Guido van Rossum: Python

[0] The following is a conversation with Guidoven Rossum, creator of Python, one of the most popular programming languages in the world, used in almost any application that involves computers, from web, backend development, to psychology, neuroscience, computer vision, robotics, deep learning, natural language processing, and almost any subfield of AI.

[1] This conversation is part of MIT course on artificial general intelligence and the artificial intelligence podcast.

[2] If you enjoy it, subscribe on YouTube, iTunes, or your podcast provider of choice, or simply connect with me on Twitter at Lex Friedman, spelled F -R -I -D.

[3] And now, here's my conversation with Guido von Rossum.

[4] You were born in the Netherlands in 1956.

[5] Your parents in the world around you was deeply impacted by World War II, as was my family from the Soviet Union so with that context what is your view of human nature are some humans inherently good and some inherently evil or do we all have both good and evil within us I did not expect such a deep one I guess we all have good and evil potential in us and a lot of it depends on circumstances and context out of that world at least on the Soviet Union side in Europe sort of out of suffering out of challenge out of that kind of set of traumatic events often emerges beautiful art music literature in an interview I read or heard you said you enjoy Dutch literature when you were a child can you tell me about the books that had an influence on you in your childhood?

[6] Well, as a teenager, my favorite writer was, my favorite Dutch author was a guy named Willem Friedrich Hermann's, who's writing, certainly his early novels, were all about sort of ambiguous things that happened during World War II.

[7] I think he was a young adult during the world.

[8] that time and he wrote about it a lot and very interesting very good books I thought I think in a non -fiction way no it was all fiction but it was very much set in in the ambiguous world of resistance against the Germans where often you couldn't tell whether someone was truly the resistance or really a spy for the Germans and and some of the characters in his novels sort of cross that line and you never really find out what exactly happened and in his novels there's always a good guy and a bad guy the nature of good and evil is it clear there's a hero it's no his heroes are often more his main characters are often anti heroes and and and And so they're not very heroic.

[9] They're often, they fail at some level to accomplish their lofty goals.

[10] And looking at the trajectory through the rest of your life, has literature, Dutch, or English, or translation had an impact outside the technical world that you existed in?

[11] Hmm.

[12] I still read novels.

[13] I don't think that it impacts me that much directly.

[14] It doesn't impact your work.

[15] It's just, it's a...

[16] It's a separate world.

[17] My work is highly technical, and sort of the world of art and literature doesn't really directly have any bearing on it.

[18] You don't think there's a creative element to the design, you know, some would say art, design of a language, art?

[19] I'm not disagreeing with that.

[20] I'm just saying that sort of I don't feel direct influences from more traditional art on my own creativity.

[21] Of course, you don't feel, it doesn't mean it's not somehow deeply there in your subconscious.

[22] Who knows?

[23] Who knows?

[24] So let's go back to your early teens.

[25] Your hobbies were building electronic circuits, building mechanical models, If you can just put yourself back in the mind of that young Guido, 12, 13, 14, was that grounded in a desire to create a system, so to create something, or was it more just tinkering, just the joy of puzzle solving?

[26] I think it was more the latter, actually.

[27] I maybe towards the end of my high school, period, I felt confident enough that I designed my own circuits that were sort of interesting somewhat, but a lot of that time I literally just took a model kit and followed the instructions, putting the things together.

[28] I mean, I think the first few years that I built electronics kits, I really did not have enough understanding of sort of electronics to really understand what I was doing.

[29] I mean, I could debug it and I could sort of follow the instructions very carefully, which has always stayed with me. But I had a very naive model of like how a transistor works.

[30] and I don't think that in those days I had any understanding of coils and capacitors which actually sort of was a major problem when I started to build more complex digital circuits because I was unaware of the sort of the analog part of the how they actually work and I would have things that the schematic looked, everything looked fine and it didn't work.

[31] And what I didn't realize was that there was some megahertz level oscillation that was throwing the circuit off because I had a sort of two wires were too close or the switches were kind of poorly built.

[32] But through that time, I think it's really interesting.

[33] interesting and instructive to think about because there's echoes of it are in this time now.

[34] So in the 1970s, the personal computer was being born.

[35] So did you sense in tinkering with these circuits?

[36] Did you sense the encroaching revolution in personal computing?

[37] So if at that point we would sit you down and ask you to predict the 80s and the 90s, do you think you would be able to do so successfully to unroll?

[38] this the process this happened no I I had no clue I I remember I think in the summer after my senior year or maybe it was the summer after my junior year well at some point I think when I was 18 I went on a trip to the math Olympiad in Eastern Europe and there was like I was part of the Dutch team and there were other nerdy kids that sort of had different experiences and one of them told me about this amazing thing called a computer and I had never heard that word my own explorations in electronics were sort of about very simple digital circuits and I had sort of I had the idea that I somewhat understood how a digital calculator worked.

[39] And so there is maybe some echoes of computers there, but I didn't, I never made that connection.

[40] I didn't know that when my parents were paying for magazine subscriptions using punched cards, that there was something called a computer that was involved, that read those cards and transferred the money between accounts.

[41] I was also not really interested in those things.

[42] It was only when I went to university to study math that I found out that they had a computer and students were allowed to use it.

[43] And there were some, you're supposed to talk to that computer by programming it.

[44] What did that feel like finding?

[45] Yeah, that was the only thing you could do with it.

[46] The computer wasn't really connected to the real world.

[47] the only thing you could do was sort of you typed your program on a bunch of punched cards you gave the punched cards to the operator and an hour later the operator gave you back your printout and so all you could do was write a program that did something very abstract and i don't even remember what my first forays into programming were but they were sort of doing simple math exercises and just to learn how a programming language worked.

[48] Did you sense, okay, first year of college, you see this computer, you're able to have a program and it generates some output.

[49] Did you start seeing the possibility of this, or was it a continuation of the tinkering with circuits?

[50] did you start to imagine that one the personal computer but did you see it as something that is a tool like a tool like a word processing tool maybe maybe for gaming or something or did you start to imagine that it could be you know going to the world of robotics like you you know the Frankenstein picture that you could create an artificial being there's like another entity in front of you you did not I don't think I really saw it that way I was really more interested in the tinkering.

[51] It's maybe not a sort of a complete coincidence that I ended up sort of creating a programming language, which is a tool for other programmers.

[52] I've always been very focused on the sort of activity of programming itself and not so much what happens with the program you write.

[53] I do remember, and I don't remember, maybe in my second or third year, probably my second, actually, someone pointed out to me that there was this thing called Conway's Game of Life.

[54] You're probably familiar with it.

[55] In the 70s, I think, as long as you came up with it.

[56] So there was a scientific American column by someone who did.

[57] a monthly column about mathematical diversions and also blanking out on the guy's name.

[58] It was very famous at the time and I think up to the 90s or so.

[59] And one of his columns was about Conway's Game of Life and he had some illustrations and he wrote down all the rules and sort of there was the suggestion that this was philosophically interesting, that that was why Conway had called it that.

[60] and all I had was like the two pages photocopy of that article.

[61] I don't even remember where I got it.

[62] But it spoke to me and I remember implementing a version of that game for the batch computer we were using where I had a whole Pascal program that sort of read an initial situation from input and read some numbers that that said do so many generations and print every so many generations and then out would come pages and pages of of sort of things patterns of different kinds and yeah yeah and i remember much later i've done a similar thing using python but i'd sort of that original version i wrote at the time I found interesting because I combined it with some trick I had learned during my electronics hobbyist times.

[63] I essentially first on paper I designed a simple circuit built out of logic gates that took nine bits of input, which is the sort of the cell and its neighbors, and produced the new value for that cell.

[64] And it's like a combination of a half adder and some other clipping.

[65] It's actually a full adder.

[66] And so I had worked that out.

[67] And then I translated that into a series of bullion operations on Pascal integers, where you could use the integers as bitwise values.

[68] And so I could basically.

[69] generate 60 bits of a generation in like eight instructions or so.

[70] Nice.

[71] So I was proud of that.

[72] It's funny that you mentioned.

[73] So for people who don't know Conway's Game of Life, it's a cellular automata where there's single compute units that kind of look at their neighbors and figure out what they look like in the next generation.

[74] based on the state of their neighbors, and this is a deeply distributed system, in concept at least, and then there's simple rules that all of them follow, and somehow out of this simple rule, when you step back and look at what occurs, it's beautiful, there's an emergent complexity, even though the underlying rules are simple, there's an emerging complexity.

[75] Now, the funny thing is you've implemented this, and the thing you're commenting, on is you're proud of a hack you did to make it run efficiently.

[76] When you're not commenting on, like this is a beautiful implementation, you're not commenting on the fact that there's an emergent complexity that you've, you've coded a simple program, and when you step back and you print out the following generation after generation, that's stuff that you may have not predicted what happened is happening.

[77] right and is that magic i mean that's the magic that all of us feel when we program when you create a program and then you run it and whether it's hello world or it shows something on screen if there's a graphical component or are you seeing the magic and the mechanism of creating that i think i went back and forth uh as a student we had an incredibly small budget uh of computer time that we could use it was actually measured.

[78] I once got in trouble with one of my professors because I had overspent the department's budget.

[79] It's a different story.

[80] So I actually wanted the efficient implementation because I also wanted to explore what would happen with a larger number of generations and a larger sort of of the of the board and so once the implementation was flawless I would feed it different patterns and then I think maybe there was a follow -up article where there were patterns that were like gliders patterns that repeated themselves after a number of generations but translated one or two positions to the right or up or something like that and there were I remember things like glider guns well you can you can Google Conway's game of life it's still a people still go awe and o over it for a reason because it's not really well understood why I mean this is what Stephen Wolfram is obsessed about okay he's just the the we don't have the mathematical tools to describe the kind of complexity that emerges and he's kinds of systems.

[81] The only way you can do is to run it.

[82] I'm not convinced that it's sort of a problem that lends itself to classic mathematical analysis.

[83] No. And so one theory of how you create an artificial intelligence or an artificial being is you kind of have to, same with a game of life, you kind of have to create a universe and let it run.

[84] that creating it from scratch in a design way in the, you know, coding up a Python program that creates a fully intelligence system may be quite challenging that you might need to create a universe, just like the game of life is.

[85] Well, you might have to experiment with a lot of different universes before.

[86] There is a set of rules that doesn't essentially always just end up repeating itself in a trivial way.

[87] Yeah, and Steve Wolfram works with these simple rules says that it's kind of surprising how quickly you find rules that create interesting things.

[88] You shouldn't be able to, but somehow you do.

[89] And so maybe our universe is laden with rules that will create interesting things.

[90] They might not look like humans, but the emergent phenomena that's interesting may not be as difficult to create as we think.

[91] Sure.

[92] But let me sort of ask, at that time, you know, some of the world, at least in popular press, was kind of captivated, perhaps at least in America, by the idea of artificial intelligence, that these computers would be able to think pretty soon.

[93] And did that touch you at all?

[94] Did that, in science fiction or in reality, in any way?

[95] I didn't really start reading science fiction until much, much later.

[96] I think as a teenager, I read maybe one bundle of science fiction stories.

[97] Was it in the background somewhere, like in your thoughts?

[98] That's sort of using computers to build something intelligent always felt to me because I felt I had so much understanding of what actually going.

[99] was on inside a computer.

[100] I knew how many bits of memory it had and how difficult it was to program and sort of I didn't believe at all that you could just build something intelligent out of that that would really sort of satisfy my definition of intelligence.

[101] I think the most influential thing that I read in my early 20s was Goel Escherbach that was about consciousness and that was a big eye -opener in some sense in what sense so so yeah so on your own brain did you do you did you at the time or do you now see your own brain as a computer or is there a total separation of the way so yeah you're very pragmatically practically know the limits of memory the limits of the sequential computing or weekly paralyzed computing and you just know what we have now and it's hard to see how it creates but it's also easy to see it was in the in the 40s 50s 60s and now at least similarities between the brain and our computers oh yeah i mean i I totally believe that brains are computers in some sense.

[102] I mean, the rules they used to play by are pretty different from the rules we can sort of implement in our current hardware.

[103] But I don't believe in like a separate thing that infuses us with.

[104] intelligence or consciousness or any of that there's no soul I've been an atheist probably from when I was 10 years old just by thinking a bit about math and the universe and well my parents were atheists now I know that you you could be an atheist and still believe that there is something sort of about intelligence or consciousness that cannot possibly emerge from a fixed set of rules, I am not in that camp.

[105] I totally see that sort of given how many millions of years evolution took its time, DNA is a particular machine that sort of encodes information and an unlimited amount of information in chemical form and has figured out a way to replicate itself.

[106] I thought that that was maybe it's 300 million years ago, but I thought it was closer to half a billion years ago that that sort of originated and it hasn't really changed, that the sort of the structure of DNA hasn't changed ever since.

[107] That is like our binary, code that we have in hardware.

[108] I mean...

[109] The basic programming language hasn't changed, but maybe the programming itself...

[110] Obviously did sort of...

[111] It happened to be a set of rules that was good enough to sort of develop endless variability and sort of the idea of self -replicating molecules competing with each other for resources and one type eventually sort of always taking over, that happened before there were any fossils.

[112] So we don't know how that exactly happened, but I believe it's clear that that did happen.

[113] Can you comment on consciousness and how you see it?

[114] Because I think we'll talk about programming quite a bit.

[115] We'll talk about intelligence connecting to programming fundamentally, mentally but consciousness consciousness is this whole other thing do you think about it often as a developer of a programming language and as a human those those are pretty sort of separate topics my sort of my line of work working with programming does not involve anything that that goes in the direction of developing intelligence or consciousness.

[116] But sort of privately, as an avid reader of popular science writing, I have some thoughts, which is mostly that I don't actually believe that consciousness is an all -or -nothing thing.

[117] I have a feeling that, and I forget what I read that influenced, this, but I feel that if you look at a cat or a dog or a mouse, they have some form of intelligence.

[118] If you look at a fish, it has some form of intelligence.

[119] And that evolution just took a long time, but I feel that the sort of evolution of more and more intelligence that led to sort of the human form of intelligence followed, the evolution of the senses, especially the visual sense.

[120] I mean, there is an enormous amount of processing that's needed to interpret a scene.

[121] And humans are still better at that than computers are.

[122] And I have a feeling that there is a sort of, the real, reason that like mammals is in particular developed the levels of consciousness that they have and that eventually going from intelligence to self -awareness and consciousness has to do with sort of being a robot that has very highly developed senses has a lot of rich sensory information coming in so the that's a really interesting thought that the that whatever that basic mechanism of DNA, whatever that basic building blocks of programming, if you just add more abilities, more high -resolution sensors, more sensors, you just keep stacking those things on top that this basic programming in trying to survive develops very interesting things that start to us humans to appear like intelligence and consciousness.

[123] Yeah, so in, as far as, as robots go, I think that the self -driving cars have that sort of the greatest opportunity of developing something like that, because when I drive myself, I don't just pay attention to the rules of the road.

[124] I also look around and I get clues from that, oh, this is a shopping district.

[125] Oh, here's an old lady crossing the street.

[126] Oh, here is someone carrying a pile of mail, there's a mailbox.

[127] I bet you they're going to cross the street to reach that mailbox, and I slow down.

[128] And I don't even think about that.

[129] And so there is so much where you turn your observations into an understanding of what other consciousnesses are going to do, or what other systems in the world are going to be, oh, that tree is going to fall.

[130] Yeah.

[131] I see sort of, I see much more of, I expect somehow that if anything is going to become conscious, it's going to be the self -driving car and not the network of a bazillion computers in a Google or Amazon data center that are all networked together to do whatever they.

[132] they do so in that sense so you actually highlight because that's what I work in is an autonomous vehicles you highlight the big gap between what we currently can't do and what we truly need to be able to do to solve the problem under that formulation then consciousness and intelligence is something that basically a system should have in order to interact with us humans as opposed to some kind of abstract notion of a consciousness.

[133] Consciousness is something that you need to have to be able to empathize, to be able to fear, understand what the fear of death is, all these aspects that are important for interacting with pedestrians.

[134] You need to be able to do basic computation based on our human desires and flaws.

[135] If you sort of, yeah, if you look at that, a dog.

[136] The dog clearly knows, I mean, I'm not a dog owner, but I have friends who have dogs.

[137] The dogs clearly know what the humans around them are going to do, or at least they have a model of what those humans are going to do, and they learn.

[138] Some dogs know when you're going out, and they want to go out with you.

[139] They're sad when you leave them alone.

[140] They cry.

[141] They're afraid because they were mistreated when they were younger.

[142] We, we, we don't.

[143] We, we, we don't.

[144] We, we, we, we, don't assign sort of consciousness to dogs, or at least not all that much, but I also don't think they have none of that.

[145] So I think it's consciousness and intelligence are not all or nothing.

[146] The spectrum.

[147] It's really interesting.

[148] But in returning to programming languages and the way we think about building these kinds of things, about building intelligence, building consciousness, building artificial beings.

[149] So I think one of the exciting ideas came in the 17th century.

[150] And with Leibniz, Hobbes, Descartes, where there's this feeling that you can convert all thought, all reasoning, all the thing that we find very special in our brains, you can convert all of that into logic.

[151] You can formalize it, form a reasoning.

[152] And then once you formalize everything, all of knowledge, then you can just calculate.

[153] And that's what we're doing with our brains as we're calculating.

[154] So there's this whole idea that we, that this is possible.

[155] But they weren't aware of the concept of pattern matching in the sense that we are aware of it now.

[156] They sort of thought you, they had discovered incredible bits of mathematics like Newton's calculus.

[157] And their sort of idealism, their sort of extension of what they could do with logic and math sort of went along those lines.

[158] And they thought, there's like, yeah, logic.

[159] There's like a bunch of rules and a bunch of input.

[160] They didn't realize that how you recognize a face is not just a bunch of rules, but it's a shit ton of data.

[161] Plus, a circuit that sort of interprets the visual clues and the context and everything else and somehow can massively parallel pattern match against stored rules.

[162] I mean, if I see you tomorrow here in front of the Dropbox office, I might recognize you.

[163] Even if I'm wearing a different shirt.

[164] Yeah, but if I see you tomorrow in a coffee shop.

[165] shop in Belmont, I might have no idea that it was you or on the beach or whatever.

[166] I make those kind of mistakes myself all the time.

[167] I see someone that I only know as like oh, this person is a colleague of my wife's.

[168] And then I see them at the movies and I don't recognize them.

[169] But do you see those, you call it pattern matching.

[170] Do you see that rules is unable to encode that to everything you see all the piece of information you look around this room I'm wearing a black shirt I have a certain height I'm a human all these there's probably tens of thousands of facts you pick up moment by moment about this scene you take them for granted and you accumulate aggregate them together to understand the scene you don't think all of that could be encoded to wear at the end of the day you can just put it all on the table and calculate I don't know what that means I mean yes in the sense that there is no there is no actual magic there but there are enough layers of abstraction from sort of from the facts as they enter my eyes and my ears to the understanding of the scene that I don't think that that AI has really covered enough of of that distance it's like if you take a body and you realize it's built out of atoms, well, that is a uselessly reductionist view, right?

[171] Right.

[172] The body is built out of organs.

[173] The organs are built out of cells.

[174] The cells are built out of proteins.

[175] The proteins are built out of amino acids.

[176] The amino acids are built out of atoms, and then you get to quantum mechanics.

[177] So that's a very pragmatic view.

[178] I mean, obviously, as an engineer, I agree with that kind of view.

[179] But you also have to consider the, well, the Sam Harris view of, well, intelligence is just information processing.

[180] You just, like you said, you take in sensory information, you do some stuff with it, and you come up with actions that are intelligent.

[181] That may, he makes it sound so easy.

[182] I don't know who Sam Harris is.

[183] Oh, it's philosopher.

[184] So, like, this is how philosophers often think, right?

[185] and essentially that's what the cart was is wait a minute if there is like you said no magic so you basically says it doesn't appear like there's any magic but we know so little about it that it might as well be magic so just because we know that we're made of atoms just because we know we're made of organs the fact that we know very little how to get from the atoms to organs in a way that's recreateable means it that you shouldn't get too excited just yet about the fact that you figured out that were made of atoms right and and and the same about taking facts as are our sensory organs take them in and turning that into reasons and actions that sort of there are a lot of abstractions that we haven't quite figured out how to how to deal with those.

[186] I mean, sometimes, I don't know if I can go on a tangent or not.

[187] Please drag you back in.

[188] Sure.

[189] So if I take a simple program that parses, say I have a compiler, it parses a program.

[190] In a sense, the input routine of that compiler, of that parser is a sensing organ.

[191] and it builds up a mighty complicated internal representation of the program it just saw.

[192] It doesn't just have a linear sequence of bytes representing the text of the program anymore.

[193] It has an abstract syntax tree, and I don't know how many of your viewers or listeners are familiar with compiler technology, but there is fewer and fewer these days, right?

[194] That's also true.

[195] probably.

[196] People want to take a shortcut.

[197] But there's sort of, this abstraction is a data structure that the compiler then uses to produce outputs that is relevant, like a translation of that program to machine code that can be executed by hardware.

[198] And then that data structure gets thrown away when a fish or a fly sees sort of gets visual impulses I'm sure it also builds up some data structure and for the fly that may be very minimal a fly may may have only a few I mean in the case of a fly's brain I could imagine that there are few enough layers of abstract that it's not much more than when it's darker here than it is here.

[199] Well, it can sense motion because a fly sort of responds when you move your arm towards it.

[200] So clearly, its visual processing is intelligent, well, not intelligent, but it has an abstraction for motion.

[201] And we still have similar things in, but much more complicated in our brains.

[202] I mean, otherwise you couldn't drive a car.

[203] if you couldn't sort of if you didn't have an incredibly good abstraction for motion yeah in some sense the same abstraction for motion is probably one of the primary sources of our of information for us we just know what to do i think we know what to do with that we've built up other abstractions on top we build much more complicated data structures based on that and we build more persistent data structures, sort of after some processing, some information sort of get stored in our memory pretty much permanently and is available on recall.

[204] I mean, there are some things that you sort of, you're conscious that you're remembering it.

[205] Like, you give me your phone number.

[206] I, well, at my age, I have to write it down, but I could imagine I could remember those seven numbers or 10 digits and reproduce them in a while.

[207] if I sort of repeat them to myself a few times.

[208] So that's a fairly conscious form of memorization.

[209] On the other hand, how do I recognize your face?

[210] I have no idea.

[211] My brain has a whole bunch of specialized hardware that knows how to recognize faces.

[212] I don't know how much of that is sort of coded in our DNA and how much of that is trained over and over between the ages of zero and three.

[213] but but but somehow our brains know how to do lots of things like that that are useful in our interactions with with other humans with without really being conscious of how it's done anymore right so our actual day -day lives we're operating at the very highest level of abstraction we're just not even conscious of all the little details and delaying it there's compilers on top of it's like turtles on top of turtles or turtles all the way down it's compil all the way down.

[214] But that's essentially, you say that there's no magic.

[215] That's what I was trying to get at, I think, is with Descartes started this whole train of saying that there's no magic.

[216] I mean, there's always before him.

[217] Well, didn't Descartes also have the notion, though, that the soul and the body were fundamentally separate?

[218] Yeah, I think he had to write in God in there for political reasons.

[219] So I don't actually, I'm not a historian, but there's no. emotions in there that all of reasoning, all of human thought can be formalized.

[220] I think that continued in the 20th century with the Russell and with Gato's Incompleteness theorem, this debate of what are the limits of the things that could be formalized.

[221] That's where the touring machine came along.

[222] And this exciting idea, I mean, underlying a lot of computing, that you can do quite a lot with a computer you can you can encode a lot of the stuff we're talking about in terms of recognizing faces and so on theoretically in an algorithm they can then run on a computer and in that context i'd like to ask programming in a philosophical way so what so what it what does it mean to program a computer so you said you write a python program or a compile called a C++ program that compiles to some bytecode.

[223] It's forming layers.

[224] You're programming in a layer of abstraction that's higher.

[225] How do you see programming in that context?

[226] Can it keep getting higher and higher levels of abstraction?

[227] I think at some point, the higher levels of abstraction will not be called programming and they will not resemble what we call programming at the moment.

[228] There will not be source code.

[229] I mean, there will still be source code sort of at a lower level of the machine, just like there are still molecules and electrons and sort of proteins in our brains.

[230] And so there's still programming and system administration and who knows what keeping to keep the machine running but what the machine does is is a different level of abstraction in a sense and as far as I understand the way that for last decade or more people have made progress with things like facial recognition or the self -driving cars is all by endless endless amounts of training data where at least as a layperson and I feel myself totally as a layperson in that field, it looks like the researchers who publish the results don't necessarily know exactly how their algorithms work.

[231] And I often get upset when I sort of read a sort of a fluff piece about Facebook in the newspaper or social networks and they say, well, algorithms.

[232] And that that's like a totally different interpretation of the word algorithm.

[233] Because for me, the way I was trained or what I learned when I was eight or 10 years old, an algorithm is a set of rules that you completely understand.

[234] They can be mathematically analyzed and you can prove things.

[235] You can like prove that Aristoceney sieve produces all prime numbers and only prime numbers.

[236] numbers.

[237] Yeah, so I don't know if you know who Andre Capathi is.

[238] I'm afraid not.

[239] So he's a head of AI at Tesla now, but he was at Stanford before, and he has this cheeky way of calling this concept software 2 .0.

[240] So let me disentangle that for a second.

[241] So kind of what you're referring to is the traditional, traditional, the algorithm, the concept of an algorithm, something that's there it's clear.

[242] You can read it.

[243] You understand it.

[244] You can prove it's functioning as kind of software 1 .0.

[245] And what software 2 .0 is, is exactly what you describe, which is you have neural networks, which is a type of machine learning, that you feed a bunch of data, and that neural network learns to do a function.

[246] All you specify is the inputs and the outputs you want, and you can't look concide.

[247] You can't analyze it.

[248] All you can do is train this function to map the inputs to the outputs by giving a lot of data.

[249] In that sense, programming becomes getting a lot of data.

[250] That's what programming is.

[251] Well, there would be programming 2 .0.

[252] 2 .0.

[253] Programming 2 .0.

[254] I wouldn't call that programming.

[255] It's just a different activity, just like building organs out of cells is not called chemistry.

[256] well so let's just step back and think sort of more generally of course but you know it's like as a parent teaching teaching your kids things can be called programming in that same sense that that's how programming is being used you're providing them data examples use cases so imagine writing a function not by not with four loops and clearly readable text but more saying well here's a lot of examples of what this function should take and here's a lot of examples of when it takes those functions it should do this and then figure out the rest so that's the 2 .0 concept and the question I have for you is like it's a very fuzzy way this is the reality of a lot of these pattern recognition systems and so on.

[257] It's a fuzzy way of quote -unquote programming.

[258] What do you think about this kind of world?

[259] Should it be called something totally different than programming?

[260] If you're a software engineer, does that mean you're designing systems that can be systematically tested, evaluated, they have a very specific specification, and then this other fuzzy software 2 .0 world, machine learning world, that's something else totally?

[261] Or is there some intermixing that's possible?

[262] Well, the question is probably only being asked because we don't quite know what that software 2 .0 actually is.

[263] And it sort of, I think there is a truism that every task that AI has, has, tackled in the past at some point we realized how it was done and then it was no longer considered part of artificial intelligence because it was no longer necessary to to use that term it was just oh now we know how to do this and a new field of science or engineering has been developed and I don't know if sort of every form of learning or sort of controlling computer systems should always be called programming.

[264] So I don't know, maybe I'm focused too much on the terminology.

[265] But I expect that there just will be different concepts where people with sort of different education and a different model of what they're trying to do will develop those concepts.

[266] Yeah, and I guess if you could comment on another way to put this concept is I think the kind of functions that neural networks provide is things, as opposed to being able to upfront prove that this should work for all cases you throw at it.

[267] All you're able, it's the worst case analysis versus average case analysis.

[268] All you're able to say is, it seems on everything we've tested, to work 99 .9 % of the time.

[269] But we can't guarantee it, and it fails in unexpected ways.

[270] We can even give you examples of how it fails in unexpected ways.

[271] But it's like really good most of the time.

[272] Yeah.

[273] Is there no room for that in current ways we think about programming?

[274] programming 1 .0 is actually sort of getting to that point too where the sort of the ideal of a bug -free program has been abandoned long ago by most software developers we only care about bugs that manifest themselves often enough to be annoying and we're willing to take the occasional crash or outage or incorrect result for granted because we can't possibly we don't have enough programmers to make all the code bug free and it would be an incredibly tedious business and if you try to throw formal methods at it it gets it becomes even more tedious so every once in a while the user clicks on a link and somehow they get an error and the average user doesn't panic they just click again and see if it works better the second time which often magically it does or they go up and they try some other way of performing their tasks so that's sort of an end -to -end recovery mechanism and inside systems there is all sorts of retries and timeouts and fallbacks and I imagine that that sort of biological systems are even more full of that because otherwise they wouldn't survive do you think programming should be taught and thought of as exactly what you just said I come from is kind of you're almost denying that fact always in sort of basic programming education the sort of the programs you're having students right are so small and simple that if there is a bug you can always find it and fix it because the sort of programming as it's being taught in some even elementary middle schools in high school introduction to programming classes in college typically it's programming in the small very few classes sort of actually teach software engineering building large systems i mean every summer here at dropbox we have a large number of interns every tech company on the west Coast has the same thing.

[275] These interns are always amazed because this is the first time in their life that they see what goes on in a really large software development environment.

[276] And everything they've learned in college was almost always about a much smaller scale.

[277] And somehow that difference in scale makes a quote.

[278] qualitative difference in how you do things and how you think about it.

[279] If you then take a few steps back into decades, 70s and 80s, when you were first thinking about Python or just that world of programming languages, did you ever think that there would be systems as large as underlying Google, Facebook, and Dropbox?

[280] Did you, when you were thinking about Python?

[281] I was actually always caught by surprise by, Every stage of it.

[282] Yeah, pretty much every stage of computing.

[283] So maybe just because you've spoken in other interviews, but I think the evolution of programming languages are fascinating.

[284] And it's especially because it leads from my perspective towards greater and greater degrees of intelligence.

[285] I learned the first programming language I played with in Russia was with the turtle logo.

[286] logo yeah and if you look i just have a list of programming languages all of which i've played with a little bit and they're all beautiful in different ways from fortran cobalt lisp algal 60 basic logo again c as a few object oriented came along in the 60s simula pascal small talk all of that leads all the classics the classics yeah the classic hits right uh scheme built that's built on top of Lisp on the database side SQL C++ and all that leads up to Python.

[287] Haskell 2 and all that's before Python Matlab these kind of different communities, different languages so can you talk about that world?

[288] I know that Python came out of ABC which actually never knew that language I just having researched this conversation went back to ABC and it looks remarkably it has a lot of annoying qualities but underneath those like all caps and so on but underneath that there's elements of python that are quite they're already there that's where I got all the good stuff all the good stuff so but in that world you're swimming these programming languages where you focused on just the good stuff in your specific circle or did you have a sense of what what is everyone chasing you said that every programming language is built to scratch an itch.

[289] Were you aware of all the itches in the community?

[290] And if not, or if yes, I mean, what itch were you trying to scratch with Python?

[291] Well, I'm glad I wasn't aware of all the itches because I would probably not have been able to do anything.

[292] I mean, if you're trying to solve every problem at once, you'll solve nothing.

[293] well yeah it's it's too overwhelming and so i had a very very focused problem i wanted a programming language that sat somewhere in between shell scripting and c and now arguably there is like one is higher level one is lower level.

[294] And Python is sort of a language of an intermediate level, although it's still pretty much at the high level.

[295] And I was thinking about much more about I want a tool that I can use to be more productive as a programmer in a very specific environment.

[296] And I also have had given myself a time budget for the development of the tool and that was sort of about three months for both the design like thinking through what are all the features of the language syntactically and semantically and how do I implement the whole pipeline from parsing the source code to executing it so I think both with a timeline and the goals, it seems like productivity was at the core of it as a goal.

[297] So, like, for me, in the 90s and the first decade of the 21st century, I was always doing machine learning, AI.

[298] Programming for my research was always in C++.

[299] Wow.

[300] And then the other people who are a little more mechanical engineering, electrical engineering, are matlabby.

[301] They're a little bit more matlab, folks.

[302] because those are the world.

[303] And maybe a little bit Java, too.

[304] But people who are more interested in emphasizing the object -oriented nature of things.

[305] So then in the last 10 years or so, especially with the oncoming of neural networks and these packages that are built on Python to interface with neural networks, I switch to Python.

[306] And it's just I've noticed a significant boost that I can't exactly, because I don't think about it, but I can't exactly put into words why I'm just much, much more productive, just being able to get the job done much, much faster.

[307] So how do you think whatever that qualitative difference is?

[308] I don't know if it's quantitative.

[309] It could be just a feeling.

[310] I don't know if I'm actually more productive.

[311] But how do you think about...

[312] You probably are.

[313] Yeah, well, that's right.

[314] I think there's elements, let me just speak to one aspect I think that was affecting my productivity.

[315] is C++ was, I really enjoyed creating performant code and creating a beautiful structure or everything that, you know, this kind of going into this, especially with the newer and newer standards of templated programming, of just really creating this beautiful, formal structure that I found myself spending most of my time doing that as opposed to getting parsing a file and extracting a few keywords or whatever the task I was trying to.

[316] to do so what is it about python how do you think of productivity in general as you were designing it now sort of through the decades last three decades what do you think it means to be a productive programmer and how did you try to design it into the language there are different tasks and as a programmer it's it's useful to have different tools available that sort of are suitable for different tasks So I still write C code, I still write shell code, but I write most of my things in Python.

[317] Why do I still use those other languages?

[318] Because sometimes the task just demands it.

[319] And, well, I would say most of the time the task actually demands a certain language because the task is not write a program that solves problem X from scratch, but it's more like fix a bug in existing program X or add a small feature to an existing large program.

[320] But even if you sort of, if you're not constrained in your choice of language by context like that, there is still the fact that if you write it in a certain language then you sort of you you have this balance between how long does it time does it take you to write the code and how long does the code run right and when you're in sort of in the phase of exploring solutions you often spend much more time writing the code than running it because every time you've sort of you've run it you see that the output is not quite what you wanted and you spend some more time coding and a language like Python just makes that iteration much faster because there are fewer details there is a large library there are fewer details that you have to get right before your program compiles and runs there are libraries that do all sorts of stuff for you so you can sort of very quickly take a bunch of existing components put them together and get your prototype application running just like when I was building electronics I was using a breadboard most of the time.

[321] So I had this like sprawled out circuit that if you shook it, it would stop working because it was not put together very well, but it functioned and all I wanted was to see that it worked and then move on to the next schematic or design or add something to it.

[322] Once you've sort of figured out, oh, this is the perfect design for my radio or light sensor or whatever then you can say okay how do we design a PCB for this how do we solder the components in a small space how do we make it so that it is robust against say voltage fluctuations or mechanical disruption I mean I know nothing about that when it comes to designing electronics but I know a lot about that when it comes to to writing code So the initial steps are efficient, fast, and there's not much stuff that gets in the way.

[323] But you're kind of describing, like Darwin described the evolution of species, right?

[324] You're observing of what is true about Python.

[325] Now, if you take a step back, if the act of creating language is art, and you had three months to do it, initial steps.

[326] so you just specified a bunch of goals sort of things that you observe about Python perhaps you had those goals but how do you create the rules the syntactic structure the features that result in those so I have in the beginning and I have follow -up questions about through the evolution of Python too but in the very beginning when you're sitting there creating the lexical analyze or whatever evolution was still a big part of it because I sort of, I said to myself, I don't want to have to design everything from scratch.

[327] I'm going to borrow features from other languages that I like.

[328] Oh, interesting.

[329] So you basically, exactly, you first observe what you like.

[330] Yeah.

[331] And so that's why if you're 17 years old and you want to sort of create a programming language, you're not going to be very successful at it.

[332] because you have no experience with other languages, whereas I was in my, let's say, mid -30s, I had written parsers before.

[333] So I had worked on the implementation of ABC.

[334] I had spent years debating the design of ABC with its authors, with its designers.

[335] I had nothing to do with the design.

[336] It was designed fully as it was.

[337] ended up being implemented when I joined the team.

[338] But so you borrow ideas and concepts and very concrete sort of local rules from different languages, like the indentation and certain other syntactic features from ABC, but I chose to borrow string literals and how numbers work from C and various other things.

[339] So in then, if you take that further, so you've had this funny sounding, but I think surprisingly accurate or least practical title of benevoling dictator for life for quite, you know, for the last three decades or whatever, or no, not the actual title, but functionally speaking.

[340] So you had to make decisions, design decisions.

[341] Can you, maybe let's take Python two, so Python, releasing Python, 3 as an example it's not backward compatible to python 2 in ways that a lot of people know so what was that deliberation discussion decision like yeah what was the psychology of that experience do you regret any aspects of how that experience undergone that well yeah so it was a group process really at that point even though i was bdFL in name in name and and and Certainly, everybody sort of respected my position as the creator and the current sort of owner of the language design.

[342] I was looking at everyone else for feedback.

[343] Sort of Python 3 .0 in some sense was sparked by other people in the community pointing out, oh well there are a few issues that sort of bite users over and over can we do something about that and for Python 3 we took a number of those Python words as they were called at the time and we said can we try to sort of make small changes to the language that address those words and we had sort of in the past we had always taken backwards compatibility very seriously and so many python wards in earlier versions had already been resolved because they could be resolved while maintaining backwards compatibility or sort of using a very gradual path of evolution of the language in a certain area and so we were stuck with a number of words that were widely recognized as problems, not like roadblocks, but nevertheless sort of things that some people trip over.

[344] And you know that that's always the same thing that people trip over when they trip.

[345] And we could not think of a backwards compatible way of resolving those issues.

[346] But it's still an option to not resolve the issues.

[347] And so, yes, for a long time, we had sort of resigned ourselves to well okay the language is not going to be perfect in this way and that way and that way and we sort of certain of these i mean there are still plenty of things where you can say well that's that particular detail is better in java or in r or in visual basic or whatever uh and we're okay with that because well we can't easily change it it's not too bad we can do a little bit with user education or we can have a static analyzer or warnings in in the parser or something but there were things where we thought well these are really problems that are not going away they are getting worse in the future uh we should do something about to do something but ultimately there is a decision to be made right yes so Was that the toughest decision in the history of Python you had to make as the benevolent dictator for life?

[348] Or if not, what are there, maybe even on the smaller scale?

[349] What was the decision where you were really torn up about?

[350] Well, the toughest decision was probably to resign.

[351] All right, let's go there.

[352] Hold on a second then.

[353] Let me just, because in the interest of time, too, because I have a few cool questions for you.

[354] Let's touch a really important one because it was quite dramatic and beautiful.

[355] in certain kinds of ways.

[356] In July this year, three months ago, you wrote, now that PEP 572 is done, I don't ever want to have to fight so hard for a PEP and find that so many people despise my decisions.

[357] I would like to remove myself entirely from the decision process.

[358] I'll still be there for a while as an ordinary core developer, and I'll still be available to mentor people, possibly more available.

[359] But I'm basically giving myself a permanent vacation from being BDFL, Benevolent, Dictator for Life.

[360] And you all will be on your own.

[361] First of all, it's almost Shakespearean.

[362] I'm not going to appoint a successor.

[363] So what are you all going to do?

[364] Create a democracy, anarchy, a dictatorship, a federation.

[365] So that was a very dramatic and beautiful.

[366] A set of statements, it's almost, it's open -ended nature.

[367] called the community to create a future for Python.

[368] It's just kind of a beautiful aspect to it.

[369] Wow.

[370] So what, and dramatic, you know, what was making that decision like?

[371] What was on your heart, on your mind, stepping back now a few months later?

[372] Take me through your mind.

[373] I'm glad you liked the writing because it was actually written pretty quickly.

[374] It was literally something like after months and months of going around in circles, I had finally approved PEP 572, which I had a big hand in its design, although I didn't initiate it originally.

[375] I sort of gave it a bunch of nudges in a direction that would be better for the language.

[376] So sorry, just to ask, it's async I .O. That's the one or no?

[377] No, PEP 572 was actually a small feature, which is assignment expressions.

[378] Oh, assignment expressions, okay.

[379] That had been, there was just a lot of debate where a lot of people claimed that they knew what was Pythonic and what was not Pythonic, and they knew that this was going to destroy the language.

[380] This was like a violation of Python's.

[381] most fundamental design philosophy and I thought that was all bullshit because I was in favor of it and I would think I know something about Python's design philosophy.

[382] So I was really tired and also stressed of that thing and literally after sort of announcing I was going to accept it a certain Wednesday evening I had finally sent the email, it's accepted.

[383] Now let's just go implement it.

[384] So I went to bed feeling really relieved.

[385] That's behind me. And I wake up Thursday morning 7 a .m. And I think, well, that was the last one that's going to be such a terrible debate.

[386] And that's the last time that I let myself be so stressed out about a pep decision.

[387] I should just resign.

[388] I've been sort of thinking about retirement for half a decade.

[389] I've been joking and sort of mentioning retirement, sort of telling the community, some point in the future, I'm going to retire.

[390] Don't take that FL part of my title too literally.

[391] And I thought, okay, this is it.

[392] I'm done.

[393] I had the day off.

[394] I wanted to have a good time with my wife.

[395] We were going to a little beach down nearby.

[396] And in, I think, maybe 15, 20 minutes, I wrote that thing that you just called Shakespearean.

[397] And the funny thing is, I'm going to get so much crap for calling you Shakespearean.

[398] I didn't even realize what a monumental decision it was.

[399] Because five minutes later, I read that link to my...

[400] my message back on Twitter, where people were already discussing on Twitter, Guido resigned as the BDFL.

[401] And I had posted it on an internal forum that I thought was only read by core developers.

[402] So I thought I would at least have one day before the news would sort of get out.

[403] The on your own aspect had also an element of quite, it was quite a powerful element.

[404] of the uncertainty that lies ahead.

[405] But can you also just briefly talk about, you know, like, for example, I play guitar as a hobby for fun.

[406] And whenever I play, people are super positive.

[407] It's super friendly.

[408] They're like, this is awesome.

[409] This is great.

[410] But sometimes I enter, as an outside observer, I enter the programming community.

[411] And there seems to sometimes be camps on whatever the topic.

[412] And the two camps, the two or plus camps, are often pretty harsh at criticizing the opposing camps.

[413] As an onlooker, I may be totally wrong on this.

[414] Yeah, Holy Wars are sort of a favorite activity in the programming community.

[415] And what is the psychology behind that?

[416] Is that okay for a healthy community to have?

[417] Is that a productive force ultimately for the evolution of a language?

[418] Well, if everybody is batting each other on the back and never telling the truth, it would not be a good thing.

[419] I think there is a middle ground where sort of being nasty to each other is not okay, but there is a middle ground where there is healthy, ongoing criticism and feedback that is very productive.

[420] And you mean, at every level you see that.

[421] I mean, someone proposes to fix a very small issue in a code base.

[422] Chances are that some reviewer will sort of respond by saying, well, actually, you can do it better the other way.

[423] When it comes to deciding on the future of the Python core developer community, we now have, I think, five or six competing proposals for a constitution so that future do you have a fear of that future do you have a hope for that future i'm very confident about that future and by and large i think that the debate has been very healthy and productive uh and i actually when when i wrote that resignation email, I knew that Python was in a very good spot and that the Python core development community, the group of 50 or 100 people who sort of write or review most of the code that goes into Python, those people get along very well most of the time.

[424] a large number of different areas of expertise are represented different levels of experience in the Python core dev community, different levels of experience completely outside it in software development in general, large systems, small systems, embedded systems.

[425] So I felt okay resigning because I knew that the community can really take care of itself.

[426] And out of a grab bag of future feature developments, let me ask if you can comment maybe on all very quickly.

[427] Concurrent programming, parallel computing, async I .O. these are things that people have expressed hope complained about whatever have discussed on on redid async ios of the paralyzation in general uh packaging i was totally clueless on this i just used pip to install stuff but apparently there's pip and in poetry there's these dependency packaging systems that manage dependencies and so on they're emerging and there's a lot confusion about what's what's the right thing to use then also uh functional programming the the ever you know uh are we going to get more functional programming or not this kind of this kind of idea and of course the uh the the gill is a connected to the paralyzation i suppose the global interpreter lock problem can you just comment on whichever you want to comment on Well, let's take the gill and paralyization and asyncIO as one topic.

[428] I'm not that hopeful that Python will develop into a sort of high concurrency, high parallelism language.

[429] That's sort of the way the language is designed, the way most.

[430] users use the language, the way the language is implemented, all make that a pretty unlikely future.

[431] So you think it might not even need to, really the way people use it.

[432] It might not be something that should be of great concern.

[433] I think ASync IO is a special case because it sort of allows overlapping IO and only IO.

[434] And that is a sort of best practice of supporting very high throughput i .o many connections per second.

[435] I'm not worried about that.

[436] I think Async