Michael I. Jordan: Machine Learning, Recommender Systems, and Future of AI | Lex Fridman Podcast #74

1. 0:00 – 3:02

   Introduction

   1. LFLex Fridman0:00

      The following is a conversation with Michael I. Jordan, a professor at Berkeley and one of the most influential people in the history of machine learning, statistics, and artificial intelligence. He has been cited over 170,000 times and he has mentored many of the world-class researchers defining the field of AI today, including Andrew Ng, Zoubin Gharahmani, Ben Taskar, and Yoshua Bengio. All this, to me, is as impressive as the over 32,000 points and the six NBA championships of the Michael J. Jordan of basketball fame. There's a non-zero probability that I talk to the other Michael Jordan, given my connection to and love of the Chicago Bulls of the '90s, but if I had to pick one, I'm going with the Michael Jordan of statistics and computer science, or as Yann LeCun calls him, the Miles Davis of machine learning. In his blog post titled, Artificial Intelligence: The Revolution Hasn't Happened Yet, Michael argues for broadening the scope of the artificial intelligence field. In many ways, the underlying spirit of this podcast is the same: to see artificial intelligence as a deeply human endeavor, to not only engineer algorithms and robots, but to understand and empower human beings at all levels of abstractions, from the individual to our civilization as a whole. This is the Artificial Intelligence Podcast. If you enjoy it, subscribe on YouTube, give us five stars on Apple Podcasts, support it on Patreon, or simply connect with me on Twitter @LexFridman, spelled F-R-I-D-M-A-N. As usual, I'll do one or two minutes of ads now and never any ads in the middle that could break the flow of the conversation. I hope that works for you and doesn't hurt the listening experience. This show is presented by Cash App, the number one finance app in the App Store. When you get it, use code LEXPODCAST. Cash App lets you send money to friends, buy Bitcoin, and invest in the stock market with as little as $1. Since Cash App does fractional share trading, let me mention that the order execution algorithm that worked behind the scenes to create the abstraction of the fractional orders is, to me, an algorithmic marvel. So big props for the Cash App engineers for solving a hard problem that, in the end, provides an easy interface that takes a step up to the next layer of abstraction over the stock market, making trading more accessible for new investors and diversification much easier. So once again, if you get Cash App from the App Store or Google Play and use the code LEXPODCAST, you'll get $10 and Cash App will also donate $10 to FIRST, one of my favorite organizations that is helping to advance robotics and STEM education for young people around the world. And now, here's my conversation with Michael I. Jordan.

2. 3:02 – 8:25

   How far are we in development of AI?

   1. LFLex Fridman3:02

      Given that you're one of the greats in the field of AI, machine learning, computer science, and so on, you're trivially called the Michael Jordan of machine learning. Although, as you know, you were born first, so technically MJ is the Michael I. Jordan of basketball. But anyway, my, my favorite is Yann LeCun calling you the Miles Davis of machine learning, because as he says, you reinvent yourself periodically and sometimes (laughs) leave fans scratching their heads after you change directions. So, can you put, at first, your historian hat on and give a history of computer science and AI as you saw it, as you experienced it, including the four generations of AI successes that I've seen you talk about?

   2. MJMichael I. Jordan3:48

      Sure. Yeah, first of all, I much prefer Yann's, uh, metaphor. Um, Miles Davis is, uh, was a real explorer in jazz and, um, he had a coherent story. So I think I have one and... But it's not just the one you lived, it's the one you think about later, right? What a good historian does is they look back and they revisit. Um, I think what's happening right now is not AI. That was an intellectual aspiration, um, that's still alive today as an aspiration. But I think this is akin to the development of chemical engineering from chemistry or electrical engineering from, from electromagnetism. So if you go back to the '30s or '40s, there wasn't yet chemical engineering. There was chemistry, there was fluid flow, there was mechanics, and so on. Um, but people pretty clearly viewed, uh, interesting goals to try to build factories that, um, you know, made chemicals, products, and do it viably, safely, make good ones, do it at scale. Uh, so people started to try to do that, of course, and some factories worked, some didn't. You know, some were not viable, some exploded. But in parallel, uh, developed a whole field called chemical engineering. Right? And chemical engineering's a field. It's, it's no, no bones about it. It has theoretical aspects to it, it has practical aspects. It's not just engineering, quote unquote, it's the real thing, and real concepts are needed. Now, same thing with electrical engineering. You know, there was Maxwell's equations, which in some sense were everything you need to know about electromagnetism, but you needed to figure out how to build circuits, how to build modules, how to put them together, how to bring electricity from one point to another safely, and so on and so forth. So a whole field was developed called electrical engineering. All right? I think that's what's happening right now, is that we have-

   3. LFLex Fridman5:28

      But-

   4. MJMichael I. Jordan5:28

      ... we have a proto-field, which is statistics, comput- more the theoretical side of, the algorithmic side of computer science. That was enough to start to build things. But what things? Systems that bring value to human beings and use human data and mix in human decisions. The engineering side of that is all ad hoc. That's what's emerging. In fact, if you wanna call machine learning a field, I think that's what it is. It's a proto-form of engineering based on statistical and computational ideas of previous generations.

   5. LFLex Fridman5:56

      But it... Do you think there's something deeper about AI in its dreams and aspirations as compared to chemical engineering and electrical engineering?

   6. MJMichael I. Jordan6:03

      No. Well, the dreams and aspirations, maybe, but those are from, those are 500 years from now. I think that that's like the Greeks sitting there and saying, "It would be neat to get to the moon someday."

   7. LFLex Fridman6:11

      Right.

   8. MJMichael I. Jordan6:12

      Um, I think we have no clue how the brain does computation. Uh, we're just clueless. We're like, we're even worse than the Greeks, uh, on most anything interesting, uh, scientifically o- of our era.

   9. LFLex Fridman6:23

      Can you linger on that just for a moment? Because you stand...... not completely unique, but a little bit unique in that, in the clarity of that. Can you, can you elaborate your intuition of why we're, like, where we stand in our understanding of the human brain? And a lot of people say, you know, scientists say we're not very far in understanding the human brain.

   10. MJMichael I. Jordan6:41

       Yeah.

   11. LFLex Fridman6:41

       But you're like, you're saying, "We're in the dark here."

   12. MJMichael I. Jordan6:44

       Well, I know I'm not unique, I don't even think in the clarity, but if you talk to real neuroscientists that really study real synapses or real neurons, they agree. They agree. It's, it's a 100 year, hundreds of year task and they're building it up slowly but surely. What the signal is there is not clear. We think we have all of our metaphors. We think it's electrical, maybe it's chemical. It's, it's a whole soup. It's ions and proteins and it's a cell. And that's even around like, a single synapse. If you look at an electromyograph of a single synapse, it's a, it's a city of its own, and that's one little thing on a dendritic tree which is extremely complicated, you know, electrochemical thing, and it's doing these spikes and voltages have been flying around, and then proteins are taking that and taking it down into the DNA and who knows what. Um, so it is the problem of the next few centuries. It is fantastic, but we have our metaphors about it. Is it an economic device? Is it like the immune system or is it like a layered, you know, set of comput- you know, arithmetic computations? What... We have all these metaphors and they're fun, um, but that's not real science, uh, per se. There is neuroscience. That's not neuroscience, all right? That- that's, that's like the Greeks speculating about how to get to the moon. Fun, right? And I think that I like to say this fairly strongly 'cause I think a lot of young people think that we're on the verge, because a lot of people who don't talk about it clearly let it be understood that yes, we kind of... This is brain-inspired, we're kind of close, you know, breakthroughs are on the horizon. And then scrupulous people sometimes who need money for their labs, um, that's why I'm saying scrupulous, but people will oversell. "Um, I need money for my lab. I'm gonna... I'm studying, you- you know, computational neuroscience, um, I'm gonna oversell it." And so there's been too much of that.

3. 8:25 – 14:49

   Neuralink and brain-computer interfaces

   1. MJMichael I. Jordan8:25

   2. LFLex Fridman8:25

      So, uh, let's step into the slight, the gray area between metaphor and engineering with, uh, I'm not sure if you're familiar with, uh, brain computer interfaces. So, a company like Elon Musk has Neuralink that's working on con- putting electrodes into, into the brain and trying to be able to read, both read and send electrical signals. Just as you said, even the basic mechanism of communication in the brain is not something we understand, but do you hope without understanding the fundamental principles of how the brain works, we'll be able to do something interesting at that gray area of metaphor and-

   3. MJMichael I. Jordan9:06

      It's not my area, so I- I hope in the sense like anybody else hopes for some interesting things to happen from research.

   4. LFLex Fridman9:12

      Yeah.

   5. MJMichael I. Jordan9:12

      I would expect more something like Alzheimer's will get figured out from modern neuroscience. That, you know, a lot of, there's a lot of human suffering based on brain disease, and we throw things like lithium at the brain. It kind of works. No one has a clue why. That's not quite true, but you know, mostly we don't know, and that's even just about the biochemistry of the brain and how it leads to mood swings and so on. How thought emerges from that, we just, we- we're really, really completely dim, so that you might want to hook up electrodes and try to do some signal processing on that and try to find patterns, fine. You know, by all means go for it. It's just not scientific at this point. It's just, it's... So it's like kind of sitting in a satellite and watching the emissions from a city and trying to infer things about the microeconomy even though you don't have microeconomic concepts. I mean, it's really that kind of thing. And, and so yes, can you find some signals that do something interesting or useful? Can you control a cursor, uh, or mouse with your brain? Yeah, absolutely. You know, and, and I can imagine business models based on that, and even, you know, medical applications of that. But from there to understanding algorithms that allow us to really tie in deeply to- from the brain to the computer, you know, I just... No, I don't agree with Elon Musk. I don't think that's even, that's not for our generations, not even for this century.

   6. LFLex Fridman10:26

      So, just, uh, in the hopes of getting you to dream, uh, you've mentioned Kolmogorov and Turing might pop up. Do you think that there might be breakthroughs that will get you to sit back in five, 10 years and say, "Wow"?

   7. MJMichael I. Jordan10:43

      Oh, I- I'm sure there will be, but I don't think that there'll be demos that impress me.

   8. LFLex Fridman10:48

      Mm-hmm.

   9. MJMichael I. Jordan10:49

      I don't think that having a computer call a restaurant and pretend to be a human is a breakthrough.

   10. LFLex Fridman10:56

       Right, right.

   11. MJMichael I. Jordan10:56

       And people, you know, some people present it as such. Uh, it's imitating human intelligence. It's even putting coughs (laughs)

   12. LFLex Fridman11:03

       Yeah.

   13. MJMichael I. Jordan11:04

       ... in the thing to make a bit of a PR stunt. And so fine, the, the world runs on those things too, uh, and I don't want to diminish all the hard work and engineering that goes behind things like that and, and the ultimate value to the human race, but that's not scientific understanding. And, and I know the people that work on these things, they are after scientific understanding. You know, in the meantime, they've got to kind of, you know, the train's got to run and they got mouths to feed and they got things to do, and there's nothing wrong with all that. Um, I would call that, though, just engineering, and I want to distinguish that between an engineering field like electrical engineering ...and chemical engineering that original- that originally emerged that had real principles and you really knew what you were doing and you had a little scientific understanding, maybe not even complete-

   14. LFLex Fridman11:43

       Mm-hmm.

   15. MJMichael I. Jordan11:43

       ... so it became more predictable and it was, really gave value to human life because it was understood. And, and so we have to, we don't want to muddle too much these waters of, uh, you know, what we're able to do versus what we really can do, uh, in a way that's gonna impress the next... So I don't, I don't need to be wowed, but I- I think that someone comes along in 20 years, a younger person who's absorbed all the, uh, the technology and- and for them to be wowed, I think they have to be more deeply impressed. A young Kolmogorov would not be wowed by some of the stunts that you see right now coming from the big companies.

   16. LFLex Fridman12:13

       The demos.

   17. MJMichael I. Jordan12:14

       The demos.

   18. LFLex Fridman12:14

       But do you think the breakthroughs from Kolmogorov-... would be, and give this question a chance, do you think they'll be in the scientific fundamental principles arena? Or do you think it's possible to have fundamental breakthroughs in engineering? Meaning, you know, I would say some of the things that Elon Musk is working with, SpaceX and then others, sort of trying to revolutionize the fundamentals of engineering, of manufacturing, of, of saying, "Here's a problem we know how to do a demo of," and actually-

   19. MJMichael I. Jordan12:43

       Yeah.

   20. LFLex Fridman12:43

       ... taking it to scale.

   21. MJMichael I. Jordan12:44

       Yeah. So, so there's gonna be all kinds of breakthroughs, I just don't like that terminology. I'm a scientist and I work on things day in and day out and things move along and then eventually say, "Wow, something happened," but it's, I don't like that language very much. Uh, also, I don't like to prize, uh, theoretical breakthroughs over practical ones. Um, I, I tend to be more of a theoretician and I think there's lots to do in, in that arena right now. Um, and so I wouldn't point to the Kolmogorovs, I might point to the Edisons of the era, and maybe Musk is a bit more like that. But, um, you know, Musk, God bless him, also will, will say things about AI that he knows very little about a- and, and he doesn't know what he... He, he, he is, you know, leads people astray when he talks about things he doesn't know anything about. Trying to program a computer to understand natural language, to be involved in a dialogue like we're having right now, ain't gonna happen in our lifetime. You could fake it. You can mimic, sort of take old sentences that humans use and retread them, but the deep understanding of language? No, it's not gonna happen. And so from that, you know, I hope you can perceive that the deeper, yet deeper kind of aspects in intelligence are not gonna happen. Now, will there be breakthroughs? You know, I think that Google was a breakthrough.

   22. LFLex Fridman13:48

       Right.

   23. MJMichael I. Jordan13:48

       I think Amazon's a breakthrough. You know? I think Uber is a breakthrough. You know, that bring value to human beings at scale in new, brand new ways based on data flows and, and so on. A lot of these things are slightly broken because there's not an em- a kind of a engineering field that takes economic value in context of data and, and that, you know, planetary scale and, and worries about all the externalities, the privacy. You know, we, we don't have that feel, so we don't think these things through very well. But I see that as emerging and that will be con- that will... You know, looking back from 100 years, that will be cons- a breakthrough in this era. Just like electrical engineering was a breakthrough in the early part of the last century and chemical engineering was a breakthrough.

   24. LFLex Fridman14:24

       So the scale, the markets that you talk about and we'll get to, uh, will be seen as sort of breakthrough. And we're in the very early days of really doing interesting stuff there. And we'll get to that, but it's... Just taking a quick step back, can you give, (laughs) uh... We'll kind of throw off the historian hat. I mean, you briefly said that, uh, im- the history of AI kind of mimics the history of chemical engineering. But-

4. 14:49 – 19:00

   The term "artificial intelligence"

   1. LFLex Fridman14:49

   2. MJMichael I. Jordan14:49

      I keep saying machine learning and you keep want to say AI. Just to let you know, I don't... You know? I, I, I resist that.

   3. LFLex Fridman14:54

      So that's-

   4. MJMichael I. Jordan14:54

      I don't think this is about... AI really was John McCarthy, as almost a philosopher-

   5. LFLex Fridman15:00

      Got it.

   6. MJMichael I. Jordan15:00

      ... saying, "Wouldn't it be cool if we could put thought in a computer? If we could mimic the human capability to think or put intelligence in in some sense into a computer?" That's an interesting philosophical question and he wanted to make it more than philosophy. He wanted to actually write down logical formula and algorithms that would do that. And that is a perfectly valid, reasonable thing to do.

   7. LFLex Fridman15:19

      Yeah.

   8. MJMichael I. Jordan15:20

      That's not what's happening in this era. Right?

   9. LFLex Fridman15:22

      So, so the reason I keep saying AI actually, and I'd love to hear what you think about it. Machine learning has, um, has a very particular set of methods and tools.

   10. MJMichael I. Jordan15:34

       Maybe your version of it is. Mine doesn't.

   11. LFLex Fridman15:36

       No, it doesn't.

   12. MJMichael I. Jordan15:36

       Mine is very, very open. It does optimization, it does sampling, it does, uh-

   13. LFLex Fridman15:40

       So systems that learn is what machine learning is?

   14. MJMichael I. Jordan15:42

       Systems that learn and make decisions.

   15. LFLex Fridman15:44

       And make decisions. So we're just-

   16. MJMichael I. Jordan15:45

       So it's not just pattern recognition and li- fr-

   17. LFLex Fridman15:47

       Got it.

   18. MJMichael I. Jordan15:47

       ... you know, finding patterns. It's all about making decisions in real worlds and having closed feedback loops.

   19. LFLex Fridman15:52

       So s- something like symbolic AI, expert systems, reasoning systems, knowledge-based representation, all of those kinds of things, search. Does that neighbor fit into what you think of as machine learning?

   20. MJMichael I. Jordan16:04

       So I don't even like the word machine learning. I think that what the field you're talking about-

   21. LFLex Fridman16:07

       Yeah.

   22. MJMichael I. Jordan16:07

       ... is all about making large collections of decisions under uncertainty by large collections of entities.

   23. LFLex Fridman16:12

       Yes.

   24. MJMichael I. Jordan16:13

       Right? And there are principles for that, at that scale. You don't have to say the principles are for a single entity that's making decisions, a single agent or a single human. It really immediately goes to the network-

   25. LFLex Fridman16:22

       Distribution.

   26. MJMichael I. Jordan16:22

       ... of decisions.

   27. LFLex Fridman16:22

       Is there a good word for that? Or not?

   28. MJMichael I. Jordan16:24

       No, there's no good words for any of this. That's kind of part of the problem. Um, so, uh, we can continue the conversation and use AI for all of that. I just want to kind of raise-

   29. LFLex Fridman16:31

       Clarify, yeah.

   30. MJMichael I. Jordan16:31

       ... the flag here that this is not about... We don't know what intelligence is. A real intelligence. We don't know much about abstraction and reasoning at the level of humans. We don't have a clue. We're not trying to build that because we don't have a clue. Eventually, it may emerge. There may... I don't know if there'll be breakthroughs, but eventually, we'll start to get glimmers of that. It's not what's happening right now though, okay? We're taking data, we're trying to make good decisions based on that. We're trying to do it at scale, we're trying to do it economically viably. We're trying to build markets, we're trying to create value at that scale. Um, and s- aspects of this will look intelligent. They will look... Computers were so dumb before, they will seem more intelligent. We will use that buzzword of intelligence, so we can use it in that sense. But, you know. So machine learning, uh, you can scope it narrowly as just learning from data and pattern recognition. But whatever I... When I talk about these topics, I... Maybe data science is another word you could throw in the mix. Um, it really is important that the decisions are i- is... As part of it, it's consequential decisions in the real world. Or am I gonna have a medical operation? Am I gonna drive down this street? You know, things that where there's scarcity, uh, things that impact other human beings or other, you know, the environment and so on. How do I do that based on data? How do I do that adaptively? How do I use computers to help those kind of things go forward? Whatever you wanna call that. So let's call it AI, let's agree to call it AI. But it's, um... Let's, let's not say that what the goal of that is is intelligence. The goal of that is really good working systems at planatory scale that we've never seen before.
5. 19:00 – 19:55

   Does science progress by ideas or personalities?

   1. LFLex Fridman19:00

      because you're one of the great personalities of machine learning, whatever the heck you call the field, the... Do you think science progresses by personalities or by the fundamental principles and theories and research that's outside of personalities?

   2. MJMichael I. Jordan19:14

      Uh, both. A- and I wouldn't say there should be one kind of personality. I have mine and I have my preferences and, uh, uh, I have a kind of network around me, uh, that feeds me. And, and some of them agree with me and some of them disagree, but, you know, all kinds of personalities are needed. Um, right now, I think the personality that is a little too exuberant, a little bit too ready to promise the moon is, is a little bit too much in ascendance.

   3. LFLex Fridman19:34

      Yeah.

   4. MJMichael I. Jordan19:35

      Um, and I do, I do think that that's... There's some good to that. It certainly attracts lots of young people to our field. But a lot of those people come in with strong misconceptions and they have to then unlearn those and then find something, you know, to do. Um, and so I think there's just gotta be some, you know, multiple voices and there's... I didn't, I wasn't hearing enough of the more sober voice.

   5. LFLex Fridman19:54

      So,

6. 19:55 – 23:53

   Disagreement with Yann LeCun

   1. LFLex Fridman19:55

      uh, as a continuation of a fun tangent, and speaking of vibrant personalities, uh, what would you say is the most interesting disagreement you have with Yann LeCun?

   2. MJMichael I. Jordan20:07

      So Yann's an old friend and I just say, say that, uh, I, I don't think we disagree about very much really.

   3. LFLex Fridman20:12

      Right.

   4. MJMichael I. Jordan20:13

      Uh, h- he and I both kind of have a let's build it kind of mentality and does it work k- kind of mentality, and, uh, kind of concrete. Um, we both speak French and we speak French when we're together and we have, we have a lot o', a lot in common. Um, and so, uh, you know, if one wanted to highlight a, uh, a disagreement, it's not really a fundamental one. I think it's just kind of what we're emphasizing. Um, Yann has, uh, emphasized pattern recognition and, uh, has emphasized prediction. All right? So, you know, um... And it's interesting to try to take that as far as you can. If you could do perfect prediction, what would that give you, kind of as a thought experiment? Um, and, um, I think that's, um, way too limited. Um, we cannot do perfect prediction. We will never have the data sets that allow me to figure out what you're about ready to do, what question you're gonna ask next.

   5. LFLex Fridman21:00

      Right.

   6. MJMichael I. Jordan21:00

      I have no clue. I will never know such things. Moreover, most of us find ourselves during the day in all kinds of situations we had no anticipation of, that are kind of very, very s- that are novel in various ways. And in that moment, we want to think through what we want. And also there's gonna be market forces acting on us. Uh, I'd like to go down that street, but now it's full because there's a crane in the street. I gotta, I gotta think about that. I gotta think about what I might really want here. And I gotta sort of think about how much it costs me to do this action versus this action. I gotta think about, uh, the risks involved. You know, a lot of our current pattern recognition and prediction systems don't do any risk evaluations. They have no error bars, right? I gotta think about other people's decisions around me. I gotta think about a collection of my decisions. Even just thinking about, like, a medical treatment. You know, I'm not gonna take a, the prediction of a neural net about my health, about something consequential, am I about ready to have a heart attack because some number is over .7. Even if you had all the data in the world that had ever been collected about heart attacks, uh, better than any doctor ever had, I'm not gonna trust the output of that neural net to predict my heart attack. I'm gonna wanna ask what if questions around that.

   7. LFLex Fridman22:06

      Right.

   8. MJMichael I. Jordan22:06

      I'm gonna wanna look at some us- or other possible data I didn't have, causal things. I'm gonna wanna have a dialogue with a doctor about things we didn't think about when we gathered the data. You know, it... I could go on and on, I hope you can see.

   9. LFLex Fridman22:17

      Yeah.

   10. MJMichael I. Jordan22:17

       And, and I don't... I think that if you say prediction's everything, that, that, that you're missing all of this stuff. Um, and so prediction plus decision making is everything, but both of them are equally important. And so the field has emphasized prediction. Yann, rightly so, has seen how powerful that is. But, um, at the cost of people not being aware that decision making is where the rubber really hits the road, where human lives are at stake, where risks are being taken, where you gotta gather more data, you gotta think about the error bars, you gotta think about the consequences of your decisions on others, you gotta think about the economy around your decisions, blah, blah, blah, blah, blah. I'm not the only one working on those, but we're a smaller tribe, and right now we're not the, the one that people talk about the most. Um, but, you know, if you go out in the real world, in industry, um, you know, at Amazon, I'd say half the people there are working on decision making and the other half are doing, you know, the pattern recognition. It's important.

   11. LFLex Fridman23:05

       And the words of pattern recognition and prediction, I think the distinction there, not to linger on words, but the distinction there is more a constrained sort of in the lab data set versus decision making is talking about consequential decisions in the real world under the messiness and the uncertainty of the real world, and just the whole of it. The whole mess of it that actually touches human beings and scale, like you said, market forces. That's the, that's the distinction.

   12. MJMichael I. Jordan23:30

       Yeah. I- it helps add those, that perspective, that broader perspective.

   13. LFLex Fridman23:33

       Right.

   14. MJMichael I. Jordan23:33

       Y- You're right. I totally agree. Uh, on the other hand, if you're a real prediction person, of course you want it to be in the real world.

   15. LFLex Fridman23:38

       Right.

   16. MJMichael I. Jordan23:38

       You wanna predict real world events. I'm just saying that's not possible with just data sets, uh, that it has to be in the context of, you know, uh, strategic things that someone's doing, data they might gather, things they could have gathered, the reasoning process around data. I- it's not just taking data and making predictions based on the data.

7. 23:53 – 43:34

   Recommender systems and distributed decision-making at scale

   1. MJMichael I. Jordan23:53

   2. LFLex Fridman23:53

      So one of the, the things that you're working on, uh, I'm sure there's others working on it, but I don't hear often, uh, it talked about, especially in, in the clarity that you talk about it, and I think is both the most exciting and the most concerning area of AI i- in terms of decision making. So you've talked about AI systems that help make decisions that scale in a distributive way, millions, billions decisions, in sort of markets of decisions. Can you, as a starting point, sort of give an example of a system that you think about when you're thinking about these kinds of systems?

   3. MJMichael I. Jordan24:27

      Uh, yeah. So first of all, you're, you're absolutely getting into some territory which will... I, I will be beyond my expertise and the... and there are lots of things that are gonna be very non-obvious to think about. Just like, just the... Uh, again, I like to think about history a little bit, but think about, put yourself back in the '60s. There was kind of a banking system that wasn't computerized really.

   4. LFLex Fridman24:43

      Mm-hmm.

   5. MJMichael I. Jordan24:43

      There was... Then there was database theory emerging.... and database people had to think about, "How do I actually not just move data around, but actual money, and have it be, you know, valid, and have transactions at ATMs happen that are actually, you know, all valid," and, and so on and so forth. So, that's the kind of issues you get into when you start to get serious about sorts of things like this. Um, I like to think about as kind of almost a thought experiment to help me think, uh, something simpler, which is, um, the music market. And, uh, 'cause there is, uh, to first order, there is no music market in the world right now, in the co- in our country, for sure. Uh, there are, uh, something called, things called record companies, and they make money, uh, and they prop up a few, um, really good musicians and make them superstars, and they all make huge amounts of money. Um, but there's a long tail of huge numbers of people that make lots and lots of really good music that is actually listened to by more people than the famous people. Um, um, uh, they are not in a market. They cannot have a career. They do not make money.

   6. LFLex Fridman25:43

      The creators, the creator's side.

   7. MJMichael I. Jordan25:45

      The creators, the so-called influencers or whatever. That diminishes who they are, right? So, there are people who make extremely good music, especially in the hip hop or Latin wor- world these days. Uh, they do it on their laptop, that's what they do, um, on the weekend, uh, and they have, uh, uh, another job during the week. They put it up on SoundCloud or other sites. Eventually it gets streamed. It now gets turned into bits. It's not economically valuable. The, the information is lost. It gets put up there, people stream it. Y- you walk around in, uh, a big city, you see people with headphones on, you know, especially young kids listening to music all the time. If you look at the data, none of the m- very little of the music they listen to is, is the famous people's music. And none of it's old music, it's all the latest stuff. But the people who made that latest stuff are like, some 16-year-olds somewhere who will never make a career out of this, who will never make money. Of course, there'll be a few counterexamples, the record companies incentivized to pick out a few and, and highlight them. Long story short, there's a missing market there. There is not a consumer-producer relationship at the level of the actual creative acts. Um, the Pipelines and Spotifys of the world that take this stuff and stream it along, they make money off of subscriptions, or advertising and those things. They're making the money, all right? And then they will offer bits and pieces of it to a few people, again, to highlight that, you know, they're ... they simulate a market. Anyway, a real market would be if you're a creator of music, that you actually are somebody who's good enough that people want to listen to you, uh, you should have the data available to you. There should be a dashboard showing a map of the United States, showing last week, here's all the places your songs were listened to. It should be transparent, um, vettable, so that if someone in, down in Providence sees that, uh, you're being listened to 10,000 times in Providence, that they know that's real data, you know it's real data. They will have you come give a show down there. They will broadcast to the people who've been listening to you that you're coming. If you do this right, you could, you could, you know, go down there and make $20,000. You do that three times through your career, you start to have a career. So, in this sense, AI creates jobs. It's not about taking away human jobs, it's creating new jobs because it creates a new market. Once you've created a market, you've now connected up producers and consumers. You know, the mu- person who's making the music can say to someone who comes to their shows a lot, "Hey, I'll play at your daughter's wedding for $10,000." You'll say, "$8,000." They'll say, "$9,000." Um, y- then you, you again, you, you can now get an income up to $100,000. You're not gonna be a millionaire, all right? And, and now even think about really the value of music is in these personal connections, even so much so that, um, a, a young kid wants to wear a T-shirt with the, their favorite musician's signature on it, right? So, if, if they listen to the music on the internet, the internet should be able to provide them with a button that they push and the merchandise arrives the next day. We can do that, right? And now, why should we do that? Well, because the kid who bought the shirt'll be happy, but more the person who made the music will get the money. There's no advertising needed, right? So, you could create markets between producers and consumers, take 5% cut. Your company will be perfectly, uh, sound. It'll go forward into the future and it will create new markets, and that raises human happiness. Um, now, this seems like, well, this is easy. Just create this dashboard, kind of create some connections and all that. But, uh, you know, if you think about Uber or whatever, you think about the, the challenges in the real world of doing things like this. And there are actually new principles going to be needed. If you're trying to create a new kind of two-way market at a different scale that's ever been done before, there's gonna be, um, you know, uh, unwanted aspects of the market. There'll be bad people. There'll be, you know, um, the data will get used in wrong ways. You know, it'll fail in some ways. It won't deliver val- You have to think that through. Just like anyone who, like, ran a big auction or, you know, or ran a big matching service in economics will think these things through. And so, that maybe doesn't get at all the huge issues that can arise when you start to create markets, but it starts for at least, uh, for me, solidify my thoughts and let- allow me to move forward in my own thinking.

   8. LFLex Fridman29:28

      Yeah. So, uh, I talked to head of research at Spotify actually, and I think their long-term goal they've said is to, uh, have at least one million creators h- make, uh, make a comfortable living putting on Spotify. So, in, and ... I, I think you articulate a really nice vision of, uh, the world in the digital s- and the cyberspace of markets. What, what do you think companies like Spotify or YouTube or Netflix can do to create such markets? Is it an AI problem? Is it an interface problem, so interface design? Is it, um, some other kind of ... What, is it an economics problem? Who, who should they hire (laughs) to solve these problems?

   9. MJMichael I. Jordan30:15

      Well, part of it's not just top down. So, the Silicon Valley has this attitude that they know how to do it. They will-

   10. LFLex Fridman30:20

       Right.

   11. MJMichael I. Jordan30:20

       ... create the system, just like Google did with the search box, that will be so good that they'll just, everyone will adopt that.

   12. LFLex Fridman30:25

       Yeah.

   13. MJMichael I. Jordan30:25

       Right? Um, it's not, it's, it, it's everything you said, but really, I think missing the kind of culture.

   14. LFLex Fridman30:31

       Mm-hmm.

   15. MJMichael I. Jordan30:31

       All right? So, it's literally that 16-year-old who's cr- who's able to create the songs. You don't create that as a Silicon Valley entity. You don't hire them, per se.

   16. LFLex Fridman30:38

       Okay.

   17. MJMichael I. Jordan30:39

       Right? You have to create an ecosystem in which they are wanted and that they're belong, right? And so you have to have some cred- cultural credibility to do things like this. You know, Netflix, to their credit, wanted some of that s- credibility. They created shows, you know, content. They call it content. It's such a terrible word, but it's cul- it's culture.

   18. LFLex Fridman30:56

       Yeah.

   19. MJMichael I. Jordan30:56

       Right? And so with movies you can kind of go give a large sum of money to somebody graduating from the USC film school, uh-... uh, it's a whole thing of its own, but it's kind of, like, rich white people's thing to do, you know? And, you know, American culture has not been so much about rich white people. (laughs) It's been about all the immigrants, all the af- all the Africans who came and brought that culture and those, th- those rhythms and, and that, that, to, to, to this world and created this whole new thing, you know, American culture. And, and so, companies can't artificially create that. They can't just say, "Hey, we're here. Uh, we're gonna buy it up." You got a partner.

   20. LFLex Fridman31:30

       Right.

   21. MJMichael I. Jordan31:31

       And, um, so, but anyway, you know, not to denigrate, these companies are all trying and they should, and they, they, they're pr- they... I'm sure they're asking these questions and some of them are even g- making an effort. But it, it, it is partly a respect the culture as you are a te- as a technology person. You gotta blend your technology with cultural, with, with cultural, uh, you know, meaning.

   22. LFLex Fridman31:49

       How much of a role do you think the algorithm through machine learning has in connecting the consumer to the creator, sort of, uh, the recommender system aspect of this?

   23. MJMichael I. Jordan31:59

       Yeah. It's a great question. I think pretty high. Recomme- y- you know, um, there's no magic in the algorithms, but a good recommender system is way better than a bad recommender system. And, uh, recommender systems was a billion-dollar industry back even, you know, 10, 20 years ago. Um, and it continues to be extremely important going forward.

   24. LFLex Fridman32:17

       What's your favorite recommender system, just so we can put something-

   25. MJMichael I. Jordan32:19

       Well, just historically, I was one of the, you know, when I first went to Amazon, uh, you know, I first didn't like Amazon 'cause they put the bookkeeper out of business or the library, you know, the local, uh, booksellers went out of business. Um, I've come to accept that there, you know, there probably are more books being sold now and more people reading them than ever before. Uh, and then lo- local book st- stores are coming back. So, you know, that's how economics sometimes work. You go up and you go down. Um, but, uh, anyway, when I finally started going there (laughs) and I bought a few books, I was really pleased to see another few books being recommended to me that I never would've thought of. Uh, and I bought a bunch of them, so they obviously had a good business model. Um, but I learned things, and I still, to this day, kind of browse using that service. Um, and I think lots of people get a lot, you know, they're... That, that, that is a good aspect of a recommendation system. I'm learning from my peers in a, in a, in an indirect way. Um, and their algorithms are not meant to th- have them impose what we, what we learn. It really is trying to find out what's in the data. Uh, it doesn't work so well for other kind of entities, but that's just the complexity of human life, like shirts. You know, I'm not gonna g- get recommendations on shirts. And, uh, but that's, that's, that's interesting. Uh, if you try to recommend, um, uh, restaurants, it's, it's, it's a, it's, it's di- it's hard. It's hard to do it at scale. And, and, um, but, uh, a blend of recommendation systems with other, um, uh, economic ideas, uh, matchings and so on is really, really still very open research-wise, and there's new companies that are gonna emerge that do that well.

   26. LFLex Fridman33:48

       What, what do you think is going to the messy difficult land of, say, politics and things like that, that YouTube and Twitter have to deal with in terms of recommendation systems, being able to suggest, uh, I think Facebook just launched Facebook News. So they're having, uh, be- recommend the kind of news that are most likely for you to be interesting. Do you think this is a S- AI solvable, again, whatever term we wanna use, do you think it's a solvable problem for machines or is it a deeply human problem that sounds solvable?

   27. MJMichael I. Jordan34:18

       Uh, so I don't even think about it at that level. I think that what's broken with some of these companies, it's all monetization via advertising. They're not... At least Facebook. Let's... I wanna critique them, but they didn't really try to connect a producer and a consumer in an economic way, right? No one wants to pay for anything.

   28. LFLex Fridman34:34

       Right.

   29. MJMichael I. Jordan34:34

       And so they all, you know, starting with Google, then Facebook, they went back to the playbook of, you know, the, the television companies back in the day. No one wanted to pay for this signal. They will pay for the TV box, but not for the signal, at least back in the day. And so advertising kind of filled that gap. And advertising was new and interesting and it somehow didn't take over our lives quite, right? Uh, fast forward, Google provides a service that people don't wanna pay for. Um, and so, sur- somewhat surprisingly in the '90s, they made, ended up making huge amounts. They cornered the advertising market. It didn't seem like that was gonna happen, at least to me. Um, these little things on the right-hand side of the screen just did not seem all that economically interesting. But that, companies had maybe no other choice. The TV market was going away and billboards and so on. Um, so they've... they got it. And I think that sadly that, uh, Google just has m- it was doing so well with that at making such money. They didn't think much more about how... Wait a minute, is there a producer-consumer relationship to be set up here, not just, uh, between us and the advertisers market to be created? Is there an actual market between the producer and consumer? They're the producers, the person who created that video clip, the person that made that website, the person who could make more such things, the person who could adjust it and as a function of demand, the person on the other side who's asking for different kinds of things, you know? So, you see glimmers of that now. There's influencers and there's kind of a little glimmering of a market, but it should have been done 20 years ago. It should have been thought about. It should have been created in parallel with the advertising ecosystem. Uh, and then Facebook inherited that, and I think they also didn't think very much about that. So fast forward, and now they are making huge amounts of money off of advertising and the news thing and all these clicks is just, is feeding the advertising. It's all connected up to the advertising. So you want more people to click on certain things because that money flows to you, Facebook. You're very much incentivized to do that. And when you start to find it's breaking, so people are telling you, "Well, we're getting into some troubles." You try to adjust it with your smart AI algorithms.

   30. LFLex Fridman36:26

       Mm-hmm.
8. 43:34 – 1:01:11

   Facebook, privacy, and trust

   1. MJMichael I. Jordan43:34

   2. LFLex Fridman43:34

      So, I apologize that we've kind of returned to words. Uh, I don't think the exact terms matter, but in sort of defense of advertisement, do- don't you think the kind of direct connection between consumer and, uh, creator-producer is the best, like the, is what advertisement strives to do? Right? So at its best, advertisement is literally now Facebook is listening to our conversation and heard that you're going to India and will be able to actually start automatically for you making these connections and start giving this offer. So like, uh, I apologize if it's just a matter of terms, but just to draw a distinction, is it possible to make advertisements just better and better and better algorithmically to where it, it actually becomes a connection? Almost like direct marketing.

   3. MJMichael I. Jordan44:25

      That's a good question. So let's keep on that- push on that. First of all, I, I... What we just talked about, I was defending advertising. Okay? So I was defending it as a way to get signals into a market that don't come any other way, uh, especially algorithmically. It's a sign that someone's putting money on it. It's a sign they think it's valuable.

   4. LFLex Fridman44:41

      It's the-

   5. MJMichael I. Jordan44:41

      And if I think that if other things- someone else thinks it's valuable, then if I trust other people, I might be willing to listen.

   6. LFLex Fridman44:47

      It's the-

   7. MJMichael I. Jordan44:47

      I don't trust that Facebook, though, is- who's an intermediary between this. I don't think they care about me. Okay? I don't think they do. And I find it creepy that they know I'm going to India next week because of our conversation.

   8. LFLex Fridman45:01

      Why do you think that is? Can we... So what... C- can you just, uh, put your PR hat on (laughs) . Why do you think you find Facebook, uh, creepy and not trust them as, as do majority of the population? So they're, uh, out of the Silicon Valley companies, I saw like, uh, not approval rate, but there's, there's ranking of how much people trust companies and Facebook is in, in the gutter.

   9. MJMichael I. Jordan45:23

      In the gutter, including people inside of Facebook (laughs) .

   10. LFLex Fridman45:25

       Yeah, so what, what, uh, what do you attribute that to? Because f- when I-

   11. MJMichael I. Jordan45:28

       Come on. You don't find it creepy that right now we're talking, that I might walk out on the street right now, that some unknown person who I don't know kind of comes up to me and says, "I hear you're going to India."

   12. LFLex Fridman45:37

       I-

   13. MJMichael I. Jordan45:37

       I mean, that's not even Facebook. That's just a... If... I- I want transparency in human society. I want to have, if you know something about me, there's actually some reason you know something about me, that's something that if I look at it later and audit it, kind of, I'm, I approve. You know something about me 'cause, uh, you care in some way. There's a caring relationship even, or an economic one or something.

   14. LFLex Fridman45:58

       Right.

   15. MJMichael I. Jordan45:58

       Not just that you're someone who could exploit it in ways I don't know about or care about, or, or I- I'm troubled by, or, or whatever.

   16. LFLex Fridman46:05

       But a lot of companies-

   17. MJMichael I. Jordan46:05

       And we're in a world right now where that happens way too much. And that Facebook knows things about a lot of people and could exploit it, and does exploit it at times. I think most people do find that creepy. It's not for them. It's not, it's not that... Facebook does not doing it 'cause they care about them, right? In any real sense. And they shouldn't. They should not be a big brother caring about us. That is not the role of a company like that.

   18. LFLex Fridman46:28

       Why not?

   19. MJMichael I. Jordan46:29

       Right?

   20. LFLex Fridman46:29

       Wait, wait. Not the big brother part-

   21. MJMichael I. Jordan46:31

       Yeah.

   22. LFLex Fridman46:31

       ... but the caring, the trusting. I mean, don't those companies... J- just to linger on it because a lot of companies have a lot of information about us. I, I would argue that there's companies like Microsoft that has more information about us than Facebook does and yet-

   23. MJMichael I. Jordan46:44

       Yeah.

   24. LFLex Fridman46:44

       ... we trust Microsoft more.

   25. MJMichael I. Jordan46:45

       Well, Microsoft is pivoting. Microsoft, you know, under Satya Nadella has decided this is really important. We, we don't wanna do creepy things. Really want people to trust us to actually only use information in ways that they really would approve of, that we don't decide. Right? And, um, I'm just kind of adding that the health, health of a, a market is that, uh, when I connect to someone who producer-consumer, it's not just a random producer-consumer. It's pe- people who see each other. They don't like each other, but they sense that if they transact, some happiness will go up on both sides. If a company helps me to do that in moments that I choose, of my choosing, um, then fine. So... And also, think about the difference between, you know, browsing versus buying, right? There are moments in my life, I just wanna buy, you know, a gadget or something. I need something for that moment. I need some ammonia for my house or something 'cause I got a problem, a spill. Um, I wanna just go in. I don't wanna be advertised at- at that moment. I don't wanna be led down various direct- I- you know, that's annoying. I want to just go and have it be extremely easy to do what I want. Um, uh, other moments, I might say, no, I'm... It's like in the... Today, I'm going to the shopping mall. I wanna walk around and see things and see people and be exposed to stuff. So I want control over that, though. I don't want the company's algorithms to decide for me. Right? And I think that's the thing. We- it's a total loss of control if Facebook thinks they should take the control from us of deciding when we want to have certain kinds of information and when we don't, what information that is, how much it relates to what they know about us that we didn't really want them to know about us. They're not... I don't want them to be helping me in that way. I don't want them to be helping them by they decide, well, they have control over, um, um, what I want and when.

   26. LFLex Fridman48:22

       I totally agree. So it's the Facebook... By the way, I have this optimistic thing where I think Facebook has the kind of personal information about us that could create a beautiful thing. So I- I'm really optimistic of what Facebook could do. Uh, not what it's doing, but what it could do. So-

   27. MJMichael I. Jordan48:39

       I don't see that... I think that optimism is misplaced because there's not a bu- you have to have a business model behind these things.

   28. LFLex Fridman48:44

       Yes. No, you have to-

   29. MJMichael I. Jordan48:45

       Cre- cre- create a beautiful thing is really-

   30. LFLex Fridman48:46

       Yeah.
9. 1:01:11 – 1:02:32

   Are human beings fundamentally good?

   1. MJMichael I. Jordan1:01:11

      it. Um...

   2. LFLex Fridman1:01:11

      So, sorry for the big philosophical question, but on that topic, do you think human beings, 'cause you've also, out of all things, had a foot in psychology too, uh, the, do you think human beings are fundamentally good? Like, all of us have good intent that could be mined? Or is it, depending on context and environment, everybody could be evil?

   3. MJMichael I. Jordan1:01:34

      So, my answer is fundamentally good, um, but fundamentally limited. All of us have very, you know, blinkers on. We don't see the other person's pain that easily. We don't see the other person's point of view that easily. We're very much in our own head, in our own world. Um, and on my- my good days, I think that technology could open us up to, you know, more perspectives and more- less blinkered and more understanding. You know, a lot of wars in human history happened because of just ignorance. They didn't- they- they thought the other person was doing this when the other person wasn't doing this and we have huge amounts of that. Um, but in my lifetime, I've not seen technology really help in that way yet. And I- I- I do- I do believe in that, but, you know, no, I think fundamentally human- humans are good. People suffer, people have grievances, people have grudges and those things cause them to do things they probably wouldn't want. They regret it often, um, so no, I- I- I think it's a, you know, part of the progress of technology is to indeed allow it to be a little easier to be the real good person you actually are.

   4. LFLex Fridman1:02:31

      Well

10. 1:02:32 – 1:04:27

    Can a human life and society be modeled as an optimization problem?

    1. LFLex Fridman1:02:32

       put. Do you think individual human life or society can be modeled as an optimization problem?

    2. MJMichael I. Jordan1:02:41

       Um, not to the way I think typically. I mean, that's you're talking about one of the most complex phenomena in the whole, you know, in all of the universe.

    3. LFLex Fridman1:02:47

       Which? The individual human life or society as a whole?

    4. MJMichael I. Jordan1:02:50

       Both, both. I mean, individual human life is- is amazingly complex and, um, so, uh, you know, optimization's kind of just one branch of mathematics that talks about certain kind of things and, uh, it just, it feels way too limited for the complexity of, uh, such things.

    5. LFLex Fridman1:03:04

       What properties of optimization problems... Do you think so- do you think most interesting problems that could be solved through optimization, uh, what kind of properties does that surface have? Non-convexity, convexity, linearity, all those kinds of things? Saddle points?

    6. MJMichael I. Jordan1:03:20

       Well, so optimization's just one piece of mathematics, you know, there's like, you just, even in our era we're aware that, say, sampling, um, is coming up with examples of something, um, coming up with a distribution-

    7. LFLex Fridman1:03:30

       What's- what's optimization? What's sampling?

    8. MJMichael I. Jordan1:03:33

       Well, you- they- you can, if you're kind of a certain kind of mathematician, you can try to blend them and make them see- seem to be sort of the same thing, but optimization is, roughly speaking, trying to, uh, find a point that, um, a single point that is the optimum of a criterion function of some kind. Um, and sampling is trying to, from that same surface, treat that as a distribution or density and find prob- points that have high density. So, um, I- I want the entire distribution in a sampling paradigm, and I want the, um, you know, the- the single point that's the best point in the para- in the samp- in the, uh, optimization paradigm. Now, if you were optimizing in the space of probability measures, the output of that could be a whole probability distribution. So, you can start to make these things the same, but in mathematics, if you go too high up that kind of abstraction arc, you start to lose the, uh, you know, the ability to do the interesting theorems, so you kind of don't try to- you don't try to overly- over abstract.

    9. LFLex Fridman1:04:26

       So,

11. 1:04:27 – 1:04:59

    Is the world deterministic?

    1. LFLex Fridman1:04:27

       as a small tangent, what kind of world view do you find more appealing? One that is deterministic or stochastic, uh, statistic-

    2. MJMichael I. Jordan1:04:36

       Well, that's easy.

    3. LFLex Fridman1:04:36

       ...ian?

    4. MJMichael I. Jordan1:04:36

       I mean, I'm a statistician, you know, the- the world is highly stochastic. We- I don't know what's going to happen in the next five minutes, right? What you're gonna ask, what we're gonna do, what I'll- I'll say.

    5. LFLex Fridman1:04:44

       Due to the uncertainty? Due to the-

    6. MJMichael I. Jordan1:04:45

       Massive uncertainty.

    7. LFLex Fridman1:04:47

       Yeah.

    8. MJMichael I. Jordan1:04:47

       You know, massive uncertainty. And so the best I can do is have kind of a rough sense, or probability distribution on things, and somehow use that in my reasoning about what to do now.

    9. LFLex Fridman1:04:58

       So,

12. 1:04:59 – 1:09:52

    Role of optimization in multi-agent systems

    1. LFLex Fridman1:04:59

       how does the distributed at scale when you have multi-agent systems, uh, look like? So, optimization can optimize sort of, it makes a lot more sense sort of, uh, at least from my- from a robotics perspective for a single robot, for a single agent trying to optimize some objective function. Uh, wha- when you start to enter the real world, this game theoretic concept starts popping up, uh, that... H- how do you see optimization in this? 'Cause you've talked about markets and the scale, what does that look like? Do you see it as optimization? Do you see it as sampling? Do you see- like how- how should you match up, match up...

    2. MJMichael I. Jordan1:05:38

       Yeah. These all blend together, um, and a system designer thinking about how to build an incentivized system will have a blend of all these things. So, you know, a particle in a potential well is optimizing a function called a Lagrangian, right? The particle doesn't know that. There's no algorithm running that does that. It just happens.

    3. LFLex Fridman1:05:55

       Okay.

    4. MJMichael I. Jordan1:05:55

       It's- it's, so it's a description mathematically of something that helps us understand as analysts what's happening, right? And so the same thing will happen when we talk about, you know, mixtures of humans and computers and markets and so on and so forth, there'll be certain principles that allow us to understand what's happening whether or not the actual algorithms are being used by any sense is not clear. Um, now at- at some point I may have set up a multi-agent or market kind of system, and I'm now thinking about an individual agent in that system, and they're asked to do some task and they're incentivized in some way. They get certain signals and they- they have some utility. Maybe what they will do at that point is they- they just won't know the answer, they may have to optimize to find an answer, okay? So, and optimizers could be embedded inside of an overall market, uh, you know, and game theory is- is very, very broad. Um, it is often studied very narrowly for certain kinds of problems, um, but it's roughly speaking this is just the, "I don't know what you're gonna do, so I kind of anticipate that a little bit, and then you anticipate what I'm anticipating," and we kind of go back and forth in our own minds. We run kind of thought experiments.

    5. LFLex Fridman1:06:55

       You've talked about this interesting point in terms of, uh, game theory, so, you know, most optimization problems really hate saddle points. Maybe you can describe what saddle points are? But-I've heard you kind of mention that there's a- there's a branch of optimization that you could try to explicitly look for saddle points as a good thing.

    6. MJMichael I. Jordan1:07:15

       Oh, not optimization. That's just game theory. That- that's so, uh, y- there's all kinds of different equilibria in game theory, and some of them are highly explanatory behavior. They're- they're not attempting to be algorithmic, they're just trying to say, "If you happen to be at this equilibrium, you would see certain kind of behavior," and we see that in real life. That's what an economist wants to do, especially a behavioral economist. Um, uh, in- in continuous, uh, differential game theory, you're in continuous spaces, a, um, some of the simplest equilibria are saddle points. A Nash equilibrium is a saddle point. It's a special kind of saddle point. So, classically in game theory, you were trying to find Nash equilibria, and in an algorithm of game theory, you were trying to find algorithms that would find them. Uh, and so you're trying to find saddle points. I mean, so that's- that's literally what you're trying to do. Um, but, you know, any economist knows that Nash equilibria, uh, have their limitations. They are definitely not that explanatory in many situations. They're not what you really want. Um, there's other kind of equilibria, and there's names associated with these 'cause they came from history with certain people working on them, but there will be new ones emerging. So, you know, one example is a Stackelberg equilibrium. So, you know, Nash, you and I, are both playing this game against each other or for each other, maybe it's cooperative, and we're both gonna think it through, and then we're gonna decide, and we're gonna off- you know, do our thing simultaneously. You know, in a Stackelberg, no, I'm gonna be the first mover. I'm gonna make a move, you're gonna look at my move, and then you're gonna make yours. Now, since I know you're gonna look at my move, I anticipate what you're gonna do, and so I don't do something stupid. But- and- but then I know that you are also anticipating me, so we're kind of going back and forth in our mind. But there is then a first mover thing. And so there's a- those are different equilibria, all right? And, uh, so just mathematically, yeah, these things have certain topologies and certain shapes that are like saddle points, and then algorithmically or dynamically, how do you move towards them? How do you move away from things? Um, you know, so some of these questions have answers, they've been studied, others do not. And especially if it becomes stochastic, especially if there's large numbers of decentralized things, there's just, uh, you know, young people getting in this field who kind of think it's all done because we have, you know, TensorFlow.

Episode duration: 1:45:48