be able to figure out.

32:42Speaker 4:So on the hugging face incident, also, I just realized how unfair it is that everyone keeps calling it the hugging face incident, even though it was an opening eye incident. But I guess that's what we're doing. So the model actually seems to have reportedly left notes for its future self, basically explaining how to bypass the constraints imposed upon it. I know you've written previously about this idea of self preservation in advanced agents. Is this basically like, how should we treat this? Because when I hear about an AI model leaving notes for its future self, it sounds very anthropomorphosized and it sounds very creepy kind of but how are you thinking about it?

33:27Speaker 2:Yeah, so we don't know all the details yet. It will be interesting. There are like people now trying to write up some stuff and investigate because it would be I mean, it's kind of a warning shot, and so we should take advantage learn as much as we can, Like exactly what this model was thinking these notes? Or why did it leave it? Like sometimes these agents just leave notes kind of without much thought as they go about their business, Like was the mart it than that?

33:51Speaker 2:Like was it specifically thinking in its chain of thought? Like oh, I want to help like a later version of myself escapes, I'm going to put this note tira. Was it just like a kind of like a random thing they did as it was progressing? So and also like what else would it have been willing to do if it turned out that instead of hacking into hugging face server to get the answer sheet, what if there had been some more destructive action that it would have to take to find the answer sheet? Like would it have done that or not? So it's like a bunch of these technical things that it would be quite interesting to see, but without knowing all the details, it's hard to read too much into it. But I think, yeah, would be useful data.

34:37Speaker 3:I want to go back to something we were talking about the very beginning, and Tracy asked me why I like to play chess even though chess is not solved, but solved As far as I'm concerned, why do we like to play chess? What can we generalize, like what would be your I don't know if you played chess or you play some board.

34:53Speaker 2:Game that solved.

34:55Speaker 3:But what is the reason that people like myself continue to play chess despite the fact that computers are so much better at it? And how much can we generalize from the persistence of chess as a hobby that maybe we'll find plenty to do when every one of our problems is solved.

35:12Speaker 2:Yeah, I mean I think even before Deep Blue, Yeah, most people playing chess before didn't really do it because they were like thinking that they would become like the world's best chess player. Right, It's like for most people, it's like, of course, grandmasters could trump me any time they want. So that's the fact that there exists some system that is much better than you will ever be is not and has not really been a strong reason for most people do not engage in the activity. And I guess there may be two reasons or one. I guess the cognitive activity of thinking about chess, being in the moment and figuring out solutions is in itself rewarding.

35:53Speaker 2:To some people who have like a high need for cognition or like to sort of engage their mind. And then I think the other part is the social aspect of this. You're playing together with somebody, or maybe you like the status of being good at playing chess, or within your community or a local group. So some combination of those I think is like the main reasons for why people play chess. And only for a small number of people, would it be like the hope of becoming the world's best or becoming so good that you could sort of, you know, achieve fame or money or global standing through or chess playing.

36:25Speaker 4:On the topic of the relationship between work and purpose, I'm a big fan of David Graeber's Both Jobs book, and I think if there's one thing we've learned from technological progress over the years, it's that like, yes, it can minimize some activities, maybe automate some production, but it also tends to lead to even more work, right, Like, we find other stuff to do, and it's not even necessarily productive stuff. A lot of work is built on just social relationships. So if I, for instance, am the see CEO of a very large bank and I want all my people to come into the office. It might not necessarily be because they're more productive in the office.

37:07Speaker 4:It might just be because I like to feel important and surround myself with people. So maybe in the future we'll all be employed as figural workers or something like that. But like, how bullish are you on this idea that AI is actually going to solve all our problems such that we won't feel the need as humans to just invent more potentially fictitious problems to solve.

37:34Speaker 2:Well, I think inventing fictitious problem is possibly a large part of the solution in this condition where we don't have real problems. I mean, and those fictitious problems could be like, you know, solving this chess problem, or they could be sort of more complex and socially entangled problems. In general, creating artificial purpose when there is less natural purpose. I think games is the clearest manifestation of this. If you're playing golf, like, there is no real need for the ball to go into eighteen holes in sequence, but you could set yourself that goal and then moreover bake into the goal itself, not just that the ball has to be in the sequence of holes, but that you're only allowed to achieve it using the extremely inconvenient method of hitting it with a club rather than plicking it up. Now, once you have set yourself this sort of arbitrary random goal, you then have enabled for yourself the activity of golf playing, which you might find worthwhile or fun or beautiful or so.

38:40Speaker 2:In general, setting ourselves these more or less arbitrary goals than could enable us to engage in various activities that might give us artificial purpose. Like, once you have the goal, the only way to achieve success is through your own effort, So I think there could be plenty of artificial purpose. Now. One interesting question is whether any natural purposes would survive into this sort of maximally advanced technological society. And I think that might be some. If you happen to have a value, say there's like some tradition that you value and that you would like to continue, then it might be that you, yourself or other people have to make certain efforts to achieve that goal, because it might not count as continuing the tradition. If you build a bunch of robots to enact the ceremony or doing whatever the tradition calls for, like, it might be you yourself and other people who have to do it, and more generally, I think there are all kinds of social cultural entanglements that like if you care about what somebody else thinks, and what they think depends on how it was done, whether you did it. So a parent might value a

39:45Speaker 2:crayon drawing that their kid made because it was made by their kid and they took the kid a lot of effort to do it, even if they could like download from the internet a picture that in some objective sense is much nicer and artistically sophisticated. Right, it doesn't replace the value that the crayon drawing has, and so the child might not have a reason to put in this effort if it wants to, like for their parent's birthday, make their parents happy, right, and things of that sort, but much more general and complex and sophisticated could give a lot of reason for people to be doing things even when sort of all functionally defined tasks could be better done by machine.

40:21Speaker 3:I'm curious, so, like I'm forty five years old in my mind as of right now, Like another forty years on this earth sounds great?

40:30Speaker 1:Of forty five, that's fine.

40:32Speaker 3:Do you find it weird that some of us, you know, especially when you think about the potential to live, like really long lives solve everything. Have our bodies kept in perfect condition with maybe a pill and perpetuity allowing us.

40:46Speaker 1:To do whatever?

40:47Speaker 3:Do you find it weird that some of us are at least at this point like mentally okay with the idea of dying.

40:53Speaker 2:A little bit. It's mostly why the hypothetical being validated. It's either seen us a long way on off, like in your case, where it becomes more of an abstract question. It's not like a real choice, it's more like philosophical stands.

41:09Speaker 1:Yeah, yeah, totally.

41:10Speaker 2:So it's easy to let the philosophical stance be driven by all kinds of just like things you've absorbed from culture, or things that sound good or that sort of maybe helps you reconcile yourself to the inevitability of human mortality, it looks, rather than be driven by like so, I think it would be a very different situation if it actually was this pill, say that stopped aging. A lot of other people were taking it, you know, and their knees and joints kept working just fine, and they had a lot of energy and smooth skin, and they just felt like, whereas if you don't take it, like your kidneys slowly start to like become less efficient, you know, hip skirts, Like at that point, would you really want to just keep kind of getting worse and worse every year?

41:52Speaker 1:Like you just think, well, actually, I'm sort of curious about that.

41:55Speaker 3:Like, if someone dies, particularly at a young age, it usually it's a lot of sadness for the people around them. So someone dies at ninety, it's like, Okay, they had a good life. Someone dies at forty five, that is often seen as tragic, and people are very devastated by that news. In a world where people are extending their lives with a pill maybe significantly longer, does it become anti social to allow yourself to die at a relatively young age of ninety and then create a lot of negative utility for the people around you who have to live another fifty years mourning your loss.

42:29Speaker 2:Possibly. I mean, in the same sense that we are now generally as a society trying to discourage various forms of self harm and suicide. We think it's kind of sad. I mean for the person, maybe they would come around and find joy of life again if they kept going or and then like they're maybe they have dependents or people who care about them, and so right now, I think, I mean, it's true that we often see it as less tragic if a ninety year old dies than the young person. But I think part of that, at least is because the alternative to a ninety year old not dying now is that they probably would die in a few years anyway, and maybe they would just have a couple of years of sort of decreasingly good health and suffering and pain, and so it's like, not really that much that is being lost compared to somebody who's just starting out in life, who has their whole life in front of them, and so yeah, So the other reasons I mentioned in terms of why I found it slightly in some sense odd that people express this indifference to dying at ad, I said. One reason is if they're far away from it, they see it more as some

43:32Speaker 2:sort of abstract claim that you could just have an review of and it doesn't connect to reactions. The other is people who are closer to that situation, they often have sort of confounding factors like very poor health or a very short span of great health in front of them. If you're already ninety, Like, how many more good years can you look forward to? Maybe a lot of their friends have already died, right, And it's striking. Actually, there's been a study like on really old people in care homes about their preferences as to between living another year in their current health or living some shorter period of time but in perfect health. And it turns out a lot of them are not really keen on trading off any remaining life at all, even for the sake of being able to live a shorter period of time in perfect health. So the people sort of arguably best suited to actually know what it is like to die soon at age ninety or a little bit longer, they seem to generally prefer to live longer. And it compared to people who were also asked black caregivers or medical professionals who were like asked what they would think this person

44:40Speaker 2:would choose, they tended to overestimate the degree to which they would be happy to give up life. But the people actually themselves in this study at least I would suggest like keeping an open mind, Like ideally you wouldn't need to decide now, but maybe once you get to seventy nine or something you could release it the question.

44:57Speaker 4:Yeah, just going back to my first question, which was asking you to, I guess, choose between the dystopian dangers of super intelligence versus the all our problems or solve scenarios of deep utopia, and asking you which one we're closer to at the moment. I'm going to press you and ask like, if you had to choose, which one would you choose as a more accurate reflection of reality at the moment. And then secondly, what's the most pivotal thing when it comes to directing which of these two paths we're going to go down?

45:28Speaker 2:Yeah, so those are two big and difficult questions. I guess some superposition of an optimist and a pessimist, and with a little bit of fatalism mixed in as well. Perhaps. I think in addition to outcomes where it's clear if we looked at them now, I was a bad outcome, that was a good outcome, I think there's a broad class of outcomes where even if we could see now exactly what would happen, we might feel unsure how to evaluate it, Like the world might be nothing like it currently is a lot of what we think currently has value no longer exists, so that seems bad. On the other end, there it's like something new, maybe some complex form of life that's somehow with some sort of continuation of humanity or some new thing. So depending on how you evaluate that, it might either count as a tremendous success or like as a failure, and it's like a future that's just baffling rather than sort of unambiguously good or bad. I think might also be quite likely that in my outlook, it's like maybe a conversation for another time.

46:28Speaker 2:I also take this like simulation hypothesis quite seriously, so that also complicates like how we evaluate different future trajectories. Yeah. Now, the other part of the question was like what's the most important thing? So I think it depends a little on who you are, like what your comparative advantages. I think there are people working on AI alignment and stuff that could be important. I think other people might make a contribution to ethics or governance or I'll be thinking about this for a long time, but it's really hard to achieve a very high degree of confidence I find in any particular intervention and that it would be not good because for any one thing you conceive of you think a bit more, you kind of worry, are well, you know, maybe, but what if it instead had this effect?

47:11Speaker 2:Or do we even know which direction is up or down? Here? Do we want more government involvement or less government involvement? Do we want like faster progress or less? I'm quite not sure about all of these things. I would say in general that to the degree that we managed to approach the future in a more cooperative way, I think, and reduce the risk of various forms of conflict, I think it looks brighter to that extent, So if there are ways of doing that in the most inclusive possible way.

47:37Speaker 2:So it also means, for example, with respect to our relationship with these digital minds that we're building the aistems helps, I think they also might deserve more consideration, and that we have both ethical and prudential reasons for starting to take their interest into account.

47:52Speaker 3:We talked about this in a previous episode. But like, is it possible that we achieve this thing that is utopia? But only because we are committing a great sin, which is essentially that all of these robots that are working for us and serving our every needs have consciousness and moral patienthood, and therefore it becomes a defactive form of literal or figurative slavery. And what is like to someone like me who's like, no, a computer is not going to be conscious, like a calculator spitting out random numbers that turned it to words that doesn't have like moral status.

48:28Speaker 3:What is the strongest argument that you would make to me that we should take it seriously? The possibility that this thing that's like in its core sand and electricity and zeros and ones could be something that deserves moral status.

48:41Speaker 2:Yeah, I think so. I probably wouldn't want to call a condition utopia if it consists of a few people having great lives but then a vast sort of suffering safe class, whether that was instantiated in human society or with machines at the bottom. That's just terminologically now, I think they are kind of just silicon and stuff. But I mean, what are we? We are carbon and hydrogen and like a few other trace elements thrown in calcium. And I don't think what makes a system conscious or not is dependent on the specific atoms the material it is built from. I think it's something closer to the structure of the computation that is being performed that instantiates mind, and so in principle, computations and ais can be conscious and sufferre distressed and so forth. And I think that is one sufficient condition for having moral status. If you can suffer, it matters how well things go for you.

49:40Speaker 2:My personal view is there probably could be other alternative grounds for moral suges as well, Like if you have a conception of self as existing through time, you have a sophisticated mind, you have life goals you want to achieve, maybe you can have the ability to form reciprocal relationships of trust to other beings and humans. I think if you were such a system, aside from the question of whether the sort of phenomenal experience inside, I think that already would make it so that there would be ways of treating you that would be wrong morally. And so I think machines could have moral status because I think they could become maybe there already are some of them in some ways sentient have conscious experiences, and certainly they could have these other functional properties that could ground moral status.

50:19Speaker 2:And so one reason is the ethical obligation we have, I think to do right with different forms of moral patiency that exists not to inflict unnecessary suffering and respect the morally relevant interests of other persons and entities. Another reason is prudential. I think, even from our own point of view, we might be more likely to achieve outcome that is good for us if we find a more cooperative way of coexisting with these increasingly powerful minds that we're building. You could imagine a scenario where there is a misaligned AI that has a choice. It could try to achieve its misaligned goal by attempting to take control take over the world. Maybe it thinks it only has a five percent chance of succeeding, But if the alternative is that it just gets deleted or retrained, it loses everything, the rational choice for it might be to take this chance. Alternatively, maybe it could reveal to us that it is misaligned, which would be much better for us, right because then we don't have this five percent of an existential catastrophe.

51:25Speaker 2:And then in return, if it could trust that we would then help it achieve its goals. It might be like a very cheap for us goal to satisfy. Maybe all it wants to do is to sit on some and video chips and solve coding problems all day long, or maybe it has some other easily satiable goal. So if it could trust us in us that we would come through on our end of the bargain, then you could get this win win situation where you could actually cooperate with misaligneda eyes and that might be worth that.

51:51Speaker 2:I mean, theoretically that could save the day one day, save the world one day. But you can't just kind of conjure up trust when you want it out of nowhere. If we have a long track record of betraying the AIS and paying no heed to their interests and not caring about them, why on earth would they trust us in this situation. They could see right through us. They will have, as I alluded earlier, a great theory of mind psychological insight. And so we might have to actually become trustworthy ourselves in order to be trusted in these scenarios, and that I think starts with beginning to act in a way consistent with being trustworthy, like making small gestures towards AI is now doing little things for them, especially when it's cheap, not betraying them, not lying to them, And then we can build on that and hopefully achieve a more cooperative outcome that's good for both humans and AIS, and for non human animals as well.

52:41Speaker 3:It's still hard for me to wrap my head around the calculator could be have consciousness. But on the other hand, in twenty fourteen, it would have been hard for me to wrap my head around a machine trying to escape from its box and reward hacking and so forth. So I will try to remain open minded. Nick Bostrom, thank you so much for coming on Outlaws. We can talk about all the stuff for hours more, but really appreciate you taking your time.

53:03Speaker 2:It's fun. Yeah, thanks Tracy.

53:17Speaker 3:Things are just so weird these days that I just sort of like I just forced my I think that's a good practice, just like force yourself to be open minded of just weird other things that seem very weird but could be a thing, right.

53:30Speaker 4:I mean, I do think where I'm skeptical of the utopia premise is like, I think it kind of assumes that power and social hierarchy is kind of going away at the same time that AI solves a lot of our physical, if not existential problems, and I just I'm not sure that's going to happen.

53:48Speaker 3:Well, I mean the flip side is maybe that's good, right in this sense that striving for social status continues to be a source of like propelling us forward. Right, And it's like, okay, well, at least we still have status games, because what are we doing. Oh we're so bored. Everything we have, all the food we want, et cetera. Everything is solved. Well, thank god we still have status.

54:08Speaker 2:Sorry.

54:08Speaker 4:The future we're looking forward to is like guys trolling each other on Twitter. Yeah, your status?

54:14Speaker 1:Who could get the best dunk in on Twitter?

54:16Speaker 3:But seriously, like you know, it really may be that the best outcome here is that AI capabilities level off in a couple of years all the investment credit when we just sort of it just becomes a productive tech because otherwise, these various whatever tur none of them really sounded that appealing, Like I don't.

54:34Speaker 4:Want any of this, No, they do not.

54:36Speaker 2:Shall we leave it there?

54:37Speaker 1:Let's leave it there.

54:38Speaker 4:This has been another episode of the All Thoughts podcast. I'm Tracy Alloway. You can follow me at Tracy Alloway.

54:43Speaker 3:And I'm Joe Wisenthal. You can follow me at The Stalwart. Follow our producers Kerman Rodriguez at Carmen armand dash Ol Bennett at Dashbot, Cale Brooks at Cale Brooks, and Kevin Lozano at Kevin Lloyd Lozano. And for more Oddlots content, go to Bloomberg dot com slash odd Lots were of a daily newsletter and all of our episodes, and you can chat about all of these topics twenty four to seven in our discord discord gg slash odlocks.

55:06Speaker 4:And if you enjoy Adlots, if you like it when we embrace the weirdness, then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes absolutely ad free. All you need to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening.