AI Can Write Code. Why Isn’t Software Better?

0:00Where the fuck is all the automation? AI is so unbelievably smart, and yet it's so useless at all other stuff. It doesn't matter how much AI coding agents you use, the software actually isn't getting better. Maybe you're riding it faster, it's arguably getting worse. OpenAI has been trying to automate customer service since 2020. What I want instead is smart software. I want to expand what software itself can do, such that things that should be automatable can then be automatable. My favorite thing that you guys say is we built prod, not God. That's so good. Because if we had any other kind of like big lab leader, even if they had joy, they would cover it up.

0:38And then your view is so different. You're like, no, we're going to create a way better world. For nuanced reasons, I don't think we are on the path of RSI. In the SaaS-spocalypse story, AI can write software. But what if the bigger opportunity is putting intelligence inside the software itself? In this episode, Ben Horowitz and Martin Casado sit down with TypeSafe AI founder Diogo Almeida to talk about JEV and a different vision for how AI changes computing. Diogo argues that coding agents make it faster to produce the same kind of software we already have. JEV is aimed at expanding what software itself can do, giving developers a new primitive for turning natural language intent into decisions that programs can actually use.

1:22They get into why reliability matters more than impressive demos. what a new era of probabilistic programming could look like, and why AI might make existing software dramatically more useful, rather than simply replacing it. And beneath all of this is the question that drove Diogo to build TypeSafe in the first place. If AI is already this smart, where is all the automation? Today, we have the founder and leader of TypeSafe, Diogo, with us, who is a bit of a hero to both Martine and me. He is not only building a really interesting product, but creating what we think is a very important movement.

2:05So we're super excited about today. Welcome. Thank you for coming. Thank you. Maybe you can give us a brief on what is JEV, what is TypeSafe, why is it important? Is this a curse-friendly or no? Yeah. Oh, okay. What work are you talking about? Okay, okay, okay. Cool. So I was actually asked for like an elevator pitch, which I tend to ramble on and I don't do well. But like I realized my favorite elevator pitch for Jeff is where the fuck is all the automation? Like this is like so unbelievably tragic. Yes. So much intelligence. AI is so unbelievably smart. And yet so, not that I hate on chatbots or coding agents.

2:44I love them myself. But it's like so useless at all other stuff. And it's tragic. It's tragic that we have so much like diamond in the rough, but not polished for work. But TypeSafe is making AI for software. You know, we want to make AI powerful, not just for humans in the loop, but to actually build real software. And Jiv to us is our first model in this whole space to make it way, way better to like make automation. Yeah. And so it's been interesting because it's kind of caught fire in software world. So one of the things that made us go, what the hell is going on here is like every developer we know is calling us and going, oh, this is freaking awesome.

3:25It's great. It's fast. It's great. Everything's better. Then how does that, because everybody thinks of, well, we've got cloud code. We've got codecs. Don't we already have that? Like, what's the difference? And then how does that lead to real automation? Ooh, I wish I had like some slopped visuals because I have a favorite slopped visual for this. So I like Cloud Code and Codex. I love the description from Gary Tan and them. It's just-in-time software. Incredible way to describe what they're doing. It makes software on the fly, and you can program software in natural language, but it has the same expressive power as software.

4:01What I want instead is smart software. Instead of automating software engineering, I want to expand what software itself can do, such that things that should be automatable can then be automatable. and like in a more flowery language, I want to express things like intent. I want to expand the vocabulary of what we can do. And I can talk about like all sorts of like weird sci-fi things I want. But like programming is like hyper specifying like valuable things and then infinitely replicating them. It's so freaking cool. And I want to just make that more. Interesting. So one way to think about it is instead of kind of a tool that somewhat replaces a software engineer with a faster, maybe not even as good software engineer, what you're saying is, no, no, no, we're going to super empower the software engineers we have to write way, way better, more interesting things.

4:56Yeah, yeah. So this is, by the way, I just think so many people miss this point, and it's such a subtle point, and it's so important to actually tease it out, which is if you use something like Cloud Coder Codex, which is great, or Cursor, which is great, they write code. But that code is the same thing a human being would have write. Maybe it's better, maybe it's worse, but it's basically still code, just like code looked 10 years ago. And the thing with JEV is whether or not you're a cloud coder or a human, you have this new primitive, this new thing that you stick in your code that actually expands the power of software.

5:26So instead of writing code, it is something that you include in your code. Which, by the way, is interesting because it's this very powerful primitive, which would be great if you explain, but it's also a little bit different than how programmers think. For example, like it has this notion of probabilities or maybe... So an intelligent layer inside the software. Yeah, think of like a library that you can use natural language to describe what you want and you give it kind of a state machine and then it will choose what to do with some confidence levels, which we kind of haven't really had before.

5:58Ooh, there's a lot of tricks there. I will jump into one thing first, which is I love the first thing you said in the direction of where the fuck is all the automation. I love software so much. I wish I could be writing it all day. It's, would not recommend being a CEO to people, but whatever. And also it's wild that AI is so cool and software has been unchanged in 10 years. Yeah, exactly. You know, like that to me, like no one can square this together. And the most we can do is add like a little chat bot in the side sometimes that can take actions, but not all actions because some of the actions are not reliable enough.

6:32So I just want to give like that tiny aside. I love the point. I'm going to jump back to the point about like, this is a little bit of a different way to think about it. Yes, I think that machine native doesn't exactly match bits perfectly. And like, that's actually the art form that we are trying to do. In our onboarding on day one, I draw like the Venn diagram of like, what AI is good at, what is valuable in code, we're in the middle. So, you know, we don't output like extrapolated floats, for example, because like AI is just bad at that. But things like probabilities are not exactly novel. and it's similar to the, is Jev just a classifier argument?

7:09Jev is absolutely a classifier. Like classifiers are sick. Classifiers, they're designed to be useful. Yeah, they're designed to be useful. And actually it's the same interface as like some of those ML concepts. Of course. Because these came from like practical people who are trying to make systems work. And what I'm seeing is happening now is that Jev actually, my guess is that Jev probably is better than having an MLE team from 2019. making the stuff for you and you can just program it on the fly. Who knows what could be built? Because there were not that many good MLE teams in 2019 to build narrow things and to be able to collect data sets and measure it and all of that.

7:47And it is just the beginning. By the way, to this point, do you think there's a slider bar here where on one end it's language in, language out, like we have today. On the other end it's an existing imperative program. And then you can kind of move between the two? Or do you think this is the point in the design space, which is language in, kind of state machine out, which is going to solidify as a general purpose thing for programmers. Oh, that's a tricky one. So I will say the answer in my heart. The answer in my heart is that it is a slider. And actually, when I design for the properties we have, I might have made mistakes due to my personal preferences.

8:21But like intelligence per dollar is my North Star right now. And it could be wrong, just to be clear. Intelligence per second might be more valuable in the short term. But even like our interface, like calling the input state, this is intentional. Like it's to say. Oh, that's great. I didn't catch that. It's meant to be the inside of programs. Oh, I love that. So in my heart, because like, so we are really optimizing. A lot of the work I do is for even more complicated arrangements of the internals of program state. Can you put intelligence in there? I think that this is going to be an ever present battle to have.

8:51I, we're very intentional about our design. And also pragmatically, I think certain things happen. Like it's easier to make an AI at these milliseconds. so it'll be more like a database for a while than like a standard library thing. But I would love it to be a standard library thing too. Can I just pull back to you? Like what is the alchemy that creates a Diogo? I mean like you speak... When a man and a woman love each other. You speak like an AI researcher, you speak like a systems person, you speak like a programmer. And normally these things have been like not super overlapping and you're taking AI, which we've been pushing towards being a being and you're making it a programmer's tool.

9:29So maybe a little bit about your personal journey that... My history into AI is somewhat unorthodox. I was a mathlete. I was an award-winning mathlete. The way I describe it is I was good enough at... This is a cringe. I was good enough at math to get girls. So that's quite good. I didn't know that was a thing. Yeah, I did. You have to get quite good. And what kind of girls do you get when you're that good at math? That's an excellent... Whoa, that's... Our audience needs to know. Oh, no. We have to inspire the youth here. Don't do it. Don't do it. It's not worth it. Just be cool and chill and interesting.

10:05And don't overcompensate. Wow, I can't believe I said that. So I was a mathlete, but I actually never, oh man, this also is a little cringe. I never really liked math. I never really tried. I was just like big fish in Little Pond. And to me, math was always the path I was set on, but I hated it. Because it was always about winning competitions. but then computer science is actually a lot like math. It's basically like math, but cool and useful and fun and interesting. And I still love giving algorithms interviews. Is it the best thing for me to do? I don't know, but do I love it? Yes. And does it like allow me to like suss people out really well?

10:44Yes, it does. So I love computer science. I consider myself to be computer scientist much more before AI researcher, despite my history. And like what actually got me into it was I also won a Kaggle competition, not from sophisticated math, but from like just automating like the fuck out of it. You know, like just like more nested loops, more, you know, like it solved like a systems problem. Wow. You know, so that event eventually got me, like I was forced to speak at NeurIPS, normally an honor, but I hated it because I just wanted to be in the mines. Was that from the Kaggle team? Yes. Oh, wow.

11:20Yeah, actually the Kaggle host of it was Isabel Guillon, who was the co-inventor of the SVM. Actually, I think the first author of SVM. I'm not 100 % sure I'm the first author. And she just basically saw that I was like this person who really didn't fit into the research community and then adopted me and showed me, like it got me to meet all the AI people and that, you know, my career was just pushed into that direction. And from there, OpenAI? No, it was like a startup with Jeremy Howard. No kidding. I love Jeremy. Fantastic. Cool. And then Google Brain for a while. And then Retire for a while.

12:04And then eventually I was like just kind of tired of not doing anything. And I was like, you know what? Actually, AI is pretty damn fun. And I joined OpenAI because of that reason. And it worked out really well. Amazing. Really, really well. Incredible. So you said something there that is so unusual in today's world, which is AI is really, really fun. And then the company has such a different demeanor and view of AI than everybody else. My favorite thing that you guys say is we build prod, not God. Because if we had any other kind of like big lab leader, they'd be like trying to, even if they had joy, they would cover it.

12:53and then your view is so different. You're like, no, we're going to create a way better world, and it's going to be awesome, and there's going to be not only are there not going to be less jobs, there'll be more jobs, and there'll be way better jobs, and everybody's going to have a great time. And just being around you, you clearly believe that. So tell us about that. Because for us, TypeSafe, Jeff, it's more than a company. It's a whole movement towards a positive future. that most people in the AI world kind of don't like. Yes. Or they're not with it. I think they don't get it. Yes. You know, like it's just a classifier complaint.

13:33It's like an ML level concern while everyone else is having like a Jeff party. Because it's like, holy shit, like we can do all the things that we wanted to do. And I think if you don't like get developers, it'll be hard to understand what's really going on. So 100%, I agree with that. I do think that there's like a pretty negative world painted that I obviously disagree with. I think it really comes from this like, you know, mono model Kool-Aid that everyone believes. Right, right. One big brain to rule them all. That's one way to, that sounds much more ominous. Yes, yes. But that's what people hear.

14:13For sure, that's what people hear, yeah. But, you know, like, will that one, is that one brain really on the path to rule us all? Like, we have not automated really basic things that I don't think we want people to be doing. You know, like, there's lots of really, really basic stuff. And I think that, oh, man, it pains me when the world is discordant with the reality. And, like, part of the pain is, you know, on the where the fuck is all the automation. Like, how can we have AI be so freaking smart? and like there's so much so much financial incentive to automate stuff like yeah you could make an excuse for diffusion I don't buy it at all I shouldn't name names but like that obviously is not true part of the problem is like the discordance with the reality and the fact that AI has like so much potential is what made it really tragic for me that we had not released this so now it's like a little bit of a party for me but like I was afraid of AI and all dev users are like there's the happy AI the people on Jeff, and then there's the Morose AI, the people who are not.

15:17Yeah, yeah, yeah. It's really, it's quite a kind of fascinating dichotomy. It is really, well, I'll give you, and to your automation point, I had a funny conversation this morning with David George who runs our growth fund because we were talking about the new tools. I was like, have you tried the Muse thing? He's like, oh, it's awesome. I was like, what'd you do with it? He said, I finally canceled my New York Times subscription. And I was like, that is hard to do. But, you know, it's a kind of a, it's a very tip of the iceberg of the things that are horrible things to do that we need to automate.

15:57I think that if we were going to be really intellectually honest and we are really aiming for the North Star of automation, we cannot fall into the same anti-patterns that AI has fallen into, which is really focusing on outliers and demos, right? Like a lot of people ask me, like, what are your favorite use cases? And I'm like, I'm not sure if they work. I want them to work in the background such that like someone would trust that to run and not page them. And like people can build on top of that too. And like, you know. Composable, composable. Composable, but like other things like safe. It's a different type of safety where if you wanted to actually run with resources associated with it, with access to things, you need guarantees for that, or at least statistical guarantees.

16:39So it doesn't go wrong, break it, that type of thing? Well, I don't think our models will be doing that anytime soon unless someone does the software to do that, which would be very cool flex. I should figure out how to give credits for that. But not in a way that we're not responsible. Right, right, right. I'm just curious. How long has this intuition been percolating? Because I remember talking to you maybe in 2017. We did talk about that. Yeah. And then a lot of these ideas were in, you know, you were talking about data being important. You were talking about, like, you want to focus on the task.

17:17And like, but like, so I just like, you know, was this like, did you know that this was going to end up being a classifier? Or was this just an intuition that like, there's just kind of another way to view this entire kind of AI movement, you know? So actually a fun story about that chat in the talk from 2017. I think my talk was actually in a very similar theme. I think it was called something like AI modular in theory and flexible in practice, which is very software. So I'm a little bit consistent with that. I think that this really started right before ChatGPT. Like right when we released these things, I did not have intuition about this.

17:56And honestly, I was not even, I was very, very pleasantly surprised by the generalization capabilities of RLHF. When is this? Must be end of 2021, like fourth quarter of 2021. Like we were, it was really, really general. Like if you read the paper, it's unlike other papers that are like trying to prove their point. It was us actually, you know, scientific method-ish, trying to disprove, like, is it cheating? And, you know, my favorite query was, why is it important to eat socks before meditating? We'd made sure that was not on the internet beforehand. And, like, the models were able to, like, make plausible human-looking answers for this.

18:35And that, to us in the team, was the thing that clicked, like, this is not cheating, which you should always be afraid of cheating in ML. And then what really got me burnt was we released it. you know we did you know I'm obviously a big capabilities guy I did a lot to release that model I really thought that model had like a decent chance of being AGI and when it didn't that was like when my whole world came crashing down and I was like why? So you were kind of on the other train for a bit. Like the crazy train? Well no I'm just like RL generalizes like maybe we have AGI like RLHF generalizes pretty well.

19:16RLVR is the thing that doesn't generalize as well from what I've seen. And AGI in... Well, I was just saying more. I mean, like, you know, you were behind chat GPT. You were behind these early GPTs. That was a very different goal, which is like creating a chatbot that will talk to the human being was not a programmer's tool, et cetera. So I'm just wondering, like... Oh, well, actually, early, early, like 2020 OpenAI, when we talked about AGI, people used to describe it as Ilya in every if statement. So it's not, it's kind of like, but like part, we were talking about OpenAI culture. Part of it is that it's like intentionally vague.

19:55So it's a wide like tent so that everyone can be inside of it. But like, I am not, for nuanced reasons, I don't think we're on the path of RSI. And I still don't think we're in the path of RSI and I did then. I do think that what OpenAI defined as AGI is extremely doable. automating most of the world's economically valuable work actually sounds like, oh man, I don't, like there's a lot of work out there. A lot of it is very rote and simple and like by volume in order to be able to like outsource work, you need like simple instructions that like basic people can do. And as far as I can tell, the intelligence of that has been available in the models for like quite a while now.

20:37And like my, oh man, you know, chip on my shoulder is like, why is this not available? And then since our LHF, the industry just like kind of bifurcated into gigantic overpromise under deliver. I think GPT-3 was actually quite calibrated back in that day. But because humans evaluate how good the models are, it looks really good because they're the judge. But we've been optimizing that judge instead of the automation part. And that has been the missing thing. So I would say that it was really, really then that it like hit me, you know, like, why is this thing not more useful? And so you think that the measure that we should have is to what extent can you automate actual productive tasks?

21:17That the, when you say over-promise and under-deliver, that's the dimension in particular you're talking to. The ability to automate tasks. I, like, I think in my heart, it's like cool sci-fi. You know, and I think that, I think that that is the canary in the coal mine for cool sci-fi. Like, are you really telling me that math is solved or like even like two years ago, GPQA, that Google proof question answering is solved, but we still can't handle a drive through, right? Like, it's a very hard thing to hold in your head at once. And I think a lot of people don't have good answers to that. Can I just test one thing, which may not make sense, but I want to.

21:58I mean, isn't there an argument, though, that, like, the distribution of the real world is different than the digital world, right? It's heavy-tailed. There's a lot of exceptions. We don't have all the data. And, I mean, couldn't it be the case that the reason we're not doing productive stuff in the real world is just, like, we don't have the data for that distribution. We're not training on that distribution. and this is why it's just been basically relegated to like these lower dimensional manifolds like whatever math or code or... I don't entirely buy the data argument in my opinion. I do believe that there's a long tail for sure, like that would be kind of crazy to deny.

22:41And I don't think that in my like canary in the coal mine situation, we need to automate that long tail. Like I think that we need to be incredibly pragmatic on everything. and like building reliable software is always an investment, right? Like, you know, what were the three great virtues of a programmer? Laziness to not to do it again, hubris, and there was a third one. Yeah, no, I remember. This is from the Pearl days. Yeah, there was a third one. Very well, yeah. I wish I could remember it. But like, it's about like the laziness to like spend, you know, like 10 hours to do like the five minute task instantly and to never have to do it again.

23:18like it only make like it should be an ROI decision for people who like automate stuff like I would just like it to be automate a bowl and I think that people will just make like new kinds of work hence the Jev in Jevons new kinds of work once that stuff is doable but like as like a benchmark I feel like it's useful to see can we actually automate the stuff that it really really looks like yeah I should be able to automate OpenAI has been trying to automate customer service since 2020. You know, like, it's, you know, like, it's not. It's just pretty amazing. It's wild, you know, it's wild. I mean, inside companies, there's very little that's automated right now.

24:04And the projects haven't worked. Other than, programming has worked amazing. Can you maybe classify the types of problems you think that are easier to automate now? Because it was kind of interesting. So we've actually looked at support before the current generative wave. And it was interesting. You'd meet a company, and the company would say, we answer 95 % of all, like, you know, like, help desk calls. I'm like, that is so many. But then you actually look at the data. They're all the same. And you realize it's all password resets. And then, like, but if you did it by, like, uniqueness, it was only something like 50 % or something.

24:39So it just feels like when you're dealing with humans and natural systems, like, there's just kind of this very kind of, you know, like a long tail of exceptions. And so to what extent did like every, probably every hour I have somebody ping me and like, I'm using JEP for this new use case. I'm like, I had no idea, you know, like, you know. And so like, to what extent did you even predict like the broad range of use cases for it? Like, did you assume that was going to happen? And have you been surprised by that? Extremely surprised. Did not assume it would happen. This launch was like not something, like if anyone expected this, they are probably insane.

25:15Right? Like it is, I don't think someone could expect a chat GPT for developers because chat GPT was for, you know, like, you know, the normal users. And it's weird. I actually don't even know what percentage of the people who are part of the JF party are developers themselves. I can't imagine non-developers using it. I don't know how they would use it. But even my non-developer friends are just like part of the party and Twitter and memeing and everything like that. So number one, phenomenal number two um this will be hard to convey in this short message because like it's been like blood sweat and tears for years now like the amount i care about reliability is um it's it's a lot like reliability is what this thing is if you don't understand that it'll be very hard to make like a copycat that's benchmarked like it's i feel like every nine of reliability is going to be so valuable for everyone, even if it's not the most valuable thing market cap wise, because it will just enable new applications.

26:21And like we are fighting for like all sorts of like weird nines of reliability that like we don't even fully understand because we are just like, you know, like really getting this like electric motor of AI, of intelligence, like into people's like workstations and they can figure out what to do with it. What does reliability mean in this context is this is just like availability of the model or is it like i call the model and it returns the same thing or like how do i think about reliability yeah so for something that's inherently kind of stuck so uh not so much the former thing and the second thing is closer like i would describe the first thing as kind of like uptime or slas the second thing i would maybe call closer to determinism yeah something thirdly i would consider more like robustness so robustness i would kind of describe as similar intelligence every time.