First Code, Then AGI: Software’s Event Horizon with Poolside Founders Jason Warner & Eiso Kant

Show Notes

Is the prevailing approach to Artificial General Intelligence (AGI) missing a crucial step – deep, focused specialization? 

For the first time since co-founding Poolside, CEO Jason Warner & CTO Eiso Kant reunite on a podcast articulating their distinct vision for AI's future with our host, Conor Bronsdon. Poolside has intentionally diverged from general-purpose models, developing highly specialized AI meticulously designed for the specific, complex task of coding, viewing it as a direct and robust pathway towards achieving AGI, and revolutionizing how software is created.

Jason and Eiso dive deep into the core tenets of their strategy: an unwavering conviction in reinforcement learning through code execution feedback and the burgeoning power of synthetic data, which they believe will help expand the surface area of software by an astounding 1000x. They candidly discuss the "devil's trade" of data privacy, Poolside's commitment to enterprise-grade AI for high-consequence systems, and why true innovation requires moving beyond flashy demos to solve real-world, critical challenges. 

Looking towards the horizon, they also share their insights on the evolving role of software engineers, where human agency, taste, and judgment become paramount in a landscape augmented by AI "coworkers." They also explore the profound societal implications of their work and the AI industry more generally, touching upon the "event horizon" of intelligent systems and the immense responsibility that comes with being at the forefront of this technological wave.

Chapters

00:00 Introduction and Guest Welcome

01:19 Founding of Poolside

02:56 Vision for AGI and Reinforcement Learning

05:36 Defining AGI and Its Implications

10:03 Training Models for Software Development

17:08 Scaling and Synthetic Data

20:12 Focus on High-Consequence Systems

26:17 Privacy and Security in AI Solutions

28:09 Earning Trust with Developers

31:08 Reinforcement Learning and Compute

34:29 The Vision for AI's Future

39:50 Will Developers Still Exist?

47:07 Poolside Cloud's Ambitions

49:37 Conclusion

Follow the hosts

Follow⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Atin⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Follow⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Conor⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Follow⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Vikram⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Follow⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠Yash⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Follow Today's Guest(s)

Website: poolside.ai

LinkedIn: Jason Warner

LinkedIn: Eiso Kant

Check out Galileo

⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Try Galileo⁠⁠⁠⁠

⁠⁠Agent Leaderboard

What is Chain of Thought | AI Agents, Infrastructure & Engineering?

AI is reshaping infrastructure, strategy, and entire industries. Host Conor Bronsdon talks to the engineers, founders, and researchers building breakthrough AI systems about what it actually takes to ship AI in production, where the opportunities lie, and how leaders should think about the strategic bets ahead.

Chain of Thought translates technical depth into actionable insights for builders and decision-makers. New episodes weekly.

Conor Bronsdon is an angel investor in AI and dev tools, Technical Ecosystem Lead at Modular, and previously led growth at AI startups Galileo and LinearB.

Disclaimer: All views, opinions and statements expressed on this account are solely my own and are made in my personal capacity. They do not reflect, and should not be construed as reflecting, the views, positions, or policies of Modular. This account is not affiliated with, authorized by, or endorsed by Modular in any way.

[0:00] Speaker:
Poolside is an AGI company. We're going after AGI because we wanna build intelligence on a compute. And we saw that there was an opportunity to become one of the four or five companies in the world that could possibly achieve that.

[0:16] Conor Bronsdon:
Welcome back to Chain of Thought, everyone. I am your host, Conor Bronsden, and I'm delighted to be joined by two titans of the AI industry, the co founders of Jason Warner, CEO, and Isso Cant, CTO. Poolside is taking a unique approach to building AI for software development, moving beyond general purpose models towards something that is fundamentally different.

[0:39] Conor Bronsdon:
They've raised $626,000,000 to accelerate the journey towards AGI by focusing on code specifically, and we're gonna dig into that approach. Jason, Isso, thank you so much for joining me. I know it's been ages since you two have been on a podcast together, so it's a unique opportunity for us. It's fantastic having you both here. Thanks for having us. Thank you, Connor. Looking forward to it. Let's dive in. It's really fun having you both here because not only can we dive into some of these deep topics, but I used to listen to you both on your engineering leadership podcast together.

[1:09] Conor Bronsdon:
And so I know that you two have a great dynamic here, so I'm I'm counting on you guys to crush it and, you know, really share your perspective on the AI landscape, what it takes to build critical AI systems. So let's just start at the beginning of Poolside. What was the core problem or opportunity that you both saw in the AI landscape that led you to found Poolside in 2023?

[1:30] Speaker:
I'll get to 2023 in minute. I promise I'll answer this question, but you have to remember, ISO and I met in 2017. So we go all the way back to there where ISO had probably the world's first AI for source code company on the planet. It was called Source, and they were doing something unique and novel and new in the world at the time. And I had just taken over the CTO job at GitHub with an ambition to turn GitHub from basically a collaborative code host into an end to end software on the platform.

[2:01] Speaker:
Very importantly, this next part infused by intelligence and the intelligence was going to be key. So Isilon and met when I tried to acquire his company, didn't work out. We went to go do the podcast later, but we basically bonded over neural networks and their applicability to software because that's what we care about. That's what we think about. That's what we had kind of obsessed over.

[2:21] Speaker:
And it didn't work out from a GitHub perspective in terms of what we wanted to go do. And then obviously GitHub gets acquired and in the office of the CTO with several folks, we ended up incubating GitHub Copilot under a very, very different mandate and very different structure and all sort of stuff. But you go back to that time to speed run over back to 2023,

[2:40] Speaker:
which was, you know, everything in the world was playing out the way we thought it might. And we looked at that and said, we want to go in or do we want to sit in the sidelines still and can basically learn to paint and sail because you could see the end game starting to emerge from there. And ISO and I had always said we wanted to build intelligence on compute. That's even Poolside. Poolside is an AGI company. We're going after AGI because we wanna build intelligence on compute.

[3:07] Speaker:
And we saw that there was an opportunity to become one of the four or five companies in the world that could possibly achieve that. And and there was a if you take back the kind of post to the post chat GPT moment, the world at the beginning of 2023 had the point of view that kind of generally was held that to reach human level intelligence and capabilities,

[3:29] Speaker:
all we have to do is just scale up next token prediction, scale up model size, provide more data, and we are, quote unquote AGI, and we held a very different point of view at that moment in time. We said that yes, skill matters incredibly, scale of compute matters immensely, but the scaling access that we were most excited about was the scaling access of reinforcement learning.

[3:54] Speaker:
And if you go back two years ago, that was not a widely held belief. Actually, a lot of people thought we were kind of crazy for having that belief. So it wasn't just that we saw, Hey, the world is on a trajectory of closing the gap between models and human level intelligence and capabilities. It was also that we realized that we had our own unique point of view on how to get there.

[4:15] Speaker:
And, and kind of that over many conversations, Jason and I had over, over quite a few months culminated into, into deciding to create poolside. I think one thing to reinforce on this is is always funny to me because in '23, when we went out, we were talking about, starting this company with others. People thought we were nuts. Again, just to scale up the cluster. We already understand how to do this. And we

[4:42] Speaker:
knew there was a different path, but, you know, more efficient, more structured, whatever you want to call it, but it was going be different. And the access, as ISO likes to point out, was very different in this way. In '25, it looks more than prescient. It looks like, holy shit. What did we miss in '23? We saw this. And I think this kind of goes to the the ideas

[5:04] Speaker:
around poolside in general, but just generally speaking, what we're going to need to do to scale up to achieve this. Here we are sitting almost midway through 2025. And I don't think I could be more convicted in our approach, but I couldn't be more convicted in the future because it's super clear. Basically, what happens from here to the next, to the next, to the next. And it does involve scaling up compute, but it you need to understand

[5:28] Speaker:
what these things do to understand how to scale. So you mentioned this conviction,

[5:33] Conor Bronsdon:
this belief in this is what happens next. And a lot of that conversation is around this term AGI. But often in those conversations, the people who are on different sides of it may not have the same definition of what AGI means to them. Isol, I'd like to ask you, what does AGI mean when you and Jason are talking about it here? So, we try to always be careful with the term because it's moving target. It's got 50 different definitions.

[6:00] Speaker:
And so, the, the one that I think is most useful for the moment we're in is the world getting to a point where we have human level intelligence and capabilities for the vast majority of knowledge work that we do behind a laptop. And if you break those things down into the three parts of it, I think we are still very much in a moment in time where AI is not yet embodied.

[6:23] Speaker:
We're very early, excitingly, we're early in robotics, we're early in those fields. So the first horizon really is all of the the world of bits that, you know, that we operate and do so much of our work in, and their model. Then if we take it from an intelligence perspective, we have seen early models, you know, with great capabilities and and understanding a lot of knowledge and building this kind of world view and world model

[6:48] Speaker:
as represented by knowledge, but really poor reasoning and thought over long time horizons and complex tasks. Software development is kind of this proxy task for intelligence, Right? It requires you to be able to understand the world, have a lot of knowledge because software development touches everything, but you also need to be able to, you know, go from a high level objective all the way to a full system build that require a lot of reasoning and thought, you know, to get there. And then there's the actual

[7:15] Speaker:
ability to interact in an environment. Right. So the ability to use a computer, the ability to use tools, the ability to really, you know, get to the point that's our ability to be intelligent also is our ability to act. And so I think those three components are the horizon that we're on right now. So I try to usually shy away from the term AGI because it's probably a moving target for the next decade.

[7:40] Speaker:
But I do think if we look at economically valuable work in the world that drives the cost of goods and services down, that pushes the frontier science and technology, that's something we're on the knowledge work side, we're on a trajectory now in the world for solving for that. And I've heard you both talk about this

[8:00] Conor Bronsdon:
and the rationale for why you're focusing on code, and the fact that there is this massive database for code that is out in the world. There's a lot of good code. There's a lot of bad code out there. But there's so many parameters that can be leveraged to take on this problem. And your approach at Poolside has been different than a lot of your competitors. Jason, you've you've used this compelling analogy

[8:25] Conor Bronsdon:
saying everybody is building a four door sedan, but we're building a truck. What makes Poolside's approach fundamentally different from the sedans of the AI space?

[8:36] Speaker:
Well, I think that obviously it goes into as a reminder for folks, if you're not familiar, again, we are in the pantheon of the open AIs and the anthropics of the world and that we are pretraining models from scratch. So there's a lot of things that we do that other folks might not do if they were fine tuning somebody else's model or just using somebody else's model behind the scenes. And so we make very principled

[8:59] Speaker:
day zero decisions on a lot of things that are different than what others would do. And I think going to the analogy, it just means that we're fundamentally thinking about this different because we think about the use case very differently. So, you know, if you want to allow me a second to abuse the analogy of the four door sedan to a truck, you can't slap the brakes from a sedan on a truck. It just won't work.

[9:24] Speaker:
You can't, you can't based upon the same chassis, it won't achieve the results. And so far with general purpose models effectively, what we've been doing is abusing the four door sedan for truck like utility. You know, we've been putting the tow hitch on it. We're putting it on the farm or the work, work site or whatnot. It's because we didn't have a truck. We don't, we didn't know what that looked like. And largely it's indistinguishable

[9:47] Speaker:
from 50,000 feet because they got four wheels and a chassis, you know, so it's hard for some people who might not understand that, but when you use it, you know, difference. If I, if I now throw out another analogy, but I think it's, it's, you know, one of the things that we've always taken from day zero at Poolside is that we we've really tried to look as an intelligence in a human inspired manner.

[10:13] Speaker:
And I think the truck, you know, versus the sedan analogy applied to humans is that we have very different types of intelligence areas where different people are far more knowledgeable, but also have, you know, really sit on different ends of the spectrum of where their skill sets lie. And one of the things that we've really looked in that is that a lot of that has to do with

[10:38] Speaker:
when in your training, when in your learning, you start really branching off and really focusing and making sure the data and the feedback, which is where reinforcement learning work comes from, starts focusing on software development capabilities. So in a world where we have infinite compute and infinite data, we're all the same model, right? We're all at, we're all at this super intelligence model that everything could apply, but we don't live in a world of infinite data or infinite compute.

[11:05] Speaker:
We actually have constrained resources. We can't make a parameter, you know, 10 train parameter model and serve it to users because it, you know, can't, no one can afford to call it. So we're all in a compute budget, compute budget and inference time that people can use. And then at the kind of frontier, we make sure we try and make sure we have similar compute budgets for training. Now, the reason I'm going on this tangent here a little bit is that

[11:30] Speaker:
at the very early part of our training, it looks extremely similar to how you would train any general purpose foundation models. And actually our models share a lot of the data as general purpose foundation models. They're great at planning your trip to London or writing a poem or a bedtime story. They're not just about code or software development. You have to understand the world and build up knowledge and intelligence.

[11:51] Speaker:
But then about a third into the training, we start biasing them massively towards software development. And then once we get through the parts of the training that are really around learning from the data that's out there, what's represented on the web and in data sets that we've built and increasingly more synthetic, to be honest, like very much so, that our view is that, you know, in the future it will likely almost all be synthetic.

[12:17] Speaker:
And then towards the end of the training are really strong efforts in reinforcement learning for code execution feedback, giving models time to think and reason to over complex software tasks across what is now almost a million repositories that we have fully containerized with their full test suite in this huge diversity of domains and languages and getting these models, chance to do tasks and learn from when they're right and wrong. So it's really about like from an infra perspective,

[12:46] Speaker:
a lot of things that initially look similar. If we tomorrow chose to change a couple of dials in our pipelines, the models would become much more like general purpose models. But we've decided to really apply our compute and our efforts to really pushing them towards our software development capabilities. It's not that similar, like maybe a bad, bad way to think about this. It's not dissimilar to whatever budgets we all have in other domains in our life.

[13:11] Speaker:
So we all have time budgets. We all have energy budgets. We all have whatever budgets. So you apply them. So this is a compute budget. We have a, we have an understanding of how we have to apply it, but there's this concept. And again, everyone knows me knows I'm really good at just throwing analogies out there. But kids in life are supposed to be general purpose for a long time. Athletes, kid, athletic kid is supposed to be general purpose. And at some point, if they want to specialize,

[13:38] Speaker:
you do it. You should go pro essentially. Keep keep doing this. But the more the the sooner you overly specialize, you can have you'll have collapsed. You basically have over specialization and all that sort of stuff. But then the late if you never specialize, you basically end up with no advantages. And so you have to actually build this out. So, you know, baseball players are you're supposed to be an athlete than a pitcher. Football players are supposed to be football and basketball and soccer players, and then you could specialize.

[14:05] Speaker:
Yes. And continue to do this. It's actually similar in this way. We all have budgets. They have time budgets, energy budgets, and all that sort of stuff. And at some point, they need to figure out what they're going to go do. So let's get really clear for our listeners here.

[14:19] Conor Bronsdon:
Poolside is purpose building models like Malibu, specifically for software engineering. And you're using this approach of reinforcement learning via code execution feedback to do it. But what advantages does that specialized approach actually offer developers and technical leaders who want to leverage poolside compared to using more generalized AI coding assistance or models? I think it's a, it's a really good question.

[14:44] Speaker:
Our view is that for, for us to succeed as a company, we only succeed if we are constantly pushing the frontier of, of capabilities. At the end of the day, you don't want to work with the dumb model, the dumb agent. It's just that straightforward. So all of us at the end of the day are in a race to human level capabilities. So there's a lot of overlap in the work that I think we all do, but we really push that towards software development. Now the other thing where that we focused on from day zero is we said,

[15:15] Speaker:
we lived in this world, you know, over the last couple of years of models, APIs, applications, all of these things that are kind of being put together. It felt a little bit like the Android world to me, but there hasn't been the Apple equivalent where everything is built from end to end to seamlessly work together. And that's what we've been doing. We've been building the model, we've been building the context engine, the ability for it to adapt and learn from its environment,

[15:41] Speaker:
and the application experience is always one thing. One of the things that my favorite thing to see is when you see someone on the product engineering side coming up with an idea, trying it out, and then realizing, Oh, the models aren't very good at this. And then you see someone from the applied research team jumping in and saying, oh, we can actually improve that in the next iteration. And twenty four hours later, you see a new version coming out.

[16:01] Speaker:
And so it's, it's that deep integration that we focus on. And on top of that, we've taken a first approach to that to try to deliver that kind of experience in the enterprise. Over time, it will become available to everyone. It's our mission and has always been to try to affect every line of code and every developer in the world, everyone who wants to build software.

[16:20] Speaker:
But it's that Apple like end to end building together,

[16:25] Conor Bronsdon:
that I think is quite unique in our culture and our approach that we've taken. I love the grand ambition of what Poolside is trying to do. And and I so I remember hearing an interview from you where you talked about this race and how you didn't want to regret not running as fast as you could in it, which speaks, I think, to how you and Jason are pushing Poolside forward.

[16:46] Conor Bronsdon:
Yet there's this dirty little secret that I've heard Jason talk about, which is that the AI industry, hey, we all pretty much have access to the same data. So how much can you really differentiate on your approach and, and what you're doing at poolside if the actual inputs are largely the same? So,

[17:08] Speaker:
think we live in a, you know, two years ago, and, and, and at least probably when we were talking about this, know, we, you spoke about this world where, yes, we all use the web and variants of it and such, and that's that same dataset. But as we are increasingly moving to scaling up more and more compute into reinforcement learning for our models to learn, We're increasingly living more and more in a world of synthetic data, and where it's about giving the model complex

[17:35] Speaker:
tasks with lots of diversity, where they can go and explore towards a correct solution. And everything from the thoughts and actions that they take, all of this becomes data. And so you end up in a place where you end up building quite a large mountain of proprietary data that really becomes yours as you're, as you're building models. It's not just us, I think that's for everyone in the field.

[17:58] Speaker:
So it's on one hand, you know, if you think about a model at the end of the day, it's a compression of a data into a neural net. And that forces this generalization that we look at as intelligence. Where now I think at the limit, a lot of us will look very similar over the course of five plus years. I think we'll all get good at the things that each other are better at and have advantages on over time.

[18:23] Speaker:
But so it is not just about the model. I think it's the model is one part of it. I think the product and everything becomes really one thing. I think we're starting to see that a little bit in our space as well. But I don't want to understate what happens when you have a team just focused with an immense amount of compute on becoming the world's best in in one domain,

[18:46] Speaker:
and you build out the scale of engineering for that, it becomes a compounded set of advantages. And now two years in, we see these advantages have really stacked up. What you're saying about synthetic data really resonates with me as well. We've seen the same success

[19:01] Conor Bronsdon:
with our evaluation approaches here at Galileo, particularly when working with enterprises to create feedback loops that drive improvement through reinforcement learning. So it's exciting to hear you're seeing the same at at Poolside, and I'll say, would love to to learn. Jason, how many folks are working at Poolside today? How much does the company scale at this point? We have over a 100 people,

[19:24] Speaker:
decent sized go to market team, very large. Most of our folks are obviously in applied research and distributed systems because every if you listen to what we've said also, it's not just a model problem as you understood it to be, but it's a model plus systems problem. And then we're building the middleware and applications as well. So it's all of those things. But, you know, we're two years in. We expect to grow more. We also don't wanna go

[19:47] Speaker:
massive in terms of the team. We're a small but mighty team. We we believe very, very, very much that a small set of highly opinionated mission driven people are going to outperform other people in this market.

[20:00] Conor Bronsdon:
Absolutely. And as Issa mentioned, this opportunity to compound the advantage by this extreme amount of focus is a really interesting approach that has huge potential. And I've also heard both of you express concerns about the focus as an industry on low consequence systems and flashy demos and the danger of applying these learnings directly to banking or healthcare systems,

[20:26] Conor Bronsdon:
enterprises that are where b to b businesses actually make money. Jason, why is this distinction so critical and what are the risks engineers and leaders should be aware of when evaluating AI tools? Iso, I'd love for you to chime in as well. I'm probably one of the,

[20:43] Speaker:
maybe weirdest people to comment on this to a degree too, given my background, CTO at GitHub, VP of engineering at Heroku before that. Same thing at Canonical people make a bunch of Linux. So Ubuntu, Heroku, GitHub. If you look at the history of those companies, most of them started as incredibly virally popular is what they were. And they, all of them were trying to jump to enterprise at some point with varying degrees of success, including varying degrees of success.

[21:12] Speaker:
But with a great go to market team, they were able to. And what I'd see happening at the moment is people are making the mistake of a Flappy Birds clone or an asteroid single shot is saying this is the same thing as building the underlying systems that control money movement, MRI machines, drone software. And it couldn't be further from the truth, in terms of all of those things. The mistake,

[21:39] Speaker:
this is a massive mistake that Silicon Valley has always made in my opinion. And I made this in my own history, which was you could get every single X, you could go to the Y or if you could go, you know, this and you'd always jump over. It's hard to do both of those things. Our, my job, our job, and we'd always said poolside is going to affect every single line of code on the planet, every single developer on the planet. And we're going to start fundamentally

[22:03] Speaker:
with the hardest possible place to achieve that, which is these very large enterprise environments. And you in those environments, if you intimately understand them, you understand that they are not looking to quote unquote vibe code their way to a solution. What they're trying to do is they're trying to achieve an outcome underneath all of these different regimes, whether they be safety or regulatory or compliance or all these internal mechanisms that need to be taken into consideration.

[22:34] Speaker:
And so it's not, has nothing to do with words. It has nothing to do with even like emotions or whatnot. It has to do with this idea of understanding the customer and understanding the user and what happens every day for them. At the, at the root of it, what it is is if someone says, oh, look what you can do with flashy thing in thirty seconds and a one shot over here,

[22:56] Speaker:
massive defense contractors should just do this. It shows a, a, such a staggering misunderstanding of that customer and user that I want you nowhere near them because my I want to sleep well at night. Hot take. But I I think, you know, if if you put that in context where models came from and and where we're heading to with AI, Now we we came from from code completion to chat to now early agentic, and we're on a trajectory to to truly autonomous agents, right? That that look like you're adding to your workforce, like you're adding to your team.

[23:32] Speaker:
And in that environment, we're not far out in the world. It's likely less than three years where you truly have human level capabilities in software development, and you have agents that are collaborating with you in your organization, and in high consequence software environments, the software that, as Jason said, essentially, allows our world to operate on from electricity

[23:56] Speaker:
to banking, to healthcare, that's a huge shift from where people are today. And more so than that, agents are not a static model call. They're going to have to learn from your data. They're going to have to get access to all of your systems. They're going to build up a history of thoughts and actions that they've taken that will be centrally available and other agents can access.

[24:22] Speaker:
That data will be used for versions of models to be fine tuned and to improve. So, we're on this trajectory of AI, in our view, really becoming a coworker and becoming part of your organization. That is where the rubber meets the road in the real world. Now, doesn't mean Jason and don't get equally excited about and have fun vibe coding it out, you know, like it's

[24:46] Speaker:
a 100% there, but where we probably spend more time obsessing over is on the model side on really, how do we push the intelligence and capabilities? How do we allow these models to continually learn in enterprise environments? And then on, on everything that's built around the model, how can you bring this safely behind the firewall of a customer? How can you bring the data?

[25:07] Speaker:
You know, can you bring the model to the data instead of sending off the data to the model somewhere else? Because at the end of the day, this is becoming a highly, highly critical infrastructure for our world and for organizations to use. And I think that's the key thing is that this, this is the end here in the limit. What these things look like is that critical infrastructure and you have to have a certain orientation around it. So, as I said, we use poolside all the time to, to vibe code our way to stuff. It's fine. I do it all, you know, maybe maintain very side projects or things that we're doing. And it's not like it's not good at that.

[25:40] Speaker:
But what I call these these high consequence environments in computer science, there's this term called NP hard. And it's a set of category of problems that if you solve one, solve all. Well, if you solve these environment problems, you solve you can make it work for these folks in these spaces. You can make it work for anyone in those spaces. But if you would just solve

[25:58] Speaker:
the the I wanna vibe code my way to something, the no code, low code is shorter replacement sort of thing right here. You've not solved that or the problem. You still have to do it. But if you can do it over there, you can do it anywhere. Jason, I've also heard you talk about this idea of a a devil's trade that the industry is making, where companies are asked to send private data like their source code to AI providers with this promise that it won't be misused.

[26:24] Conor Bronsdon:
Given that source code is often a company's most valuable asset, how should the developers, the technical leaders listening, evaluate the privacy and security implications of different AI solutions that maybe don't have an enterprise focus yet?

[26:40] Speaker:
Well, I've always said at the when you're when you're in a selling motion for something like this, which is, in my view, this is probably the most important digital technology of our lifetime. So when you're talking about enterprises adopting this, it's very different than a developer signing up with a credit card call and going to try something and stuff they just don't care about.

[27:01] Speaker:
And maybe they care, but they don't really care. But if you think about what enterprises are doing in this motion, as you're really understanding it, you have three people that you're selling to. Three peoples whose requirements you have to satisfy. You've the CTO and CIO, effectively a primary buyer, but you also have the general counsel or the GC, and you have the CISO,

[27:22] Speaker:
the security officer. And there's decisions across the board on what you build and how you expose it. I need to satisfy all three of those people. And it's going to become incredibly more critical in the future as people become aware of what's actually happening as opposed to what they think might be happening. But, you know, the devil's trade I've talked about is I never want to ask a customer

[27:46] Speaker:
to have to send me their most valuable asset, essentially their source code And to quote unquote trust me. I want to show them that I've understood their concerns and I'm meeting them where they need to be, and all of that sort of stuff. Because we've all seen what happens in this space. And it's like, yo bro, trust me on this. And we're going to earn the right. Working with enterprises is earning the right every day to continue to to satisfy them. Earning with developers is not dissimilar, but a developer is an individual. There's no homogenous set of developers and enterprises are the same, but they they typically boil down to similar sets of concern.

[28:24] Speaker:
And so we just need to earn the right to satisfy them and do this. I don't want to ask them to send me a source code. Thankfully, from a technical perspective, this is something that how we're unique. We don't need to get the source code to continue to train better and better models. That goes back to the previous techniques of reinforcement learning, synthetic data. We don't have to ask them to do this. But from a satisfied customer perspective, I don't want to ask them to do this. I also think there's a, there's a second thing here, which is very few products in the enterprise that are products we love.

[28:56] Speaker:
A lot of the products we love are the consumer products. And one of the things that I, I'm, you know, I think is inherent all of us developers is we only want to use the product we really love. We know it doesn't matter if we're at home or if we're the enterprise, like wherever we are, we want to use the product that we think is best that we love. And so by kind of obsessing over, you know, all sides of this, the intelligence in the model,

[29:20] Speaker:
but the user experience is making it work end to end apple like, but then still being able to deliver that in these like highly complicated environments that have, you know, the data locality and security boundaries and, and all of these things that come with it, but what has kind of been our, you know, our, our dream and, and, and what we've been working towards. And it's, it's really nice seeing all of that come together today

[29:43] Speaker:
in these places because it's, it's that trade off is one that you don't want to make. I think the best products are the ones that, you know, you're happy to use anywhere. You're to use them at home and you're happy to use them at work. And I do think that if we fast forward, you know, a couple of years, the form factor is going to drastically change for AI. Yeah. Whatever

[30:04] Speaker:
the form factor is today, which, you know, is heavily, you know, built around editor extensions and editors and CLI tools. The reality is, as AI gets more and more capable and closes more of the gap between what we're capable of doing behind our laptop and what models are capable, the form factor will evolve. But what will actually stay is, is the full system.

[30:26] Speaker:
Everything that lives, that becomes the substrate inside an enterprise, the access to all of the data that's there, the model that is learning from, from the interactions that people have with it, the management of the agents, like all of these things are starting to become that layer while the form factor on top will probably be changing every six months for the next couple of years. And

[30:48] Speaker:
there will be incredibly new form factors thought of by people that are not poolside and that are others, and you want to make sure you can empower them. So it's it's gonna be, it's gonna be a really exciting and and and weird time at the same, the same moment. I think weird is a really good description.

[31:03] Conor Bronsdon:
It's very exciting to see us overcoming some of the underlying technical challenges, parameter budgets, the model serving, and progressing along this road towards human level intelligence at some of these tasks. And Poolside is taking a kind of two lever approach here, leveraging compute, but also focusing heavily on data via reinforcement learning, as you've mentioned, synthetic data generation.

[31:32] Conor Bronsdon:
I'd love to understand a bit more about how this two lever approach potentially accelerates progress compared to

[31:39] Speaker:
just throwing more compute at it. So, so I think there, there are two levers we talk about publicly. So I think that that's, that's one thing unless you want to spill some secrets. I, I, that, that won't be But, but I do think, I do think there's a, and by the way, in the last four or five months, I think the world has, has woken up to reinforcement learning very much so. We've seen it in the first reasoning models that others have brought out. And so

[32:05] Speaker:
I think we are very, very strong believers that the more compute that we can drive towards reinforcement learning for both verified rewards, code execution and math and the things that really push those capabilities and having the world's largest environment there for code execution by several orders of magnitude is this huge advantage. We've also been doing a lot of work in non verified rewards. You know, how do we just truly improve the model's capability of thought

[32:32] Speaker:
and reasoning as a subset of thought? And so, so our view is, is that that's really, that's a scaling axis. That's where the majority of compute will go. But you mentioned, you know, not just throwing more compute at it. I think we've said this before. If you are not investing at training compute at the same scale of of your peers in the frontier, you will fill. And we're we've always been very open and and aware of this reality and and have spoken about this. There is a direct relationship today

[33:06] Speaker:
between intelligence and models across any measure and breadth and the amount of compute that was required to produce them. But the question is where are you applying that compute? Are you just making the model larger? Are you just training it on another epoch of 10,000,000,000,000 tokens? We don't think that's the access while scale and size of model does help quite a bit. Don't get me wrong. And so does more data, but our view is, is that compute is best spent on reinforcement learning. And we've been building that for two years to get us to a place where we'd have the world's largest environment

[33:39] Speaker:
and also have the bread and butter foundation model building really well done. And so now we've got something still to prove. I want also be very open. Like I have a lot of respect. Would say right now I look at, you know, the latest cloud model and say, there's areas where they're stronger than us still. Now our job is to make sure that that in twelve months from now and even earlier is is not the discussion.

[34:02] Speaker:
But we also feel quite a privilege that we get to work on this and that we get to have the resources to scale up.

[34:10] Conor Bronsdon:
I I love that. And I do think there is a really healthy competition happening right now. As you mentioned, you know, Cloud CloudSonet is really good at some things. Poolside is extremely good at other parts of, software engineering. And I think it's so interesting to look at this in the context of this bet, this conviction that you two had back when chat GPT had its major moment in 2023.

[34:36] Conor Bronsdon:
You both said, hey. Look. This is the future. We see where this is going. We see this event horizon you've talked about. And Jason, you brought this up earlier. You have this even deeper conviction today about what the future looks like and what happens next. Tell us about that vision. This is where

[34:53] Speaker:
you can start to be one of those people that talks a little bit too ethereally about the future too. And we always like to root in the moment and practical. But as I alluded to earlier, I do think that this is the most important digital technology. Neural networks are the most important digital technology of our lifetime. The application to solve societal problems, the application to solve hard,

[35:14] Speaker:
interesting problems for us is within our grasp in a short order period of time. We solve a known set of problems today. We can do a certain set of things, and that's fascinating and it's great. But what the potential here is is profound. And so it's it's similar to how you might think about this in a way, which is I have X before me. I know exactly how to solve it. I'm going to go do that.

[35:41] Speaker:
Well, it's the unknowns, you know, I don't know how to solve that problem. Can I apply some compute budget to it effectively? Do I have a, do I have one of these things that allows me to apply some compute budget to it to go after that problem and have it be solved? It's remarkable that we're even talking about this. And that's weirdly where I think of the convergence is happening in the future. That's what we're all talking about here. The privilege of a lifetime is to work on something like that, something so profound that it matters at the societal level. And that's what ISO was alluding to.

[36:11] Speaker:
And I think that this is for us, one one of the most interesting, maybe the the most, impactful things is that you can see viscerally what that means. There's some event horizons out there where none of us can see past, but I look at that set when we start to, you know, even approaching the event horizon, you know, obviously we're working in the domain of software right now. It's pretty obvious what happen will over the next couple of years as you build this out.

[36:38] Speaker:
But you can see the effect in other industries and other other spots too. And I don't know, it's beyond humbling to think that you're one of the five on the planet that's allowed to go after this using these various problems in that way. And, you know, we're just gonna keep doing it one foot after the other on towards the event horizons. Jason talks on it. I really like this notion of known compute budget problems and unknown compute budget problems.

[37:02] Speaker:
And so I think if we play it out over the next three, five, ten, fifteen years, everything today that is a broadly known compute budget effort, doesn't matter if that's accounting, software development, work that we do in large groups of people, many of us do it, and we do it behind a laptop, and in the future, embodied in robotics in the real world. All of those things will lead to

[37:28] Speaker:
perfect automation. I think, I think that's now what that does though, and I think this is where it gets exciting, where it's like, there's that you can just spin up compute, you know, to tackle it is that it leaves an entire field open for us of unknown compute budget problems. And the unknown compute budget problems is this infinite frontier of technology progress.

[37:50] Speaker:
We will never not find more science, more technology that we can build, more exploration that we can do if that's the oceans or the stars. If that's the next, you know, breakthrough in biology, if that's in healthcare material sciences. And so I think what we will see is human level intelligence as a skill up on compute is going to be a resource for all of us that we can get access to and use and scale up.

[38:15] Speaker:
Once we are through the things that drive the cost of goods and services down to their raw material cost, we will start increasingly more and more spending on the frontier science where we don't know how much intelligence it takes. And we will also start building, taking this first primitive of intelligence as human intelligence and really applying it to very specific areas.

[38:38] Speaker:
I think Alpha-four is an amazing example of this, where it's a very specific type of task. I think the combination of human level reasoning and thought and capabilities, combined with some very concrete goals that humans could never figure out. The next breakthrough in material sciences or the drug discovery problem. That's where that combination is going to get very interesting. So I think we will look back on this in ten years and say, oh, it's so cute that we solve human level intelligence.

[39:06] Speaker:
Look at what we now have these super intelligences that we can apply to healthcare and to all of these areas of science. And along the way we will, we will find it very deeply integrated in our lives. It will be normal that there will both be knowledge work done by AI as robots will be doing construction and manufacturing,

[39:25] Conor Bronsdon:
and and we will find our meaning amongst all of this. I do think it's interesting to think about this philosophical question, and it's increasingly one that the industry is, I think, beginning to grapple with as we see the first order impacts that are coming and start to project into some of the second order impacts this could have on society. One of those first order impacts

[39:47] Conor Bronsdon:
is clearly on software development, the area where poolside is explicitly focused. And I'd love to understand what is the trajectory, Jason, that you're seeing for what the day to day for software engineers

[40:04] Speaker:
and technical folks will look like in the next couple of years as this transition occurs? So as we you know, first order effect, that's effectively how you ask this. What is what are those first sets of dominoes to fall? And another way to ask this question might possibly be, do software developers exist in the future? Because we get that question while I'm sure you I think yes for the record. Well, so so I think this is this is not one of those yes or no questions. There's a whole nuanced conversation around this.

[40:31] Speaker:
My simple answer to this without going into too much time on a podcast is going to be developers exist. The relationship to the tasks and jobs change. Totally. We've had this throughout history, throughout all of development history, we've had this. It's no different here. The tool is different. It's massively different, but it doesn't mean necessarily developers go away. But what does change though, I think enterprises themselves, the actual amalgamation of all the people, the enterprises, there's, there's more change there than there is for some of the developers.

[41:04] Speaker:
But what I give recommendation for developers is many software developers' whole identity is wrapped up in the fact that they're software developers. And a lot of folks in that way are, I am a software developer, I'm a Java developer, I'm a Ruby developer, I'm a Python developer, I'm a distributed systems engineer or ML or whatever. And a lot of them are, you know, that I have perfectly memorized or know the ins and outs of the SDKs or the APIs or whatever.

[41:31] Speaker:
Those types of things go away. Those are not the skills that are necessarily the valuable ones in that that future world. But the ones that still allow you the taste, discernment, and judgment, the ability to understand the unknown possible implications, the ask the questions of, like, hey. Does this cover x or y or z? Does this satisfy the customer? That sort of stuff. That stuff exists.

[41:54] Speaker:
And I think that this is kind of one of those big changes. And I also think their day to day looks different. This is where, again, a lot of the things that happen in the moment don't matter long term. So let's just take the moment like poolside of the full stack model middleware applications that sit on top of it. A lot of the application surface area that happens is just this call IDEs today.

[42:17] Speaker:
A lot of that application surface area as developers use AI via them, that surface area in the IDEs is there because model gaps exist today. So as models themselves and the systems that help support them become more capable, the surface area available to the IDEs diminishes. But it's also because the people who are spending the vast majority of their time are not consigned to IDEs the way they were they once were. They move that time to something else. Because if you have 10 developers on a project, you might have 10 human developers, but you also might have a thousand digital developers working on something. It's a very different interaction mode. So this is one when we're talking to enterprises or even developers every day,

[43:03] Speaker:
which is you have to understand the implications of x to understand what's gonna happen for y. And pulling this all apart is a is kind of an interesting overall conversation.

[43:14] Conor Bronsdon:
Iso, what about you? How do you see the the role of technical folks like like yourself changing in these next couple years? I look at

[43:24] Speaker:
AI getting more capable and and and reaching human level capabilities as we're adding to the global supply of of software development. People can build software. But I also think that, you know, software eats the world has barely started. There's this huge surface area and you can just walk around anywhere or walk into the DMV or like anywhere. And you realize that there's so much more software to be built,

[43:52] Speaker:
to actually create a world where we have more automation, we bring down the cost of goods and services for everybody. And so I think the surface area of software can definitely use more software engineers. Now the question becomes is that there's going to be a lot of tasks that are part of a day to day software developer. Where the AI can, can now do it for you in a second that before would have taken you, you know, twenty minutes.

[44:19] Speaker:
And so the, now we are inherently lazy as software engineers, right? Like it's it's in our blood. If I know a system can do it for me in ten seconds and it would take me twenty minutes and that system's offline, I'll go make coffee for thirty minutes. Right? Like, and and so, and we're already seeing that with AI assistance today. Right? If the AI is down and now you have to go do that thing before that, you know, the model can do 10 times faster, you're not touching it anymore. I am guilty of that for sure. And I think we all inherently are. And so it's entirely, so that gap will entirely close.

[44:55] Speaker:
And then it becomes about your agency. Your agency to go tackle problems, to go build things, to find new surface areas where software is valuable. As Jason said, the construct, the change at enterprises and company structures is probably larger than it is for you as an individual. And it's true if you really have held your identity on to the fact that you are the world's best person at optimizing,

[45:19] Speaker:
you know, CUDA code. And now one day you wake up and you realize that that optimization can be end to end, you know, done by AI faster than you. You have to change, you know, you're, you have to let go and change your identity on that. And, and so I think more people will create software. I think the surface area of software will grow still a thousand X. I think the nature of the day to day role will look increasingly more of one where you have agency and are delegating. That's to a single agent or if it's two dozen or a 100, it has to a variable number that you're scaling up and down as, as you're tackling something.

[45:53] Speaker:
You, you find yourself less time in the editor. And I think it was sucker work on the interview this week. I just saw the snippet. I didn't listen to interview yet, but, you know, you look a little more like a tech lead. And I think that's frankly, a

[46:06] Conor Bronsdon:
nice starting point to talk about it. Yeah. I I definitely hear what you're both saying here. And I think there are these distinct traits that are important ones for great engineers that maybe haven't been focused on necessarily by the industry broadly. So, you know, agency obviously is one everyone's talking about, like, you know, just going and doing things, leveraging the compute you have access to. Creativity is obviously one. How are you thinking through a problem? How are you problem solving?

[46:33] Conor Bronsdon:
And then, Jason, you mentioned one that I've been thinking about a lot lately, which is taste. Paki McCormack wrote an article about taste in reference to one of his portfolio companies back in, I think, 2022 before AI had really exploded. And I've been thinking about it a lot lately from the standpoint of really having that defined sense of what does good mean and how to create it is becoming more and more important tied in with the agency that you're talking about, ISO.

[47:01] Conor Bronsdon:
And I think that's one thing that poolside is really optimizing for. You're saying, hey, we're gonna focus on large enterprises. We're gonna nail this. We're gonna have taste in our problem solving. We're going to really make sure we're solving the hardest problem here. But I also know you're planning to make solutions, more generally available with Poolside Cloud. What's the roadmap look like for Poolside Cloud, and how do you see

[47:26] Conor Bronsdon:
your Poolside's truck becoming accessible to a wider range of developers and organizations.

[47:32] Speaker:
Global ambition for Poolside in the near term is every line of code affected or every developer affected some way somehow. So when Poolside Cloud launches an example, power applications of all variety around the world with our state of the art models and things of that nature, as well as give people access directly to our own full stack. So everything that we do all the way up to the editor. But if they've someone wanted to use somebody else's editor,

[47:56] Speaker:
but powered by poolside, great. They can do they can do that. And I think that this kind of goes to, again, what we think is important in the long term is that you understand what's happening in effect, the the the idea of number of lines of code created or manipulated or, number of places where people are actually doing interesting work. Don't mean to talk about it too much. I think it's kind of obvious there. But I also also pointed out that there's a philosophical difference

[48:20] Speaker:
to approach, and we do talk about this internally quite a bit, which is we want to be Apple when many other people might be Android. And so we're not, we're, we're taking a very opinionated approach to this. I mean, the first opinion is expressed in that the domain is important, which does not, let's do general, let's go specific in software to start. And we have very specific viewpoints

[48:44] Speaker:
on how to build out the full stack on top of it. At the end of the day, it's all about customer preference. It's all about users and developers, and developers are highly opinionated. And again, they're not a monolithic block. The Ruby developers versus the Erlang developers versus the Python, very, very different types of communities. So you have to understand

[49:03] Speaker:
how to satisfy some and, all in different ways. And then you have enterprises and everything else in between. So in, in, in some ways it's a massive problem to, to go undertake. But this is also why you build upon twenty years of experience from myself or others in the industry. And you've this, you've dedicated your entire life to developers. You understand exactly what I'm saying, that they are very different than dedicating yourselves to accountants or,

[49:29] Conor Bronsdon:
dentists or doctors. And it's understanding the people themselves and what they care about at the end of the day. I I love it. It's a great note to end on. And and Jason, Iso, thank you so much for both sharing your perspectives on Chain of Thought today. It's been an honor to have you both together for your first podcast interview together since you started the company. I think that's a really cool opportunity for us. For everyone listening,

[49:52] Conor Bronsdon:
where can they go to find more information about Poolside and follow the work you're doing? Poolside.ai

[49:57] Speaker:
is our home page. It's got information on how to, get in touch with us. If you're interested in installing Poolside and trying us, it's first of all,

[50:06] Conor Bronsdon:
thank you. The other is going to be that, just get in touch. We'll figure out how to to to get you in the pike. Thanks so much, guys. We'll link everything, including Poolside's website, in the show notes. I really appreciate you both for the long ranging discussion. I feel like we could have gone for another half hour. I knew I should have booked more time. So to everyone listening at home, we will be sure to have these two back at some point. So be sure to subscribe wherever you get your podcasts, and check out the Galileo

[50:31] Conor Bronsdon:
YouTube channel for more content like webinars, events, deep dives, and, of course, every episode of this incredible chain of thought podcast. And you can watch Iso and Jason and I interact. You can maybe see a little, feature of Iso's dog who, who joined us pretty very briefly on the YouTube. And, obviously, check out the rest of our episodes with our lovely guests. Gentlemen, thank you so much again. This was a ton of fun. Thank you, Connor. Appreciate it.