Knowledge Fight Interactive Search Tool

All Episodes

Previous Next

Dec. 13, 2020 - Lex Fridman Podcast

01:56:19

Michael Littman: Reinforcement Learning and the Future of AI | Lex Fridman Podcast #144

Large

Audio Only |

Scroll

Time	Text
	The following is a conversation with Michael Littman, a computer science professor at Brown University doing research on and teaching machine learning, reinforcement learning, and artificial intelligence.
	He enjoys being silly and lighthearted in conversation, so this was definitely a fun one.
	Quick mention of each sponsor, followed by some thoughts related to the episode.
	Thank you to Simply Save, a home security company I use to monitor
	and protect my apartment.
	ExpressVPN, the VPN I've used for many years to protect my privacy and the Internet.
	Masterclass, online courses that I enjoy from some of the most amazing humans
	in history and BetterHelp, online therapy with a licensed professional.
	Please check out these sponsors in the description to get a discount
	and to support this podcast.
	As a side note, let me say that I may experiment with doing some solo episodes
	in the coming month or two.
	The three ideas I have floating in my head currently is to use one
	particular moment in history to Two, a particular movie.
	Or three, a book to drive a conversation about a set of related concepts.
	For example, I could use 2001 A Space Odyssey or Ex Machina to talk about AGI for one, two, three hours.
	Or I could do an episode on the, yes, rise and fall of Hitler and Stalin, each in a separate episode, using relevant books and historical moments for reference.
	I find the format of a solo episode very uncomfortable and challenging, but that just tells me that it's something I definitely need to do and learn from the experience.
	Of course, I hope you come along for the ride.
	Also, since we have all this momentum built up on announcements, I'm giving a few lectures on machine learning at MIT this January.
	In general, if you have ideas for the episodes, for the lectures, or for just short videos on YouTube, Let me know in the comments that I still definitely read despite my better judgment and the wise sage device of the great Joe Rogan.
	If you enjoy this thing, subscribe on YouTube, review it with Five Stars on Apple Podcasts, follow on Spotify, support on Patreon, or connect with me on Twitter at Lex Friedman.
	And now, here's my conversation with Michael Littman.
	I saw a video of you talking to Charles Isbell about Westworld, the TV series.
	You guys were doing a kind of thing where you're watching new things together, but let's rewind back.
	Is there a sci-fi movie or book or shows that was profound that had an impact on you philosophically or just specifically something you enjoyed nerding out about?
	Yeah, interesting. I think a lot of us have been inspired by robots in movies.
	The one that I really like is there's a movie called Robot and Frank, which I think is really interesting because it's very near-term future where robots are being deployed as helpers in people's homes.
	And we don't know how to make robots like that at this point, but it seemed very plausible.
	It seemed very realistic or imaginable.
	And I thought that was really cool because they're awkward, they do funny things that raise some interesting issues, but it seemed like something that would ultimately be helpful and good if we could do it right.
	Yeah, he was an older, cranky gentleman, right?
	He was an older, cranky jewel thief, yeah.
	It's kind of a funny little thing, which is...
	You know, he's a jewel thief, and so he pulls the robot into his life, which is something you could imagine taking a home robotics thing and pulling into whatever quirky thing that's involved in your existence.
	Yeah, this is meaningful to you. Exactly so.
	Yeah, and I think from that perspective, I mean, not all of us are jewel thieves, and so when we bring our robots into our lives- Speak for yourself, yeah.
	It explains a lot about this apartment, actually.
	But no, the idea that people should have the ability to, you know, make this technology their own, that it becomes part of their lives.
	And I think that's – it's hard for us as technologists to make that kind of technology.
	It's easier to mold people into what we need them to be.
	And just that opposite vision, I think, is really inspiring.
	And then there's an anthropomorphization where we project certain things on them.
	Because I think the robot was kind of dumb.
	But I have a bunch of Roombas I play with and you immediately project stuff onto them.
	Much greater level of intelligence.
	We'll probably do that with each other too.
	Much greater degree of compassion.
	That's right. One of the things we're learning from AI is where we are smart and where we are not smart.
	Yeah. You also enjoy, as people can see, and I enjoyed myself, watching you sing and even dance a little bit?
	A little bit of dancing?
	A little bit of dancing.
	That's not quite my thing. As a method of education, or just in life, you know, in general.
	So, easy question.
	What's the definitive, objectively speaking, top three songs of all time?
	Maybe something that...
	You know, to walk that back a little bit, maybe something that others might be surprised by, the three songs that you kind of enjoy.
	That is a great question that I cannot answer, but instead, let me tell you a story.
	Pick a question you do want to answer.
	That's right. I've been watching the presidential debates and vice presidential debates, and it turns out, yeah, you can just answer any question you want.
	Let me interrupt you.
	No, I'm sorry. Yeah, well said.
	I really like pop music.
	I've enjoyed pop music ever since I was very young.
	So 60s music, 70s music, 80s music, this is all awesome.
	And then I had kids, and I think I stopped listening to music, and I was starting to realize that my musical taste had sort of frozen out.
	And so I decided in 2011, I think, to start listening to the top 10 Billboard songs each week.
	So I'd be on the treadmill and I would listen to that week's top 10 songs so I could find out what was popular now.
	And what I discovered is that I have no musical taste whatsoever.
	I like what I'm familiar with.
	And so the first time I'd hear a song is the first week that was on the charts.
	I'd be like, ugh. And then the second week, I was into it a little bit.
	And the third week, I was loving it.
	And by the fourth week, it was just part of me.
	I'm afraid that I can't tell you my favorite song of all time because it's whatever I heard most recently.
	Yeah, that's interesting.
	People have told me that there's an art to listening to music as well.
	If you listen to a song just carefully, explicitly just force yourself to really listen.
	I did this when I was part of a jazz band and fusion band in college.
	You start to hear the layers of the instruments.
	You start to hear the individual instruments.
	You can listen to classical music or to orchestra this way.
	You can listen to jazz this way.
	It's funny to imagine you now Walking that forward to listening to pop hits now as like a scholar.
	Listening to like Cardi B or something like that.
	Or Justin Timberlake.
	Is he? No, not Timberlake.
	Bieber. They've both been in the top 10 since I've been listening.
	They're still up there.
	Oh my God, I'm so cool. If you haven't heard Justin Timberlake's Top 10 in the last few years, there was one song that he did where the music video was set at essentially NeurIPS.
	Oh, wow. Oh, the one with the robotics.
	Yeah, yeah, yeah, yeah, yeah. Yeah, yeah.
	It's like at an academic conference and he's doing a demo.
	He was presenting, right? It was sort of a cross between the Apple, like Steve Jobs kind of talk and NeurIPS.
	So, you know, it's always fun when...
	AI shows up in pop culture.
	I wonder if he consulted somebody for that.
	That's really interesting.
	So maybe on that topic, I've seen your celebrity multiple dimensions, but one of them is you've done cameos in different places.
	I've seen you in a TurboTax commercial as like, I guess, the brilliant Einstein character.
	And the point is that TurboTax doesn't need somebody like you.
	It doesn't. It doesn't need a brilliant person.
	Very few things need someone like me.
	But yes, they were specifically emphasizing the idea that you don't need to be a computer expert to be able to use their software.
	How did you end up in that world?
	I think it's an interesting story.
	So I was teaching my class.
	It was an intro computer science class for non-concentrators, non-majors.
	And sometimes when people would visit campus, they would check in to say, hey, we want to see what a class is like.
	Can we sit on your class?
	So... A person came to my class who was the daughter of the brother of the husband.
	Husband of the best friend of my wife.
	Anyway, basically a family friend came to campus to check out Brown and asked to come to my class and came with her dad.
	Her dad is, who I've known from various kinds of family events and so forth, but he also does advertising.
	And he said that he was recruiting scientists for this ad, this TurboTax set of ads.
	And he said, we wrote the ad with the idea that we get, like, the most brilliant researchers.
	But they all said no.
	So can you help us find, like, B-level scientists?
	I'm like, sure.
	That's who I hang out with.
	So that should be fine.
	So I put together a list and I did what some people called a Dick Cheney.
	So I included myself on the list of possible candidates, you know, with a little blurb about each one and why I thought it would make sense for them to do it.
	And they reached out to a handful of them, but then they ultimately, they YouTube stalked me a little bit and they thought, oh, I think he could do this.
	And they said, okay, we're going to offer you the commercial.
	I'm like, what?
	So it was such an interesting experience because they have another world.
	The people who do nationwide kind of ad campaigns and television shows and movies and so forth,
	it's quite a remarkable system that they have going.
	Because they- Is it like a set?
	Yeah, so I went to, it was just somebody's house that they rented in New Jersey.
	But in the commercial, it's just me and this other woman.
	In reality, there were 50 people in that room and another, I don't know, half a dozen
	kind of spread out around the house in various ways.
	There were people whose job it was to control the sun.
	They were in the backyard on ladders putting filters up to try to make sure that the sun didn't glare off the window in a way that would wreck the shot.
	So there was like six people out there doing that.
	There was three people out there giving snacks.
	healthy snacks because that was a separate craft table.
	There was one person whose job it was to keep me from getting lost.
	And I think the reason for all this is because so many people are in one place at one time.
	They have to be time efficient.
	They have to get it done.
	The morning they were going to do my commercial.
	In the afternoon, they were going to do a commercial of a mathematics professor from Princeton.
	They had to get it done.
	No wasted time or energy.
	And so there's just a fleet of people all working as an organism.
	And it was fascinating.
	I was just the whole time just looking around like, this is so neat.
	Like one person whose job it was to take the camera off of the cameraman.
	Mm-hmm. So that someone else, whose job it was to remove the film canister, because every couple's takes, they had to replace the film because, you know, film gets used up.
	It was just...
	I don't know. I was geeking out the whole time.
	It was so fun. How many takes did it take?
	It looked the opposite.
	Like, there was more than two people there.
	It was very relaxed. Right, right.
	Yeah. I mean, the person who I was in the scene with is a professional.
	She's, you know...
	She's an actress? Improv comedian from New York City.
	And when I got there, they had given me a script, such as it was.
	And then I got there and they said, we're going to do this as improv.
	I'm like, I don't know how to improv.
	Like, this is not, I don't know what this, I don't know what you're telling me to do here.
	Don't worry, she knows. I'm like, okay, we'll see how this goes.
	I guess I got pulled into the story because, like, where the heck did you come from?
	I guess in the scene.
	Like, how did you show up in this random person's house?
	I don't know. Yeah, well, I mean, the reality of it is I stood outside in the blazing sun.
	There was someone whose job it was to keep an umbrella over me because I started to schvitz.
	I started to sweat. And so I would wreck the shot because my face was all shiny with sweat.
	So there was one person who would dab me off, had an umbrella.
	But yeah, like, the reality of it, like, why is this strange stalkery person hanging around outside somebody's house?
	We're not sure.
	We'll have to wait for the book.
	So you make, like you said, YouTube.
	You make videos yourself.
	You make awesome parody songs that kind of focus in on a particular aspect of computer science.
	How much... Those seem really natural.
	How much Production Valley goes into that?
	Do you also have a team of 50 people?
	The videos, almost all the videos, except for the ones that people would have actually seen, are just me.
	I write the lyrics. I sing the song.
	I generally find a...
	Like a backing track online because I'm like you, can't really play an instrument.
	And then I do, in some cases, I'll do visuals using just like PowerPoint.
	Lots and lots of PowerPoint to make it sort of like an animation.
	The most produced one is the one that people might have seen, which is the overfitting video that I did with Charles Isbell.
	And that was produced by the Georgia Tech and Udacity people because we were doing a class together.
	I usually do parody songs kind of to cap off a class at the end of a class.
	So that one you're wearing, so it was just a thriller.
	Yeah. You're wearing the Michael Jackson, the red leather jacket.
	The interesting thing with podcasting that you're also into is that I really enjoy...
	Is that there's not a team of people.
	It's kind of more...
	Because, you know, there's something that happens when there's more people involved than just one person, that just the way you start acting...
	I don't know. There's a censorship.
	You're not given, especially for slow thinkers like me, you're not, and I think most of us are, if we're trying to actually think, we're a little bit slow and careful.
	It kind of, large teams get in the way of that.
	And I don't know what to do with it.
	To me, it's very popular to criticize, quote-unquote, mainstream media.
	But there is legitimacy to criticizing them.
	I love listening to NPR, for example.
	But it's clear that there's a team behind it.
	There's constant commercial breaks.
	There's this kind of rush of like, okay, I have to interrupt you now because we have to go to commercial.
	It creates...
	It destroys the possibility of nuanced conversation.
	Yeah, exactly. Evian, which Charles Isbell, who I talked to yesterday, told me that Evian is naive backwards, which the fact that his mind thinks this way is just quite brilliant.
	Anyway, there's a freedom to this podcast.
	He's Dr. Awkward, which, by the way, is a palindrome.
	That's a palindrome that I happen to know from other parts of my life, and I just figured I'd use it against Charles.
	Dr. Awkward. So what was the most challenging parody song to make?
	Was it the Thriller one? No, that was really fun.
	I wrote the lyrics really quickly, and then I gave it over to the production team.
	They recruited an acapella group to sing.
	It went really smoothly.
	It's great having a team, because then you can just focus on the part that you really love, which in my case is writing the lyrics.
	Yeah. For me, the most challenging one, not challenging in a bad way, but challenging in a really fun way, was one of the parody songs I did is about the halting problem in computer science.
	The fact that you can't create a program that can tell for any other arbitrary program whether it's actually going to get stuck in an infinite loop or whether it's going to eventually stop.
	Yes. And so I did it to an 80s song because I hadn't started my new thing of learning current songs.
	And it was Billy Joel's The Piano Man.
	Nice. Which is a great song.
	Great song. Yeah, yeah. Sing Me a Song, The Piano Man.
	Yeah, it's a great song. So the lyrics are great because, first of all, it rhymes.
	Not all songs rhyme. I've done Rolling Stones songs, which turn out to have no rhyme scheme whatsoever.
	They're just sort of yelling and having a good time.
	Which makes it not fun from a parody perspective, because you can say anything.
	But the lines rhymed, and there was a lot of internal rhymes as well.
	And so figuring out how to sing with internal rhymes, a proof of the halting problem, was really challenging.
	And I really enjoyed that process.
	What about, last question on this topic, what about the dancing in the thriller video?
	How many takes does that take? So, I wasn't planning to dance.
	They had me in the studio, and they gave me the jacket, and it's like, well, you can't, if you have the jacket and the glove, like, there's not much you can do.
	So, I think I just danced around, and then they said, why don't you dance a little bit?
	There was a scene with me and Charles dancing together.
	They did not use it in the video, but we recorded it.
	Yeah, yeah, no, it was pretty funny.
	And Charles, who has this beautiful, wonderful voice, doesn't really sing.
	He's not really a singer. And so that was why I designed the song with him doing a spoken section and me doing the singing section.
	Yeah, it's very like Barry White. Yeah, just smooth baritone.
	Yeah, yeah. It's great.
	Yeah, it was awesome. So, one of the other things Charles said is that, you know, everyone knows you as like a super nice guy, super passionate about teaching and so on.
	What he said, don't know if it's true, that despite the fact that you are- Killed a man in cold blood.
	Like, okay, I will admit this, finally, for the first time, that was me.
	It's the Johnny Cash song of the Manarino just to watch him die.
	That you actually do have some strong opinions on some topics.
	So, if this, in fact, is true, what strong opinions would you say you have?
	Is there ideas... You think maybe in artificial intelligence, machine learning, maybe in life that you believe is true that others might, you know, some number of people might disagree with you on.
	So I try very hard to see things from multiple perspectives.
	There's this great Calvin and Hobbes cartoon where Calvin's dad is always kind of a bit of a foil.
	Calvin had done something wrong.
	The dad talks him into seeing it from another perspective.
	This breaks Calvin because he's like, oh my gosh, now I can see the opposite sides of things.
	It becomes like a cubist cartoon where there is no front and back.
	Everything's just exposed. And it really freaks him out.
	And finally, he settles back down.
	It's like, oh, good. No, I can make that go away.
	But I'm that. I live in that world where I'm trying to see everything from every perspective all the time.
	So there are some things that I've formed opinions about that would be harder, I think, to disavow me of.
	One is the superintelligence argument and the existential threat of AI is one where I feel pretty confident in my feeling about that one.
	Like, I'm willing to hear other arguments, but like, I am not particularly moved by the idea that if we're not careful, we will accidentally create a superintelligence that will destroy human life.
	Let's talk about that. Let's get you in trouble and record your own video.
	It's like Bill Gates, I think he said some quote about the internet, that that's just going to be a small thing.
	It's not going to really go anywhere.
	And I think Steve Ballmer said, I don't know why I'm sticking on Microsoft, that's something that smartphones are useless.
	There's no reason why Microsoft should get into smartphones, that kind of...
	So, let's talk about AGI. As AGI is destroying the world, we'll look back at this video and see.
	Now, I think it's really interesting to actually talk about it because nobody really knows the future, so you have to use your best intuition.
	It's very difficult to predict it.
	But you have spoken about AGI and the existential risks around it and sort of based on your intuition that...
	We're quite far away from that being a serious concern relative to the other concerns we have.
	Can you maybe unpack that a little bit?
	Yeah, sure, sure, sure.
	So as I understand it, for example, I read Bostrom's book and a bunch of other reading material about this sort of general way of thinking about the world.
	And I think the story goes something like this, that we will at some point create computers that are smart enough that they can help design the next version of themselves and Which itself will be smarter than the previous version of themselves and eventually bootstrapped up to being smarter than us, at which point we are essentially at the mercy of this sort of more powerful intellect, which in principle...
	over what its goals are. And so if its goals are at all out of sync with our goals, like
	for example, the continued existence of humanity, we won't be able to stop it. It'll be
	way more powerful than us and we will be toast. So there's some, I don't know, very smart people
	who have signed on to that story. And it's a compelling story. I want to, now I can really
	I once wrote an op-ed about this, specifically responding to some quotes from Elon Musk, who has been on this very podcast more than once.
	And- AI summoning the demon.
	That's a thing he said.
	But then he came to Providence, Rhode Island, which is where I live, and said to the governors of all the states, you're worried about entirely the wrong thing.
	You need to be worried about AI. You need to be very, very worried about AI. So, and journalists kind of reacted to that, and they wanted to get people's take, and I was like, okay, my belief is that one of the things that makes Elon Musk so successful and so remarkable as an individual— He believes in the power of ideas.
	He believes that if you have a really good idea for getting into space, you can get into space.
	If you have a really good idea for a company or for how to change the way that people drive, you just have to do it, and it can happen.
	It's really natural to apply that same idea to AI. You see these systems that are doing some pretty remarkable computational tricks, demonstrations, and then to take that idea and just push it all the way to the limit and think, okay, where does this go?
	Where is this going to take us next?
	And if you're a deep believer in the power of ideas, then it's really natural to believe that those ideas could be taken to the extreme and kill us.
	So I think his strength is also his undoing because that doesn't mean it's true.
	It doesn't mean that that has to happen, but it's natural for him to think that.
	So another way to phrase the way he thinks, and I find it very difficult to argue with that line of thinking.
	So Sam Harris is another person from a neuroscience perspective that thinks like that, is saying, well, is there something fundamental in the physics of the universe that prevents this from eventually happening?
	And Nick Bostrom thinks in the same way.
	They're kind of zooming out, yeah, okay, we humans now are existing in this time scale of minutes and days, and so our intuition is in this time scale of minutes, hours, and days.
	But if you look at the span of human history, is there any reason you can't see this in 100 years?
	Is there something fundamental about the laws of physics that prevent this?
	And if it doesn't, then it eventually will happen.
	Or we will destroy ourselves in some other way.
	And it's very difficult, I find, to actually argue against that.
	Yeah. Me too.
	And not sound like you're just rolling your ass.
	In science fiction, we don't have to think about it.
	But even worse than that, which is, like, I don't have kids, but, like, I gotta pick up my kids now.
	Like, okay. I see. There's more pressing short-term.
	Yeah, there's more pressing short-term things that, like, stop it with this existential crisis.
	We have much shorter things.
	Like, now, especially this year, there's COVID. So, like, any kind of discussion like that is...
	Like there's, you know, there's pressing things today.
	And then so the Sam Harris argument, well, like any day the exponential singularity can occur is very difficult to argue against.
	I mean, I don't know. But part of his story is also he's not going to put a date on it.
	It could be in a thousand years.
	It could be in a hundred years. It could be in two years.
	It's just that as long as we keep making this kind of progress, it's ultimately has to become a concern.
	I kind of am on board with that, but the piece that I feel like is missing from that way of extrapolating from the moment that we're in is that I believe that in the process of actually developing technology that can really get around in the world and really process and do things in the world in a sophisticated way, we're going to learn a lot about what that means, which we don't know now because we don't know how to do this right now.
	can just turn on a deep learning network and it eventually give it enough compute and it'll
	eventually get there.
	Well, sure, that seems really scary because we won't we won't be in the loop at all.
	We won't we won't be helping to design or target these kinds of systems.
	But I don't I don't see that.
	That feels like it is against the laws of physics because these systems need help, right?
	They need to surpass the difficulty, the wall of complexity that happens in arranging something in the form that that will happen in.
	Yeah. Like, I believe in evolution.
	Like, I believe that there's an argument, right?
	So there's another argument, just to look at it from a different perspective, that people say, well, I don't believe in evolution.
	How could evolution—it's sort of like a random set of parts assemble themselves into a 747, and that could just never happen.
	Yeah. So it's like, okay, that's maybe hard to argue against, but clearly 747s do get assembled.
	They get assembled by us.
	Basically, the idea being that there's a process by which we will get to the point of making technology that has that kind of awareness.
	And in that process, we're going to learn a lot about that process, and we'll have more ability to control it or to shape it or to...
	It's not something that is going to spring into existence like that 747 and we're just going to have to contend with it completely unprepared.
	It's very possible that in the context of the long arc of human history, it will in fact spring into existence.
	But that springing might take, like if you look at nuclear weapons, like even 20 years is a springing in the context of human history.
	And it's very possible, just like with nuclear weapons, that we could have, I don't know what percentage you want to put at it, but the possibility of- Could have knocked ourselves out.
	Yeah, the possibility of human beings destroying themselves in the 20th century with nuclear weapons, I don't know, if you really think through it, you could really put it close to, I don't know, 30%, 40%, given the certain moments of crisis that happen.
	So I think one Like, fear in the shadows that's not being acknowledged is it's not so much the AI will run away, is that as it's running away, we won't have enough time to think through how to stop it.
	Right. The fast takeoff or foom.
	Yeah. I mean, my much bigger concern, I wonder what you think about it, which is...
	We won't know it's happening.
	So I kind of think that there's an AGI situation already happening with social media, that our minds, our collective intelligence of human civilization is already being controlled by an algorithm.
	And we're already super...
	The level of a collective intelligence, thanks to Wikipedia, people should donate to Wikipedia to feed the AGI, Man, if we had a superintelligence that was in line with Wikipedia's values, it's a lot better than a lot of other things I could imagine.
	I trust Wikipedia more than I trust Facebook or YouTube as far as trying to do the right thing from a rational perspective.
	Now, that's not where you were going, and I understand that.
	But it does strike me that there's sort of smarter and less smart ways of exposing ourselves to each other on the internet.
	Yeah, the interesting thing is that Wikipedia and social media are very different forces.
	You're right. I mean, Wikipedia, if AGI was Wikipedia, you'd be just like this cranky, overly competent editor of articles.
	You know, there's something to that.
	But the social media aspect is...
	So the vision of AGI is as a separate system that's super intelligent.
	That's super intelligent, that's one key little thing.
	I mean, there's the paperclip argument that's super dumb, but super powerful systems.
	But with social media, you have a relatively, like algorithms we may talk about today, very simple algorithms that when, so something Charles talks a lot about, which is interactive AI, when they start like, Having at scale, like tiny little interactions with human beings, they can start controlling these human beings.
	So a single algorithm can control the minds of human beings slowly to where we might not realize it could start wars, it could start, it could change the way we think about things.
	It feels like In the long arc of history, if I were to sort of zoom out from all the outrage and all the tension on social media, that it's progressing us towards better and better things.
	It feels like chaos and toxic and all that kind of stuff.
	It's chaos and toxic, yeah.
	But it feels like actually the chaos and toxic is similar to the kind of debates we had from the founding of this country.
	There was a civil war that happened over that period.
	And ultimately it was all about this tension of something doesn't feel right about our implementation of the core values we hold as human beings and they're constantly struggling with this.
	And that results in people calling each other, just being shitty to each other on Twitter.
	But, ultimately, the algorithm is managing all that, and it feels like there's a possible future in which that algorithm controls us into the direction of self-destruction, whatever that looks like.
	Yeah, so, alright, I do believe in the power of social media to screw us up royally.
	I do believe in the power of social media to benefit us, too.
	I do think that we're in a...
	Yeah, it sort of almost got dropped on top of us, and now we're trying to, as a culture, figure out how to cope with it.
	There's a sense in which—I don't know, there's some arguments that say that, for example, I guess college-age students now, late college-age students now, people who were in middle school when social media started to really take off— Maybe really damaged.
	This may have really hurt their development in a way that we don't have all the implications of quite yet.
	That's the generation who, and I hate to make it somebody else's responsibility, but they're the ones who can fix it.
	They're the ones who can figure out, how do we keep the good of this kind of technology without letting it eat us alive?
	And if they're successful, We move on to the next phase, the next level of the game.
	If they're not successful, then we're going to wreck each other.
	We're going to destroy society.
	So you're going to, in your old age, sit on a porch and watch the world burn because of the TikTok generation?
	I believe... Well, so this is my kid's age, right?
	And certainly my daughter's age.
	And she's very tapped in to social stuff.
	But she's also... She's trying to find that balance, right, of participating in it and in getting the positives of it, but without letting it eat her alive.
	And I think sometimes she ventures—I hope she doesn't watch this—sometimes I think she ventures a little too far and is consumed by it, and other times she gets a little distance.
	And if, you know, if there's enough people like her out there, they're going to navigate this choppy waters.
	That's an interesting skill, actually, to develop.
	I talked to my dad about it.
	You know, I've now somehow, this podcast in particular, but other reasons, has received a little bit of attention.
	And with that, apparently, in this world, even though I don't shut up about love and I'm just all about kindness, I have now a little mini army of trolls.
	It's kind of hilarious, actually, but it also doesn't feel good.
	But it's a skill to learn to not look at that.
	Like, to moderate, actually, how much you look at that.
	The discussion I have with my dad, it's similar to, it doesn't have to be about trolls, it could be about checking email, which is...
	Like, if you're anticipating, you know, there's, my dad runs a large institute at Drexel University, and there could be stressful, like, emails you're waiting.
	Like, there's drama of some kinds.
	And so, like, there's a temptation to check the email.
	If you send an email and you kind of, and that pulls you in into, it doesn't feel good.
	And it's a skill that he actually complains that he hasn't learned, I mean, he grew up without it, so he hasn't learned the skill of how to shut off the internet and walk away.
	And I think young people, while they're also being quote unquote damaged by like, You know, being bullied online, all of those stories, which are very horrific.
	You basically can't escape your bullies these days when you're growing up.
	But at the same time, they're also learning that skill of how to be able to shut off the disconnect, be able to laugh at it, not take it too seriously.
	It's fascinating. We're all trying to figure this out.
	Just like you said, it's been dropped on us and we're trying to figure it out.
	Yeah, I think that's really interesting. And I guess I've become a believer in the human design, which I feel like I don't completely understand.
	Like, how do you make something as robust as us?
	Like, we're so flawed in so many ways.
	And yet, and yet...
	We dominate the planet and we do seem to manage to get ourselves out of scrapes eventually.
	Not necessarily the most elegant possible way, but somehow we get to the next step.
	And I don't know how I'd make a machine do that.
	Generally speaking, like if I train one of my reinforcement learning agents to play a video game and it works really hard on that first stage over and over and over again and it makes it through, it succeeds on that first level.
	And then the new level comes and it's just like, okay, I'm back to the drawing board.
	And somehow humanity, we keep leveling up and then somehow managing to put together the skills necessary to achieve success, some semblance of success in that next level too.
	And, you know, I hope we can keep doing that.
	You mentioned reinforcement learning, so you have a couple years in the field.
	No, quite a few.
	Quite a long career in artificial intelligence broadly, but reinforcement learning specifically.
	Can you maybe give a hint about your sense of the history of the field?
	And in some ways it's changed with the advent of deep learning, but as long roots, like how is it weaved in and out of your own life?
	How have you seen the community change or maybe the ideas that it's playing with change?
	I've had the privilege, the pleasure of having almost a front row seat to a lot of this stuff, and it's been really, really fun and interesting.
	So when I was in college in the 80s, early 80s, the neural net Thing was starting to happen, and I was taking a lot of psychology classes and a lot of computer science classes as a college student, and I thought, you know, something that can play tic-tac-toe and just, like, learn to get better at it, that ought to be a really easy thing.
	So I spent almost all of my, what would have been vacations during college, like, hacking on my home computer, trying to teach it how to play tic-tac-toe.
	Or programming language. Basic, oh yeah.
	Basic. That's my first language, that's my native language.
	Is that when you first fell in love with computer science?
	Just like programming basic on that?
	What was the computer?
	Do you remember? I had a TRS-80 Model 1 before they were called Model 1s because there was nothing else.
	I got my computer in 1979?
	I would have been bar mitzvahed, but instead of having a big party that my parents threw on my behalf, they just got me a computer, because that's what I really, really, really wanted.
	I saw them in the mall in Radarshack, and I thought, what, how are they doing that?
	I would try to stump them.
	I would give them math problems like one plus, and then in parentheses, two plus one.
	And I would always get it right.
	I'm like, how do you know so much?
	Like, I've had to go to algebra class for the last few years to learn this stuff,
	and you just seem to know.
	So I was smitten, and I got a computer.
	And I think ages 13 to 15, Thank you.
	I have no memory of those years.
	I think I just was in my room with the computer.
	Listening to Billy Joel. Communing, possibly listening to the radio, listening to Billy Joel.
	That was the one album I had on vinyl at that time.
	And then I got it on cassette tape and that was really helpful because then I could play it.
	I didn't have to go down to my parents' Wi-Fi or Hi-Fi, sorry.
	And at age 15, I remember kind of walking out and like, okay, I'm ready to talk to people again.
	Like, I've learned what I need to learn here.
	And so, yeah, so that was my home computer.
	And so I went to college and I was like, oh, I'm totally going to study computer science.
	I opted—the college I chose specifically had a computer science major.
	The one that I really wanted—the college I really wanted to go to didn't, so— Bye-bye to them.
	Which college did you go to? So I went to Yale.
	Princeton would have been way more convenient, and it was just a beautiful campus, and it was close enough to home, and I was really excited about Princeton, and I visited.
	I said, so, computer science major?
	They're like, well, we have computer engineering.
	I'm like, oh, I don't like that word, engineering.
	I like computer science.
	I really, I want to do, like, you're saying hardware and software?
	They're like, yeah. I'm like, I just want to do software.
	I couldn't care less about hardware.
	And you grew up in Philadelphia? I grew up outside Philly, yeah.
	So the local schools were like Penn and Drexel and Temple.
	Like everyone in my family went to Temple at least at one point in their lives except for me.
	So yeah, Philly family.
	Yale had a computer science department and that's when you...
	It's kind of interesting you said 80s and neural networks.
	That's when neural networks was a hot new thing or a hot thing, period.
	So is that in college when you first learned about neural networks?
	Yeah. And it was in a psychology class, not in a CS class.
	Oh, wow. Was it psychology or cognitive science?
	Do you remember what context?
	Yeah, yeah, yeah. So I've always been a bit of a cognitive psychology groupie.
	So I study computer science, but I like to hang around where the cognitive scientists are because, I don't know, brains, man, they're wacky.
	Cool. And they have a bigger picture view of things.
	They're a little less engineer-y, I would say.
	They're more interested in the nature of cognition and intelligence and perception and how the vision system works.
	They're asking always bigger questions.
	Now, with the deep learning community, there's a lot of intersections, but I do find that the neuroscience folks, actually, and cognitive psychology, cognitive science folks, Are starting to learn how to program, how to use artificial neural networks.
	And they are actually approaching problems in totally new, interesting ways.
	It's fun to watch grad students from those departments approach a problem of machine learning.
	Right. They come in with a different perspective.
	Yeah, they don't care about your ImageNet data set or whatever.
	They want to understand the basic mechanisms at the neuronal level, at the functional level of intelligence.
	It's kind of cool to see them work.
	But yeah, okay, so you're always a groupie of cognitive psychology.
	Yeah, yeah. And so it was in a class by Richard Gehrig.
	He was kind of my favorite psych professor in college, and I took like three different classes with him.
	And yeah, so they were talking.
	Specifically, the class I think was kind of a...
	There was a big paper that was written by Steven Pinker and Prince.
	I'm blanking on Prince's first name, but Pinker and Prince.
	They wrote kind of a...
	They were at that time kind of like...
	I'm blanking on the names of the current people.
	The cognitive scientists who are complaining a lot about deep networks.
	Oh, Gary.
	Gary Marcus. Gary Marcus.
	And who else?
	I mean, there's a few, but Gary is the most feisty.
	Sure. Gary's very feisty.
	And with his co-author, they're kind of doing these kind of takedowns where they say, okay, well, yeah, it does all these amazing things, but here's a shortcoming, here's a shortcoming, here's a shortcoming.
	And so the Pinker-Prince paper is kind of like that generation's version of Marcus and Davis, right?
	Where they're trained as cognitive scientists, but they're looking skeptically at the results
	in the artificial intelligence neural net kind of world and saying, yeah, it can do this and this and this,
	but like, it can't do that, and it can't do that, and it can't do that.
	Maybe in principle or maybe just in practice at this point, but the fact of the matter is,
	you've narrowed your focus too far to be impressed.
	You're impressed with the things within that circle, but you need to broaden that circle a little bit.
	You need to look at a wider set of problems.
	And so I was in this seminar in college that was basically a close reading of the Pinker Prince paper, which was like really thick.
	There was a lot going on in there.
	And it talked about the reinforcement learning idea a little bit.
	I'm like, oh, that sounds really cool because behavior is what is really interesting to me about psychology anyway.
	So, making programs that, I mean, programs are things that behave.
	People are things that behave.
	Like, I want to make learning that learns to behave.
	In which way was reinforcement learning presented?
	Is this talking about human and animal behavior, or are we talking about actual mathematical constructs?
	Right, so that's a good question.
	I think it wasn't actually talked about as behavior in the paper that I was reading.
	I think that it just talked about learning.
	And to me, learning is about learning to behave.
	But really, neural nets at that point were about learning, like supervised learning, so learning to produce outputs from inputs.
	So I kind of tried to invent reinforcement learning.
	When I graduated, I joined a research group at Bellcore, which had spun out of Bell Labs recently at that time because of the divestiture of long distance and local phone service in the 1980s, 1984.
	And I was in a group with Dave Ackley, who was the first author of the Boltzmann machine paper, so the very first neural net paper that could handle XOR, right?
	So XOR sort of killed neural nets, the very first, the zero-width order neural nets.
	Yeah. The Perceptrons paper, and Hinton, along with his student Dave Ackley, and I think there was other authors as well, showed that, no, no, no, with Boltz machines, we can actually learn nonlinear concepts.
	And so everything's back on the table again.
	And that kind of started that second wave of neural networks.
	So Dave Ackley, he became my mentor at Bellcore, and we- I've talked a lot about learning and life and computation and how all these things fit together.
	Now Dave and I have a podcast together.
	So I get to kind of enjoy that sort of his perspective once again, even all these years later.
	And so I said, so I said, I was really interested in learning, but in the concept of behavior.
	And he's like, oh, well, that's reinforcement learning here.
	And he gave me Rich Sutton's 1984 TD paper.
	So I read that paper.
	I honestly didn't get all of it, but I got the idea.
	I got that they were using, that he was using ideas that I was familiar with in the context of neural nets and like sort of backprop.
	Yeah.
	Yeah. Yeah.
	Yeah. And so I was, you know, my mind was blown.
	And so Rich came and he gave a talk at Belcor and he talked about what he was super excited, which was they had just figured out at the time, Q learning.
	So Watkins had visited the Rich Sutton's lab at UMass or Andy Bartow's lab that Rich was a part of.
	And he was really excited about this because it resolved a whole bunch of problems that he didn't know how to resolve in the earlier paper.
	And so...
	For people who don't know, TD, temporal difference, these are all just algorithms for reinforcement learning.
	Right. And TD, temporal difference in particular, is about making predictions over time.
	And you can try to use it for making decisions, right?
	Because if you can predict how good an action...
	Outcomes will be in the future.
	You can choose one that has better.
	But the theory didn't really support changing your behavior.
	The predictions had to be of a consistent process if you really wanted it to work.
	And one of the things that was really cool about Q-learning, another algorithm for reinforcement learning, is it was off-policy, which meant that you could actually be learning about the environment and what the value of different actions would be while actually figuring out how to behave optimally.
	So that was a revelation.
	Yeah, and the proof of that is kind of interesting.
	I mean, that's really surprising to me when I first read that in Rich Sutton's book on the matter.
	It's kind of beautiful that a single equation can capture all— One equation, one line of code, and you can learn anything.
	So equation and code, you're right.
	The code that you can arguably, at least if you squint your eyes, can say this is all of intelligence.
	You can implement that in a single...
	I think I started with Lisp, which is...
	Shout out to Lisp.
	Like a single line of code, key piece of code, maybe a couple...
	That you could do that. It's kind of magical.
	It feels so good to be true.
	Well, and it sort of is.
	Yeah. It seems they require an awful lot of extra stuff supporting it.
	But nonetheless, the idea is really good.
	And as far as we know, it is a very reasonable way of trying to create adaptive behavior, behavior that gets better at something over time.
	Did you find the idea of optimal at all compelling, that you could prove that it's optimal?
	So like one part of computer science that it makes people feel warm and fuzzy inside is when you can prove something like that a sorting algorithm, worst case, runs in n log n and it makes everybody feel so good.
	Even though in reality it doesn't really matter what the worst case is, what matters is like, does this thing actually work?
	In practice on this particular actual set of data that I enjoy.
	Did you? So here's a place where I have maybe a strong opinion, which is like, you're right, of course, but no, no.
	So what makes worst case so great, if you have a worst case analysis so great, is that you get modularity.
	You can take that thing and plug it into another thing and still have some understanding of what's gonna happen when you click them together.
	If it just works well in practice, in other words, with respect to some distribution that you care about, when you go plug it into another thing, that distribution can shift.
	It can change, and your thing may not work well anymore.
	And you want it to, and you wish it does, and you hope that it will, but it might not, and then, ah!
	So you're saying you don't like machine learning?
	But we have some positive theoretical results for these things.
	You can come back at me with, yeah, but they're really weak.
	And yeah, they're really weak.
	And you can even say that sorting algorithms, like if you do the optimal sorting algorithm, it's not really the one that you want.
	And that might be true as well.
	But it is. The modularity is a really powerful statement.
	I really like that. As an engineer, you can then assemble different things.
	You can count on them to be...
	I mean, it's interesting.
	It's a balance with everything else in life.
	You don't want to get too obsessed.
	I mean, this is what computer scientists do, which they tend to get...
	They're obsessed and they over-optimize things, or they start by optimizing them, they over-optimize.
	So it's easy to get really granular about this thing, but the step from an N squared to an N log N sorting algorithm is a big leap for most real-world systems, no matter what the actual Behavior of the system is, that's a big leap. And the same can probably be said for other kind of first leaps that you would take on a particular problem.
	Like, it's picking the low-hanging fruit, or whatever the equivalent of doing not the dumbest thing, but the next to the dumbest thing.
	I see, picking the most delicious, reachable fruit.
	Yeah, most delicious, reachable fruit.
	I don't know why that's not a saying.
	Yeah. Okay, so you, then this is the 80s, and this kind of idea starts to percolate of learning.
	Yeah, well, at that point, I got to meet Rich Sutton, so everything was sort of downhill from there, and that was really the pinnacle of everything.
	But then I felt like I was kind of on the inside.
	So then as interesting results were happening, I could like check in with Rich or with Jerry Tesaro, who had a huge impact on kind of early thinking in temporal difference learning and reinforcement learning and showed that you could solve problems that we didn't know how to solve any other way.
	And so that was really cool.
	So as good things were happening, I would hear about it from either the people who were doing it or the people who were talking to the people who were doing it.
	And so I was able to track things pretty well through the 90s.
	So wasn't most of the excitement on reinforcement learning in the 90s era with, what is it, TD Gamma?
	Yeah. What's the role of these kind of little, like, fun game-playing things and breakthroughs about, you know, exciting the community?
	Was that, like, what were your, because you've also built a cross, or part of building a cross, Puzzle solver, solving program called Proverb.
	So you were interested in this as a problem, like, in forming, in using games to understand how to build intelligent systems.
	So, like, what did you think about TD Gamble?
	Like, what did you think about that whole thing in the 90s?
	Yeah, I mean, I found the TD Gammon result really just remarkable.
	So I had known about some of Jerry's stuff before he did TD Gammon.
	he did a system, just more vanilla, well, not entirely vanilla,
	but a more classical back-proppy kind of network for playing backgammon,
	where he was training it on expert moves.
	So it was kind of supervised.
	But the way that it worked was not to mimic the actions, but to learn internally an evaluation function.
	So to learn, well, if the expert chose this over this, that must mean that the expert values this more than this.
	And so let me adjust my weights to make it so that the network evaluates this
	as being better than this.
	So it could learn from human preferences.
	It could learn its own preferences.
	And then when he took the step from that to actually doing it as a full-on reinforcement learning problem where you didn't need a trainer, you could just let it play, that was remarkable, right?
	And so I think as...
	and in the recent past as well, people extrapolate.
	It's like, oh, well, if you can do that, which is obviously very hard,
	then obviously you could do all these other problems that we wanna solve that we know are also really hard.
	And it turned out very few of them ended up being practical, partly because I think neural nets,
	certainly at the time, were struggling to be consistent and reliable.
	And so training them in a reinforcement learning setting was a bit of a mess.
	I had, I don't know, generation after generation of master students who wanted to do...
	Value function approximation, basically reinforcement learning with neural nets.
	And over and over and over again, we were failing.
	We couldn't get the good results that Jerry Tesaro got.
	I now believe that Jerry is a neural net whisperer.
	He has a particular...
	Ability to get neural networks to do things that other people would find impossible.
	And it's not the technology, it's the technology and Jerry together.
	Yeah, which I think speaks to the role of the human expert in the process of machine learning.
	Right. It's so easy. We're so drawn to the idea that it's the technology that is where the power is coming from that I think we lose sight of the fact that sometimes you need a really good – just like – I mean, no one would think, hey, here's this great piece of software.
	Here's like, I don't know, GNU Emacs or whatever.
	Yeah. And doesn't that prove that computers are super powerful and basically going to take over the world?
	It's like, no, Stallman is a hell of a hacker, right?
	So he was able to make the code do these amazing things.
	He couldn't have done it without the computer, but the computer couldn't have done it without him.
	And so I think people discount the role of people like Jerry who have just a particular set of skills, right?
	On that topic, by the way, as a small side note, I tweeted, Emacs is greater than Vim yesterday, and deleted the tweet 10 minutes later when I realized it started a war.
	I was like, oh, I was just kidding.
	I was just being provocative.
	So people still feel passionately about that particular piece of great stuff.
	Yeah, I don't get that, because Emacs is clearly so much better.
	I don't understand...
	But, you know, why do I say that? Because, like, I spent a block of time in the 80s making my fingers know the Emacs keys, and now, like, that's part of the thought process for me.
	Like, I need to express—and if you take that—if you take my Emacs key bindings away, I become— I can't express myself.
	I'm the same way with, I don't know if you know what it is, but the Kinesis keyboard, which is this butt-shaped keyboard.
	Yes, I've seen them.
	They're very, I don't know, sexy, elegant.
	They're just beautiful. Yeah, they're gorgeous, way too expensive.
	But the problem with them, similar with Emacs, is once you learn to use it, It's harder to use other things.
	It's hard to use other things.
	There's this absurd thing where I have small, elegant, lightweight, beautiful little laptops, and I'm sitting there in a coffee shop with a giant Kinesis keyboard and a sexy little laptop.
	It's absurd. I used to feel bad about it, but at the same time, you just kind of have to...
	Sometimes it's back to the Billy Joel thing.
	You just have to... Throw that Billy Joe record and throw Taylor Swift and Justin Bieber to the wind.
	See, but I like them now because, again, I have no musical taste.
	Now that I've heard Justin Bieber enough, I really like his songs.
	And Taylor Swift, not only do I like her songs, but my daughter's convinced that she's a genius, and so now I basically am signed on to that.
	That's true. So yeah, that speaks to the, back to the robustness of the human brain.
	That speaks to the neuroplasticity that you can just, you can just like a mouse teach yourself to, or probably a dog teach yourself to enjoy Taylor Swift.
	I'll try it out. I don't know.
	I try it. You know what?
	It has to do with just like acclimation, right?
	Just like you said, a couple of weeks.
	Yeah. That's an interesting experiment.
	I'll actually try that. That wasn't the intent of the experiment.
	Just like social media, it wasn't intended as an experiment to see what we can take as a society, but it turned out that way.
	I don't think I'll be the same person on the other side of the week listening to Taylor Swift, but let's try.
	It's more compartmentalized.
	Don't be so worried. I get that you can be worried, but don't be so worried because we compartmentalize really well.
	And so it won't bleed into other parts of your life.
	You won't start, I don't know...
	Wearing red lipstick or whatever.
	Like, it's fine. It's fine.
	Change fashion and everything. It's fine.
	But you know what? The thing you have to watch out for is you'll walk into a coffee shop once we can do that again.
	And recognize the song?
	And you'll be... No, you won't know that you're singing along until everybody in the coffee shop is looking at you.
	And then you're like, that wasn't me.
	Yeah, that's the, you know, people are afraid of AGI. I'm afraid of the Taylor Swift takeover.
	Yeah, and I mean, people should know that TD Gammon was, would you call it, do you like the terminology of self-play by any chance?
	Sure. So like systems that learn by playing themselves?
	Just, I don't know if it's the best word, but...
	So what's the problem with that term?
	I don't know. Silly, it's like the Big Bang.
	It's like talking to serious physicists.
	Do you like the term Big Bang?
	When it was early, I feel like it's the early days of self-play.
	I don't know, maybe it was used previously, but I think it's been used by only a small group of people.
	Oh. And so I think we're still deciding, is this ridiculously silly name a good name for potentially one of the most important concepts in artificial intelligence?
	Well, okay, it depends how broadly you apply the term.
	So I used the term in my 1996 PhD dissertation.
	Wow, the actual terms of self-play.
	Yeah, because Tesaro's paper was something like training up an expert backgammon player through self-play.
	So I think it was in the title of his paper.
	If not in the title, it was definitely a term that he used.
	There's another term that we got from that work is rollout.
	So I don't know if you ever hear the term rollout.
	That's a backgammon term that has now applied generally in computers, well, at least in AI,
	because of TDGammon.
	Yeah.
	That's fascinating.
	So how is self-play being used now?
	And like, why is it, does it feel like a more general powerful concept?
	It's sort of the idea of, well, the machine's just gonna teach itself to be smart.
	Yeah, so that's where, maybe you can correct me, but that's where the continuation of the spirit
	and actually like literally the exact algorithms of TDGammon are applied by DeepMind and OpenAI
	to learn games that are a little bit more complex.
	That when I was learning artificial intelligence, Go was presented to me with artificial intelligence, the modern approach.
	I don't know if they explicitly pointed to Go in those books as unsolvable kind of thing, implying that these approaches hit their limit for these particular kind of games.
	Something, I don't remember if the book said it or not, but something in my head, if it was the professors, instilled in me the idea like this is the limits of artificial intelligence, of the field.
	It instilled in me the idea that if we can create a system that can solve the game of Go, we've achieved AGI. That was kind of, I didn't explicitly say this, but that was the feeling.
	And so I was one of the people that it seemed magical when A learning system was able to beat a human world champion at the game of Go.
	And even more so from that, that was AlphaGo, even more so with AlphaGo Zero, then kind of renamed and advanced into AlphaZero, beating a world champion or world-class player without any supervised learning on expert games.
	We're doing only through by playing itself.
	So that is, I don't know what to make of it.
	I think it would be interesting to hear what your opinions are on just how exciting, surprising, profound, interesting, or boring the breakthrough performance of AlphaZero was.
	Okay, so AlphaGo knocked my socks off.
	That was so remarkable.
	Which aspect of it? That they got it to work.
	That they actually were able to leverage a whole bunch of different ideas, integrate them into one giant system.
	Just the software engineering aspect of it is mind-blowing.
	I've never been a part of a program as complicated as the program that they built for that.
	And just the, you know, like Jerry Tesaro is a neural net whisperer, like, you know, David Silver is a kind of neural net whisperer, too.
	He was able to coax these networks and these new, way-out-there architectures to do these, you know, solve these problems that, as you said, you know, when we were learning from AI, no one had an idea how to make it work.
	It was remarkable that, you know, These, you know, these techniques that were so good at playing chess and that could beat the world champion in chess couldn't beat, you know, your typical Go playing teenager in Go.
	So the fact that, you know, in a very short number of years, we kind of ramped up to trouncing people in Go just blew me away.
	So you're kind of focusing on the engineering aspect, which is also very surprising.
	I mean, there's something different about large, well-funded companies.
	I mean, there's a compute aspect to it, too.
	Sure. Like, that, of course, I mean, that's similar to Deep Blue, right, with IBM. Like, there's something important to be learned and remembered about a large company taking the ideas that are already out there and investing a few million dollars into it or more.
	And so you're kind of saying the engineering is kind of fascinating, both on the...
	What AlphaGo is probably just gathering all the data, right, of the expert games, like organizing everything, actually doing distributed supervised learning and...
	To me, see, the engineering I kind of took for granted, to me, philosophically being able to persist in the face of long odds, because it feels like, for me, I would be one of the skeptical people in the room thinking that you can learn your way to beat Go.
	It sounded like, especially with David Silver, it sounded like David was not confident at all.
	It's funny how confidence works.
	Yeah. It's like you're not cocky about it, but...
	Right, because if you're cocky about it, you kind of stop and stall and don't get anywhere.
	Yeah, but there's like a hope that's unbreakable.
	Maybe that's better than confidence.
	It's a kind of wishful hope and a little dream, and you almost don't want to do anything else, so you kind of keep doing it.
	That seems to be the story.
	Yeah. But with enough skepticism that you're looking for where the problems are and fighting through them.
	Yeah. Because you know there's got to be a way out of this thing.
	Yeah. And for him, it was probably, there's a bunch of little factors that come into play.
	It's funny how these stories just all come together.
	Like everything he did in his life came into play, which is like a love for video games and also a connection to, so the 90s had to happen with TD Gammon and so on.
	Yeah. In some ways, it's surprising.
	Maybe you can provide some intuition to it that not much more than TD Gammon was done for quite a long time on the reinforcement learning front.
	Is that weird to you?
	I mean, like I said, the students who I worked with, we tried to basically apply that architecture to other problems, and we consistently failed.
	There were a couple... A couple of really nice demonstrations that ended up being in the literature.
	There was a paper about controlling elevators, right?
	where it's like, okay, can we modify the heuristic that elevators use for deciding,
	like a bank of elevators for deciding which floors we should be stopping on
	to maximize throughput, essentially?
	And you can set that up as a reinforcement learning problem,
	and you can have a neural net represent the value function so that it's taking where are all the elevators,
	where are the button pushes, this high dimensional, well, at the time,
	high dimensional input, a couple dozen dimensions, and turn that into a prediction as to,
	oh, is it gonna be better if I stop at this floor or not?
	And ultimately, it appeared as though for the standard simulation distribution
	for people trying to leave the building at the end of the day,
	that the neural net learned a better strategy than the standard one that's implemented
	in elevator controllers.
	So that was nice.
	There was some work that Satinder Singh et al.
	did on... Handoffs with cell phones, you know, deciding when should you hand off from this cell tower to this cell tower.
	Oh, okay. Communication networks, yeah.
	Yeah, and so a couple things seemed like they were really promising.
	None of them made it into production that I'm aware of.
	And neural nets as a whole started to kind of implode around then.
	And so there just wasn't a lot of air in the room for people to try to figure out, okay, how do we get this to work in the RL setting?
	And then they found their way back in 10 plus years.
	So you said AlphaGo was impressive, like it's a big spectacle.
	Is there- Right, so the AlphaZero.
	So I think I may have a slightly different opinion on this than some people.
	So I talked to Satinder Singh in particular about this.
	So Satinder was like Rich Sutton, a student of Antibarto.
	So they came out of the same lab, very influential machine learning, reinforcement learning researcher.
	Now at DeepMind, as is Rich, though different sites, the two of them.
	He's in Alberta. Rich is in Alberta, and Satinder would be in England, but I think he's in England from Michigan at the moment.
	But he was, yes, he was much more impressed with...
	AlphaGo Zero, which is, didn't get a kind of a bootstrap in the beginning with human-trained games.
	It just was purely self-play.
	Though the first one, AlphaGo, was also a tremendous amount of self-play.
	They started off, they kick-started the action network that was making decisions,
	but then they trained it for a really long time using more traditional temporal difference methods.
	So as a result, it didn't seem that different to me.
	It seems like, yeah, why wouldn't that work?
	Once it works, it works.
	But he found that removal of that extra information to be breathtaking.
	That's a game changer.
	To me, the first thing was more of a game changer.
	But the open question, I mean, I guess that's the assumption, is the expert games might contain within them a...
	A humongous amount of information.
	But we know that it went beyond that, right?
	We know that it somehow got away from that information because it was learning strategies.
	I don't think AlphaGo is just better at implementing human strategies.
	I think it actually developed its own strategies that were more effective.
	And so from that perspective, okay, well, so it made at least one...
	Quantum leap in terms of strategic knowledge.
	Okay, so now maybe it makes three.
	Like, okay. But that first one is the doozy, right?
	Getting it to work reliably and for the networks to hold on to the value well enough.
	Like, that was... That was a big step.
	Well, maybe you could speak to this on the reinforcement learning front.
	Starting from scratch and learning to do something, like the first random behavior to crappy behavior to somewhat okay behavior.
	It's not obvious to me That that's not like impossible to take those steps.
	Like if you just think about the intuition, like how the heck does random behavior become somewhat basic intelligent behavior?
	Not human level, not superhuman level, but just basic.
	But you're saying to you kind of the intuition is like, if you can go from human to superhuman level intelligence on this particular task of game playing, then you're good at taking leaps.
	So you can take many of them.
	That the system, I believe that the system can take that kind of leap.
	Yeah, and also I think that beginner knowledge in Go, like you can start to get a feel really quickly for the idea that...
	Certain parts of the being and certain parts of the board seems to be more associated with winning, right?
	Because it's not stumbling upon the concept of winning.
	It's told that it wins or that it loses.
	Well, it's self-play, so it both wins and loses.
	It's told which side won.
	And the information is kind of there to start percolating around to make a difference as to, well, these things have a better chance of helping you win and these things have a worse chance of helping you win.
	And so, you know, it can get to basic play, I think, pretty quickly.
	Then once it has basic play, well, now it's kind of forced to do some search to actually experiment with, okay, well, what gets me that next increment of improvement?
	Yeah. How far do you think, okay, this is where you kind of bring up the Elon Musk and the Sam Harris's, right?
	How far is your intuition about these kinds of self-play mechanisms being able to take us?
	Because it feels, one of the The ominous but stated calmly things that when I talked to David Silver he said is that they have not yet discovered a ceiling for AlphaZero, for example, on the game of Go or chess.
	Oh, interesting. It keeps, no matter how much the compute they throw at it, it keeps improving.
	So it's possible, it's very possible that if you throw some 10x compute that it will improve by 5x or something like that.
	And when stated calmly, it's so like, oh, yeah, I guess so.
	But then you think, well, can we potentially have continuations of Moore's Law in a totally different way, like broadly defined Moore's Law?
	Right, exponential improvement.
	Exponential improvement, like, are we going to have an alpha zero that swallows the world?
	But notice it's not getting better at...
	Other things. It's getting better at go.
	And I think that's a big leap to say, okay, well, therefore, it's better at other things.
	Well, I mean, the question is how much of the game of life can be turned into?
	Right. So that, I think, is a really good question.
	And I think that we don't, I don't think we as a, I don't know, community really know
	the answer to this.
	But so, okay, so I went to a talk by some experts on computer chess.
	So in particular, computer chess is really interesting because, you know, for, of course,
	for a thousand years, humans were the best chess-playing things on the planet.
	And then computers, like, edged ahead of the best person.
	And they've been ahead ever since.
	It's not like people have overtaken computers.
	But computers and people together have overtaken computers.
	So at least last time I checked, I don't know what the very latest is, but last time I checked that there were teams of people who could work with computer programs to defeat the best computer programs.
	In the game of Go? In the game of chess.
	In the game of chess. Right. And so using the information about how...
	These things called ELO scores, this sort of notion of how strong a player are you.
	There's kind of a range of possible scores.
	And you increment in score, basically, if you can beat another player of that lower score 62% of the time or something like that.
	Like there's some threshold of if you can somewhat consistently beat someone, then you are of a higher score than that person.
	And there's a question as to how many times can you do that in chess, right?
	And so we know that there's a range of human ability levels that cap out with the best playing humans, and the computers went a step beyond that.
	And computers and people together have not gone, I think, a full step beyond that.
	The estimates that they have is that it's starting to asymptote, that we've reached kind of the maximum, the best possible chess playing.
	And so that means that there's kind of a finite strategic depth, right?
	At some point, you just can't get any better at this game.
	Yeah, I mean, I don't...
	So, I'll actually check that.
	I think it's interesting because if you have somebody like Magnus Carlsen who's using these chess programs to train his mind, like to learn about chess.
	To become a better chess player, yeah. And so, like, that's a very interesting thing because we're not static creatures.
	We're learning together.
	I mean, just like we're talking about social networks, those algorithms are teaching us just like we're teaching those algorithms.
	So that's a fascinating thing, but I think the best chess playing programs are now better than the pairs.
	Like they have competition between pairs, but it's still, even if they weren't, it's an interesting question, where's the ceiling?
	So the David, the ominous David Silver kind of statement is like, we have not found the ceiling.
	Right. So the question is, okay, so I don't know his analysis on that.
	From talking to Go experts, the strategic depth of Go seems to be substantially greater than that of chess, that there's more kind of steps of improvement that you can make, getting better and better and better and better.
	But there's no reason to think that it's infinite.
	Infinite, yeah. And so it could be that what David is seeing is a kind of asymptoting, that you can keep getting better, but with diminishing returns.
	And at some point, you hit optimal play.
	Like, in theory, all these finite games, they're finite.
	They have an optimal strategy.
	There's a strategy that is the minimax optimal strategy.
	And so at that point...
	You can't get any better, you can't beat that strategy.
	Now that strategy may be, from an information processing perspective, intractable.
	You need, all the situations are sufficiently different that you can't compress it at all.
	It's this giant mess of hard-coded rules.
	And we can never achieve that, but that still puts a cap on how many levels of improvement
	that we can actually make.
	But the thing about self-play is if you put it, although I don't like doing that,
	in the broader category of self-supervised learning, is that it doesn't require too much or any human.
	Human labeling, yeah. Yeah, human labeling or just human effort.
	The human involvement past a certain point.
	And the same thing you could argue is true for the recent breakthroughs in natural language processing with language models.
	Oh, this is how you get to GPT-3.
	Yeah, see how that did the...
	That was a good transition, yeah.
	You're a pro. I practiced that for days leading up to this.
	But that's one of the questions is, Can we find ways to formulate problems in this world that are important to us humans, more important than the game of chess, to which self-supervised kinds of approaches could be applied?
	Whether it's self-play, for example, maybe you could think of autonomous vehicles in simulation, that kind of stuff, or just robotics applications in simulation.
	Or in the self-supervised learning where unannotated data or data that's generated by humans naturally without extra cost, like Wikipedia or like all of the internet, can be used to learn something about To create intelligent systems that do something really powerful, that pass the Turing test or that do some kind of superhuman level performance.
	So, what's your intuition, like, trying to stitch all of it together about Our discussion of AGI, the limits of self-play, and your thoughts about maybe the limits of neural networks in the context of language models.
	Is there some intuition in there that might be useful to think about?
	Yeah, yeah, yeah. So first of all, the whole transformer network Family of things is really cool.
	It's really, really cool.
	I mean, if you've ever...
	Back in the day, you played with, I don't know, Markov models for generating text, and you've seen the kind of text that they spit out, and you compare it to what's happening now.
	It's amazing.
	It's so amazing.
	Now, it doesn't take very long interacting with one of these systems before you find the holes, right?
	It's not smart in any kind of general way.
	It's really good at a bunch of things.
	And it does seem to understand a lot of the statistics of language extremely well.
	And that turns out to be very powerful.
	You can answer many questions with that.
	But it doesn't make it a good conversationalist, right?
	And it doesn't make it a good storyteller.
	It just makes it good at imitating of things that it has seen in the past.
	The exact same thing could be said by people who voting for Donald Trump about Joe Biden supporters
	and people voting for Joe Biden about Donald Trump supporters is.
	That they're not intelligent?
	They're just following the...
	Yeah, they're following things they've seen in the past.
	And it doesn't take long to find the flaws in their natural language generation abilities.
	Yes, yes. So we're being very...
	That's interesting.
	...critical of ASPs.
	Right. So I've had a similar thought, which was that the stories that GPT-3 spits out are amazing and very human-like.
	And it doesn't mean that computers are smarter than we realize, necessarily.
	It partly means that people are dumber than we realize, or that much of what we do day to day is not that deep.
	Like, we're just kind of going with the flow.
	We're saying whatever feels like the natural thing to say next.
	Not a lot of it is creative or meaningful or intentional.
	But enough is that we actually get by, right?
	We do come up with new ideas sometimes, and we do manage to talk each other into things sometimes,
	and we do sometimes vote for reasonable people sometimes.
	But it's really hard to see in the statistics because so much of what we're saying is kind of rote.
	And so our metrics that we use to measure how these systems are doing don't reveal that
	because it's in the interstices that is very hard to detect.
	But is your, do you have an intuition that with these language models, if they grow in size, it's already surprising that when you go from GPT-2 to GPT-3, that there is a noticeable improvement.
	So the question now goes back to the ominous David Silver and the ceiling.
	Right, so maybe there's just no ceiling, we just need more compute.
	I mean, okay, so now I'm speculating.
	As opposed to before when I was completely on firm ground.
	I don't believe that you can get something that really can do language and use language as a thing that doesn't interact with people.
	I think that it's not enough to just take everything that we've said written down and just say, that's enough.
	You can just learn from that and you can be intelligent.
	I think you really need to be pushed back at.
	I think that conversations, even people who are pretty smart,
	maybe the smartest thing that we know, maybe not the smartest thing we can imagine,
	but we get so much benefit out of talking to each other and interacting.
	That's presumably why you have conversations live with guests
	is that there's something in that interaction that would not be exposed by,
	oh, I'll just write you a story and then you can read it later.
	And I think because these systems are just learning from our stories,
	they're not learning from being pushed back at by us, that they're fundamentally limited
	into what they could actually become on this route.
	They have to get, you know, shut down.
	They have to have an argument with us and lose a couple times before they start to realize, oh, okay, wait, there's some nuance here that actually matters.
	Yeah, that's actually subtle sounding, but quite profound that the interaction with humans is essential.
	And the limitation within that is profound as well because the timescale, like the bandwidth at which you can really interact with humans is very low.
	So it's costly. So you can't, one of the underlying things about self-play is it has to do, you know, a very large number of interactions.
	And so you can't really deploy reinforcement learning systems into the real world to interact.
	Like you couldn't deploy a language model into the real world to interact with humans because it was just not getting enough data relative to the cost it takes to interact.
	Like the time of humans is expensive.
	Which is really interesting. That takes us back to reinforcement learning and trying to figure out if there's ways to make algorithms that are more efficient at learning.
	Keep the spirit in reinforcement learning and become more efficient.
	In some sense, that seems to be the goal.
	I'd love to hear what your thoughts are.
	I don't know if you got a chance to see a blog post called Bitter Lesson.
	Oh, yes. By Rich Sutton that makes an argument, and hopefully I can summarize it.
	Perhaps you can.
	I mean, I could try and you can correct me, which is he makes an argument that it seems if we look at the long arc of the history of the artificial intelligence field, It calls 70 years, that the algorithms from which we've seen the biggest improvements in practice are the very simple, like dumb algorithms that are able to leverage computation.
	And you just wait for the computation to improve.
	Like all the academics and so on have fun by finding little tricks and congratulate themselves on those tricks.
	And sometimes those tricks can be like big, that feel in the moment like big spikes and breakthroughs, but in reality, over the decades, it's still the same DOM algorithm that just waits for the compute to get faster and faster.
	Do you find that to be an interesting argument against the entirety of the field of machine learning?
	That's an academic discipline.
	That we're really just a subfield of computer architecture.
	We're just kind of waiting around for them to do their next thing.
	We really don't want to do hardware work.
	That's right. I really don't want to think about hardware.
	We're procrastinating. Yes, that's right.
	Just waiting for them to do their job so that we can pretend to have done ours.
	Yeah, I mean, the argument reminds me a lot of, I think it was a Fred Jelinek quote, early computational linguist, who said, you know, we're building these computational linguistic systems, and every time we fire a linguist, performance goes up by 10%, something like that.
	And so the idea of us building the knowledge in...
	In that case, he was finding it to be much less successful than get rid of the people who know about language from a kind of scholastic academic kind of perspective and replace them with more compute.
	And so I think this is kind of a modern version of that story, which is, okay, we want to do better on machine vision.
	You could build in all these...
	So motivated, part-based models that just feel like obviously the right thing that you have to have, or we can throw a lot of data at it, and guess what?
	We're doing better with a lot of data.
	So I hadn't thought about it until this moment in this way, but what I believe—well, I've thought about what I believe.
	What I believe is that, you know, compositionality and— What's the right way to say it?
	It's got – that's got to end, right?
	And there's hints now that Moore's Law is starting to feel some friction.
	It's starting to – the world is pushing back a little bit.
	One thing that I – I don't know.
	Do lots of people know this?
	I didn't know this. I was trying to write an essay.
	And yeah, Moore's Law has been amazing and it's enabled all sorts of things.
	But there's also a kind of counter-Moore's Law, which is that the development cost for each successive generation of chips also is doubling.
	So it's costing twice as much money.
	So the amount of development money per cycle or whatever is actually sort of constant.
	And at some point, we run out of money.
	Or we have to come up with an entirely different way of doing the development process.
	So... I guess I was always a bit skeptical of the, look, it's an exponential curve, therefore it has no end.
	Soon the number of people going to NeurIPS will be greater than the population of the Earth.
	That means we're going to discover life on other planets.
	No, it doesn't. It means that we're in a sigmoid curve on the front half, which looks a lot like an exponential.
	The second half is going to look a lot like diminishing returns.
	Yeah, but the interesting thing about Moore's Law, if you actually look at the technologies involved, it's hundreds, if not thousands, of S-curves stacked on top of each other.
	It's not actually an exponential curve.
	It's constant breakthroughs.
	And then what becomes useful to think about, which is exactly what you're saying, the cost of development, like the size of teams, the amount of resources that are invested in continuing to find new S-curves, new breakthroughs.
	Yeah, it's an interesting idea.
	You know, if we live in the moment, if we sit here today, it seems to be the reasonable thing to say that exponentials end.
	And yet, in the software realm, they just keep appearing to be happening.
	And it's so, I mean, it's so hard to disagree with Elon Musk on this.
	Because it, like, I've...
	I used to be one of those folks, I'm still one of those folks, I've studied autonomous vehicles, this is what I worked on, and it's like, you look at what Elon Musk is saying about autonomous vehicles, well obviously in a couple years, or in a year, or next month, We'll have fully autonomous vehicles.
	Like, there's no reason why we can't.
	Driving is pretty simple.
	Like, it's just a learning problem, and you just need to convert all the driving that we're doing into data and just having, you know, with the trains on that data.
	And, like, we use only our eyes, so you can use cameras, and you can train on it.
	And it's like, yeah.
	That should work.
	And then you put that hat, like the philosophical hat, but then you put the pragmatic hat and it's like, this is what the flaws of computer vision are.
	Like, this is what it means to train at scale.
	And then you put the...
	Human factors, the psychology hat on, which is like, it's actually driving us a lot.
	The cognitive science or cognitive, whatever the heck you call it, it's much harder to drive than we realize.
	There's a much larger number of edge cases.
	So building up an intuition around exponential is really difficult.
	And on top of that, the pandemic is making us think about exponentials Making us realize that we don't understand anything about it.
	We're not able to intuit exponentials.
	We're either ultra-terrified, some part of the population, and some part is like the opposite of whatever, carefree, and we're not managing it very well.
	Blase. Blase. Well, wow, that's...
	Is that French? I assume so, it's got an accent.
	So it's fascinating to think what the limits of this exponential It's not just Moore's Law, it's technology, how that rubs up against the bitter lesson in GPT-3 and self-play mechanisms.
	It's not obvious. I used to be much more skeptical about neural networks.
	Now I at least give a slither possibility.
	That we'll be very much surprised.
	And also caught in a way that we are not prepared for.
	Like in applications of social networks, for example.
	Because it feels like Really good transformer models that are able to do some kind of, like very good natural language generation are the same kind of models that can be used to learn human behavior and then manipulate that human behavior to gain advertiser dollars and all those kinds of things.
	Sure, for sure. If you the capitalist system and- Right, and they arguably already are manipulating human behavior.
	Yeah. But not for self-preservation, which I think is a big—that would be a big step.
	Like, if they were trying to manipulate us to convince us not to shut them off— I would be very freaked out.
	But I don't see a path to that from where we are now.
	They don't have any of those abilities.
	That's not what they're trying to do.
	They're trying to keep people on the site.
	But see, the thing is, this is the thing about life on Earth, is they might be borrowing our consciousness and sentience.
	In a sense, they do, because the creators of the algorithms have...
	If you look at our body, We're not a single organism.
	We're a huge number of organisms with tiny little motivations.
	We're built on top of each other.
	In the same sense, the AI algorithms that are, they're not like- It's a system that includes human companies and corporations, right?
	Because corporations are funny organisms in and of themselves that really do seem to have self-preservation built in.
	And I think that's at the design level.
	I think the design to have self-preservation be a focus.
	So you're right, in that broader system, That we're also a part of and can have some influence on, it is much more complicated, much more powerful.
	Yeah, I agree with that. So people really love it when I ask what three books, technical, philosophical, fiction, had a big impact on your life, maybe you can recommend.
	We went with movies. We went with Billy Joel.
	I forgot what music you recommended.
	I didn't. I just said I have no taste in music.
	I just like pop music.
	That was actually really skillful, the way you avoided that question.
	I'm going to try to do the same with the books.
	So do you have a skillful way to avoid answering the question about three books you would recommend?
	I'd like to tell you a story.
	So my first job out of college was at Belcourt.
	I mentioned that before, where I worked with Dave Ackley.
	The head of the group was a guy named Tom Landauer.
	And I don't know how well-known he's known now, but arguably he's the inventor and the first proselytizer of word embeddings.
	So they developed a system shortly before I got to the group Yeah.
	That's called latent semantic analysis that would take words of English and embed them in, you know, multi-hundred dimensional space and then use that as a way of, you know, assessing similarity and basically doing reinforcement learning.
	Sorry, not reinforcement, information retrieval, you know, sort of pre-Google information retrieval.
	And he was trained as an anthropologist, but then became a cognitive scientist.
	I was in the cognitive science research group.
	Like I said, I'm a cognitive science groupie.
	At the time, I thought I'd become a cognitive scientist, but then I realized in that group, no, I'm a computer scientist, but I'm a computer scientist who really loves to hang out with cognitive scientists.
	And he said, he studied language acquisition in particular.
	He said, you know, humans have about this number of words of vocabulary, and most of that is learned from reading.
	And I said, that can't be true, because I have a really big vocabulary, and I don't read.
	He's like, you must. I'm like, I don't think I do.
	I mean, like, stop signs. I definitely read stop signs.
	But, like, reading books is not a thing that I do a lot of.
	Do you really, though? It might be just visual.
	Maybe the red color. Do I read stop signs?
	No, it's just pattern recognition at this point.
	I don't sound it out. So now I do...
	I wonder what that...
	Oh yeah, stoptogons.
	So... That's fascinating.
	So you don't... So I don't read very...
	I mean, obviously I read and I've read plenty of books.
	But like some people, like Charles, my friend Charles and others, like a lot of people in my field, a lot of academics, like reading was really a central topic to them I'm not in development.
	And I'm not that guy.
	In fact, I used to joke that when I got into college, that it was on kind of a help out the illiterate kind of program.
	Because in my house, I wasn't a particularly bad or good reader.
	But when I got to college, I was surrounded by these people that were just voracious in their reading appetite.
	And they were like, have you read this?
	Have you read this? Have you read this?
	And I'd be like... No, I'm clearly not qualified to be at this school.
	There's no way I should be here.
	Now I've discovered books on tape, like audiobooks, and so I'm much better.
	I'm more caught up.
	I read a lot of books. There's a small tangent on that.
	It is a fascinating, open question to me on the topic of driving.
	Whether supervised learning people, machine learning people, think you have to drive to learn how to drive.
	To me, it's very possible that just by us humans, by first of all walking, but also by watching other people, not even being inside cars as a passenger, but let's say being inside the car as a passenger, but even just being a pedestrian and crossing the road, you learn so much about driving from that.
	It's very possible that you can, without ever being inside of a car, be okay at driving once you get in it.
	Or like watching a movie, for example.
	I don't know, something like that.
	Have you taught anyone to drive?
	No. Except myself.
	I have two children, and I learned a lot about car driving, because my wife doesn't want to be the one in the car while they're learning, so that's my job.
	So I sit in the passenger seat, and it's really scary.
	I have wishes to live, and they're figuring things out.
	They start off very, very much better than I imagine like a neural network would, right?
	They get that they're seeing the world.
	They get that there's a road that they're trying to be on.
	They get that there's a relationship between the angle of the steering, but it takes a while to not be very jerky.
	And so that happens pretty quickly.
	Like the ability to stay in lane at speed, that happens relatively fast.
	It's not zero shot learning, but it's pretty fast.
	The thing that's remarkably hard, and this is, I think, partly why self-driving cars are really hard, is the degree to which driving is a social interaction activity.
	And that blew me away.
	I was completely unaware of it until I watched my son learning to drive.
	And I was realizing that he was sending signals to all the cars around him.
	And in his case, he's always had social communication challenges.
	He was sending very mixed, confusing signals to the other cars, and that was causing the other cars to drive weirdly and erratically.
	And there was no question in my mind that he would have an accident because they didn't know how to read him.
	There's things you do with the speed that you drive, the positioning of your car, that you're constantly in the head of the other drivers.
	And seeing him not knowing how to do that and having to be taught explicitly, okay, you have to be thinking about what the other driver is thinking.
	Was a revelation to me.
	I was stunned.
	So creating kind of theories of mind of the other- Theories of mind of the other cars.
	Yeah.
	Yeah, which I just hadn't heard discussed in the self-driving car talks that I've been to.
	Since then, there's some people who do consider those kinds of issues, but it's way more subtle
	There's a little bit of work involved with that when you realize, like when you especially focus not on other cars, but on pedestrians, for example, it's literally staring you in the face.
	Yeah, yeah, yeah. So that when you're just like, how do I interact with pedestrians?
	You have Pedestrians, you're practically talking to an octopus at that point.
	They've got all these weird degrees of freedom.
	You don't know what they're going to do. They can turn around any second.
	But the point is, we humans know what they're going to do.
	We have a good theory of mind.
	We have a good mental model of what they're doing.
	And we have a good model of the model they have of you.
	And the model of the model of the model...
	We're able to kind of reason about this kind of, the social game of it.
	The hope is that it's quite simple, actually, that it could be learned.
	That's why I just talked to the Waymo, I don't know if you know that company, it's Google Self-Drapping Car.
	I talked to their CTO on this podcast, and I rode in their car, and it's quite aggressive, and it's quite fast, and it's good, and it feels great.
	It also, just like Tesla, Waymo made me change my mind about maybe driving is easier than I thought.
	Maybe I'm just being speciesist, human-centric.
	Maybe a...
	It's a speciest argument.
	Yeah, so I don't know. But it's fascinating to think about the same as with reading, which I think you just said.
	You avoided the question, though I still hope you answered it somewhat.
	You avoided it brilliantly.
	There's blind spots that artificial intelligence researchers have about what it actually takes to learn to solve a problem.
	That's fascinating. Have you had Anka Dragan on?
	She's one of my favorites.
	So much energy. She's amazing.
	And in particular, she thinks a lot about this kind of I know that you know that I know kind of planning.
	And the last time I spoke with her, she was very articulate about the ways in which self-driving cars are not solved.
	Like what's still really, really hard.
	But even her intuition is limited.
	Like we're all like new to this.
	So in some sense, the Elon Musk approach of being ultra confident and just like plowing- Put it out there.
	Putting it out there, like some people say it's reckless and dangerous and so on, but partly it seems to be one of the only ways to make progress in artificial intelligence.
	These are difficult things.
	Democracy is messy.
	Implementation of artificial intelligence systems in the real world is So many years ago, before self-driving cars were an actual thing you could have a discussion about, somebody asked me, like, what if we could use that robotic technology and use it to drive cars around?
	Like, aren't people gonna be killed and then it's, you know, blah, blah, blah.
	I'm like, that's not what's gonna happen, I said, with confidence, incorrectly, obviously.
	What I think is going to happen is we're going to have a lot more, like a very gradual kind of rollout where people have these cars in like closed communities, right?
	Where it's somewhat realistic, but it's still in a box, right?
	So that we can really get a sense of what are the weird things that can happen?
	How do we have to change the way we behave around these vehicles?
	Like it obviously requires a kind of co-evolution that you can't just plop them in and see what happens.
	But of course, we're basically popping them in to see what happens.
	So I was wrong, but I do think that would have been a better plan.
	So that's, but your intuition, that's funny, just zooming out and looking at the forces of capitalism, and it seems that capitalism rewards risk-takers, and rewards and punishes risk-takers, and try it out.
	The academic...
	This approach to let's try a small thing and try to understand slowly the fundamentals of the problem.
	And let's start with one, then do two, and then see that, and then do the three.
	You know, the capitalist startup entrepreneurial dream is let's build a thousand.
	Right, and 500 of them fail, but whatever, the other 500, we learn from them.
	But if you're good enough, I mean, one thing, it's like your intuition would say, like, that's gonna be hugely destructive to everything.
	But actually, it's kind of, the forces of capitalism, people are quite, it's easy to be critical, but if you actually look at the data, at the way our world has progressed in terms of the quality of life, it seems like the competent good people rise to the top.
	This is coming from me from the Soviet Union and so on.
	It's interesting that somebody like Elon Musk is the way you push progress in artificial intelligence.
	It's forcing Waymo to step their stuff up and Waymo is forcing Elon Musk to step up.
	It's fascinating because I have this tension in my heart and just being upset by The lack of progress in autonomous vehicles within academia.
	So there's huge progress in the early days of the DARPA challenges.
	And then it just kind of stopped at MIT, but it's true everywhere else, with an exception of a few sponsors here and there.
	It's not seen as a sexy problem.
	Like the moment artificial intelligence starts approaching the problems of the real world, like academics kind of like, eh, all right, let the- They get really hard in a different way.
	In a different way, that's right.
	I think, yeah, right, some of us are not excited about that But I still think there's fundamentals problems to be solved in those difficult things.
	It's still publishable, I think.
	It's the same criticism you could have of all these conferences in Europe, CVPR, where application papers are often as powerful and as important as theory paper.
	Even theory just seems much more respectable and so on.
	I mean, the machine learning community is changing that a little bit, I mean, at least in statements, but it's still not seen as the sexiest of pursuits, which is like, how do I actually make this thing work in practice as opposed to on this toy data set?
	All that to say, are you still avoiding the three books question?
	Is there something on audiobook that you can recommend?
	Oh, yeah. I mean, yeah, I've read a lot of really fun stuff.
	In terms of books that I find myself thinking back on that I read a while ago,
	like that have stood the test of time to some degree, I find myself thinking of Program or Be Programmed a lot
	by Douglas Roskopp, which was, it basically put out the premise
	that we all need to become programmers in one form or another.
	And it was an analogy to once upon a time we all had to become readers, we had to become literate.
	And there was a time before that when not everybody was literate,
	but once literacy was possible, the people who were literate had more of a say in society
	than the people who weren't.
	And so we made a big effort to get everybody up to speed, and now it's not 100% universal, but it's quite widespread.
	Like the assumption is generally that people can read.
	The analogy that he makes is that programming is a similar kind of thing,
	that we need to have a say in, right?
	So being a reader, being literate, being a reader means you can receive all this information, but you don't get to put it out there.
	And programming is the way that we get to put it out there.
	And that was the argument that he made. I think he specifically has now backed away from this idea.
	He doesn't think it's happening quite this way.
	And that might be true that it didn't, society didn't sort of play forward quite that way.
	I still believe in the premise.
	I still believe that at some point, we have the relationship that we have to these machines
	and these networks has to be one of each individual can has the wherewithal to make the machines help them.
	Do the things that that person wants done.
	And as software people, we know how to do that.
	And when we have a problem, we're like, okay, I'll just, I'll hack up a Pro Script
	or something and make it so.
	If we lived in a world where everybody could do that, that would be a better world.
	And computers would have, I think, less sway over us.
	And other people's software would have less sway over us as a group.
	Yeah, in some sense, software engineering, programming is power.
	Programming is power. Right.
	Yeah, it's like magic.
	It's like magic spells.
	And it's not out of reach of everyone.
	But at the moment, it's just a sliver of the population who can commune with machines in this way.
	So I don't know. So that book had a big, big impact on me.
	Currently, I'm reading The Alignment Problem, actually, by Brian Christian.
	So I don't know if you've seen this out there yet.
	Is it similar to Stuart Russell's work with the control problem?
	It's in that same general neighborhood.
	I mean, they have different emphases that they're concentrating on.
	I think Stuart's book did a remarkably good job, just a celebratory good job at describing AI technology and sort of how it works.
	I thought that was great.
	It was really cool to see that in a book.
	I think he has some experience writing some books.
	That's You know, that's probably a possible thing.
	He's maybe thought a thing or two about how to explain AI to people.
	Yeah, that's a really good point.
	This book so far has been remarkably good at telling the story of the history, the recent history of some of the things that have happened.
	I'm in the first third.
	He said this book is in three thirds.
	The first third is essentially AI fairness and implications of AI on society
	that we're seeing right now.
	And that's been great.
	I mean, he's telling those stories really well.
	He went out and talked to the frontline people whose names are associated with some of these ideas,
	and it's been terrific.
	He says the second half of the book is on reinforcement learning,
	so maybe that'll be fun.
	And then the third half, third third, is on the superintelligence alignment problem.
	And I suspect that that part will be less fun for me to read.
	Yeah, it's an interesting problem to talk about.
	I find it to be the most interesting, just like thinking about whether we live in a simulation or not.
	As a thought experiment to think about our own existence.
	So in the same way, talking about alignment problem with AGI is a good way to think, similar to like the trolley problem with autonomous vehicles.
	It's a useless thing for engineering, but it's a nice little thought experiment for actually thinking about What are our own human ethical systems, our moral systems?
	By thinking how we engineer these things, you start to understand yourself.
	So sci-fi can be good at that, too.
	So one sci-fi book to recommend is Exhalations by Ted Chiang, a bunch of short stories.
	Ted Chiang is the guy who wrote the short story that became the movie Arrival.
	And all of his stories, just from a—he was a computer scientist.
	Actually, he studied at Brown. They all have this sort of really insightful bit of science or computer science that drives them.
	And so it's just— A romp, right?
	He creates these artificial worlds by extrapolating on these ideas that we know about, but hadn't really thought through to this kind of conclusion.
	And so his stuff is really fun to read.
	It's mind-warping.
	So I'm not sure if you're familiar.
	I seem to mention this every other word.
	I'm from the Soviet Union, and I'm Russian.
	I read way too much to say it.
	My roots are Russian, too, but a couple generations back.
	Well, it's probably in there somewhere.
	So maybe we can pull at that thread a little bit of the existential dread that we all feel.
	I think somewhere in the conversation you mentioned that you don't really pretty much like dying.
	I forget in which context.
	It might have been a reinforcement learning perspective.
	I don't know. No, you know what it was?
	It was in teaching my kids to drive.
	That's how you face your mortality, yes.
	From a human being's perspective, or from a reinforcement learning researcher's perspective, let me ask you the most absurd question.
	What do you think is the meaning of this whole thing?
	The meaning of life on this spinning rock?
	I mean, I think reinforcement learning researchers maybe think about this from a science perspective more often than a lot of other people, right?
	As a supervised learning person, you're probably not thinking about the sweep of a lifetime, but reinforcement learning agents are having little lifetimes, little weird little lifetimes, and it's hard not to project yourself into their world sometimes.
	But as far as the meaning of life, so when I turned 42, you may know from—that is a book I read, The Hitchhiker's Guide to the Galaxy, that that is the meaning of life.
	So when I turned 42, I had a meaning of life party where I invited people over and everyone shared their meaning of life.
	We had slides made up, and so we all sat down and— Did a slide presentation to each other about the meaning of life.
	And mine... That's great.
	Mine was balance. I think that life is balance and...
	So the activity at the party, for a 42-year-old, maybe this is a little bit non-standard, but I found all the little toys and devices that I had where you had to balance on them.
	You had to stand on it and balance, or a pogo stick I brought, a rip stick, which is like a weird two-wheeled skateboard.
	I got a unicycle, but I didn't know how to do it.
	I now can do it. I'd love watching you try.
	Yeah, I'll send you a video.
	I'm not great, but I managed.
	And so balance, yeah.
	So my wife has a really good one that she sticks to and is probably pretty accurate.
	And it has to do with... Healthy relationships with people that you love and working hard for good causes.
	But to me, yeah, balance.
	Balance in a word. That works for me.
	Not too much of anything, because too much of anything is iffy.
	That feels like a Rolling Stones song.
	I feel like they must be.
	You can't always get what you want, but if you try sometimes, you can strike a balance.
	Yeah, I think that's how it goes.
	I'll write you a parody. It's a huge honor to talk to you.
	This is really fun. I've been a big fan of yours, so can't wait to see what you do next in the world of education, in the world of parody, in the world of reinforcement learning.
	Thanks for talking today. My pleasure.
	Thank you for listening to this conversation with Michael Littman, and thank you to our sponsors.
	SimpliSafe, a home security company I use to monitor and protect my apartment.
	ExpressVPN, the VPN I've used for many years to protect my privacy on the internet.
	Masterclass, online courses that I enjoy from some of the most amazing humans in history.
	And BetterHelp, online therapy with a licensed professional.
	Please check out these sponsors in the description to get a discount and to support this podcast.
	If you enjoy this thing, subscribe on YouTube, review it with Five Stars on Apple Podcasts, follow on Spotify, support on Patreon, or connect with me on Twitter at Lex Friedman.
	And now, let me leave you some words from Groucho Marx.
	If you're not having fun, you're doing something wrong.

▲