Y Combinator Startup Podcast
00:49:51 · Jeffrey Morgan
[00:49:51] S0: We went back and forth on this for a lot, which we shouldn't have. We should have just said, of course, we're doing YC. But by and large, starting a company is a really lonely experience. Even if you have a great co founder, and Michael, my co founder was the co founder of my first company, he was my college roommate at University of Waterloo, but it's still lonely. And I think just having a set of peers, even though we did it during the pandemic, just talking to Jared and like five other groups of founders every week really helped you feel less lonely. And I think that's such an important part of it. And then of course, when we finally moved down here and there was no more COVID, the network was just incredible. And the fact that we could meet founders building on open models, building on any kind of AI, you know, we kinda knew that was gonna happen because we had known so many founders from the University of Waterloo who had done YC pre COVID. And they were like, it's really about getting together. And like that was a big part of it. And we knew that was there. And I think that's what made it a no brainer. But also just I think there are a lot of mistakes you can repeat that you don't have to. And what I love about the YC community is how transparent founders are with each other about those. And I still keep in touch with the founder of Docker, who's an investor in our company. And we're able to talk about some of these challenges we saw in the previous generation of companies that we don't necessarily have to repeat or things that worked and we can bring into the future.
00:51:18 · Garry Tan, Jeffrey Morgan
[00:51:18] S2: Yeah, if you just don't repeat one of those mistakes, sometimes is the mistake that would have killed the company. [00:51:24] S0: Potentially, yeah, yeah. Our famous saying, a bunch of our team is from companies that ended up working great and Docker's doing phenomenal now. But whether it's some of our team was early, early at VMware. And there are always ups and downs. And I think just having a group of people around the table who have a collection of those, and also what worked... Actually, what worked is actually even more important. And just being able to, like, have that muscle memory is a big part of it.
00:51:49 · Garry Tan
[00:51:49] S2: Oh, man. I was just thinking about this because we obviously hang out with and work with a lot of 18 year olds or 19 year olds. And then sometimes they're always asking like, well, what should I do? And then I'm starting to realize, like, one of the more important things is if you've never worked on a team that shipped really amazing technology to, like, a lot of people or, like like, just real clear product market fit, like, do that once. Like, even if it's a month, even if it's, like, three months, you would learn more in those three months, because then you know what good looks like. And then without that, it's like... I mean, it's not like it's impossible. Like, people at YC do figure it out because... You know, but it's that much harder. Like, the difference between having seen something that actually works from beginning to some form of, like, this is what the bug database looks like, and this is how we release, and this is the quality that's necessary, and here's, like, the bar that we hold each other to. Having seen that, it just, like, multiplies the chance that people succeed. So it makes sense that, you know, starting off with a co founding team that has seen a lot of that, pretty powerful.
00:52:52 · Jeffrey Morgan, Garry Tan, Harj Taggar, Aaron Epstein
[00:52:52] S0: Yeah. I think it it provides you a set of values you can work around. Especially when you have so much power in your hands with AI, there are just parts of it that AI can help you with, but it won't hold you accountable to it. And And how does software work? And to look at Ollama for example, I'm sure there's versions of Ollama running in the wild from two years ago. How will your software work when somebody falls in love with it and continues using it for two years? Is it still gonna be working? Well hopefully they update to the latest software or it's a cloud service. But I think you build that muscle memory. And we definitely have that from a lot of our more senior engineers on the team who are at VMware or Nicira, for example. But at the same time, I think there are a lot of lessons we learned in the previous generation of DevOps and infrastructure that aren't valid anymore in [00:53:33] S2: the AI world. Oh, Tell us about it. What have you found? [00:53:36] S4: What is not valid anymore? [00:53:37] S0: I think a good example that I classically used is there was this generation of companies called Platform as a Service And and like the 2010s. The Herokus of the world was a great example of this. [00:53:50] S1: I mean, Heroku started out
00:53:50 · Aaron Epstein, Jeffrey Morgan, Garry Tan
[00:53:50] S1: I mean, Heroku started out [00:53:50] S0: as that. Started out as a Platform as a Service. And there's this concept that if you're a layer on top of something else, that you're in kind of a vulnerable position as a startup, which is absolutely not true in the eye world. And in fact, going up the stack can sometimes be even better because you're closer to the customer. In an infrastructure world, that's also the case. And that was like an analogy that we had to like so many of these muscles, we actually had to break building a llama. Another one was, you know, these LMs are never perfect. And like in the systems world, you want everything to be exactly as it's designed to run. It's tested. It's validated. But LMs by definition are not [00:54:25] S2: That's a feature, not a bug. Exactly. [00:54:27] S0: Yeah. It's a feature. [00:54:27] S2: You want it to be a little nondeterministic, I suppose. [00:54:30] S0: And I think from building a team too, it's that with AI now, there are just problems that you don't need to staff as heavily, whereas you did ten years ago, right? If you think about what does your customer support pipeline look like? What does it look like to deliver a cloud service? Like it's a very different world with AI because how do you build a service where no engineer knows exactly how all the code works, which is obviously the case now? So there's just new lessons we're learning going from like a... Some of our team from infrastructure 1.0 in the 2000s to cloud in the 2010s to now the AI space. There are a lot of rules that break.
00:55:07 · Garry Tan
[00:55:07] S2: I mean, you're probably actually doing an incredible service to like, both sides of the ecosystem, and that, like, the end users get this, like, very clean thing that just works, especially, like, the tokens just come out, and they're very clean, and the API makes sense, and it's rational and logical. And then on the flip side, like, I mean, if you don't have a layer like Ollama, I've directly experienced this, where it's like, oh, yeah, the underlying inference provider has a weird error for, you know, if you put this parameter in this way, or it expects JSON, and, you know, it's not documented. It's just like this insane minefield. Like, the agents can kinda figure it out, but you're gonna bang your head into the wall for a couple hours before the agent figures it out. And in the meantime, you're like, this is a terrible experience, you know? And so you're in there probably helping the inference providers fix all these fundamental bugs too.