00:21:56 · Aditya Agarwal
[00:21:56] S1: Yeah. That's a good observation. I mean, obviously, as somebody who was at Facebook pretty early, you know, and I think there were three of us who actually released Facebook's first open source package, but it was all infrastructure. It was basically an RPC layer thing. But you're totally right that the more broad category of essentially how do you do basically content ranking and content like we had there was a lot of research on it in the early 2000s and then everybody stopped because if you can't get access to the newsfeed data, then what are you doing? There's no point, right? Netflix did its Netflix challenge for a while and then they made it closed source and nobody wanted to do anything. This is totally mean, so I think it's actually just really good for us as users that all these open source models exist. And certainly for me, I'm running my personal agents and I think, it was just like a week ago, I don't know, I just out of nowhere, my Hermes agent decided that it was going to take up $600 of Opus 4.8. I'm like $600 a day, that's a little expensive for even a really good assistant. Then it's pretty easy. You swap in kind of like a cheap open source model and you let it run and you see if you're happy with the results and it's like, oh, actually, I can't tell the difference. And it was fantastic. Switching gears into perhaps Snowflake, you know, like, I'm curious, you know, Snowflake, I think technically was the world's ever biggest software IPO. That's correct. That's correct, right? If I remember correctly, that's right. Was an IPO In terms of dollars raised. In terms of dollars raised. Yeah, it was like the world's biggest ever software IPO. And historically, I think you kind of made this point earlier, Snowflake is in the business of writing really complex software and then essentially selling it and deploying it essentially, right? But we also live in an age where software is a lot easier to kind of create, to produce. And so far, we talk a lot about software that we might wipe code. Maybe it's websites. Maybe it's apps. Maybe it's internal tools. Maybe it's dashboards. Maybe now it's like V1 of my startup idea. But we also have people who are now starting to build file systems. We have a couple of people in our portfolio who are like, actually, five of us traditionally a file system is pretty complex. But you know what?
00:24:00 · Aditya Agarwal
We can write a file system because we can kind of do it a lot. So, at which kind of the logical question here is that there's probably somebody out there who is along with the help of kind of these coding agents building a modern kind of analytics kind of like storage and database solution. How do you think about that?
00:24:17 · Sridhar Ramaswamy
[00:24:17] S0: Yeah. There's absolutely someone out there building an analytics solution, building... Well, there's actually Postgres. It's already open source. There's lots of stuff that people are building on top of it. There are also people building CRM systems and every other kind of software that there is. I think every company needs to look really hard at what their enduring value is. It used to be that getting enterprises to adopt your software, to getting your software embedded in how they operated day to day, that familiarity was enough of a moat That they could not, that companies would not think of switching very quickly. I think all of that is changing. So we definitely think a lot about what that enduring value is. A big part of it is getting people to get value from their investments in something like a data platform be a whole lot faster? What do we know about usage? I look at Cortex a key strategic enabler for the future, not because it can get people to implement stuff on Snowflake fast, but it can learn how what a particular company needs to do in order to get the jobs that they want done on Snowflake fast. It has the beginnings of a feedback loop for what does it mean to optimize for a particular customer. I was talking to folks, for example, this is not even a wild theory anymore, but I think all software can be reduced to a handful of skills on a data substrate. And these skills can be self evolving. We have done this for a bunch of our own internal systems. We rebuilt an SRE system from scratch. That's basically Snowflake data. Day zero, there was nothing. At the end of week one, there were like two dozen skills. At the end of week two, there were more skills than I could reasonably parse and understand what was going on. And the structures kept evolving for the last three months. And it's a full on system at this point. And so we think very much in terms of what is the deployment lifecycle? What does value creation mean? What does Snowflake bring to the table in a way that like anyone else cannot? And so my take is that we need to be faster to value than say a Claude Code or a Codex plus some open source systems. That's the kind of competition that we're going to be facing. But the ideas are very much that of think in terms of durable value, think in terms of how does usage of the product make future usage go much faster.
00:26:58 · Sridhar Ramaswamy, Aditya Agarwal
I think like ironically, ad systems are secure. Ad systems are secure in the world of AI because they involve a feedback loop that's largely invisible to the outside world. So all of us need to be thinking in terms of what states if all software can be easily replicated. [00:27:17] S1: It's a good answer, actually. I mean, to kind of almost... You're right. Code can be very easily generated, but historically, time to value in any enterprise software has been low. Like how do you kind of get your solution in there? How do you customize it? How do you tailor it for the particular customer? And now you can actually get in there and these code generation systems can help you with a lot of that. Correct. [00:27:37] S0: That's right. That's And remember the other thing that is also going on that's true for all of us is that it is a reasonable bet again that the music will stop at some point for model improvements. Yeah. Okay, we don't know when, but it will. You just want to make sure that you're next to a chair and that stops. That's part of the reason for building up businesses quickly. [00:27:59] S1: Yeah, yeah, for sure. You started touching upon this figure which is what are businesses that will have enduring value after kind of like software feels as though it is not commoditized but certainly faster to build and cheaper to build. Ad marketplaces are kind of one example. What are some other categories that you're kind of bullish on? Maybe not bearish on. Not bearish on.
00:27:59 · Aditya Agarwal, Sridhar Ramaswamy
[00:27:59] S1: Yeah, yeah, for sure. You started touching upon this figure which is what are businesses that will have enduring value after kind of like software feels as though it is not commoditized but certainly faster to build and cheaper to build. Ad marketplaces are kind of one example. What are some other categories that you're kind of bullish on? Maybe not bearish on. Not bearish on. [00:28:23] S0: I mean, think things that interact with external worlds. We talked about ad ecosystems, which are a kind of marketplace. But I think other marketplaces, whether it's an Airbnb or an Uber or any other kind of specialized marketplace even, I think they have clearly reliable enduring value. I don't particularly think that infrastructure as a category, and I think you would agree, Aditya, that there's anything special about it. It used to be that people are like, oh, SaaS software is easier to rebuild and replace than infrastructure software. I don't really think agents care. For example, with something like a Spark or even a file system implementation, if you have a reference implementation, you can build a new implementation that's compatible API by API with what's already out there. Yeah. I don't know much beyond that in terms of what has long term enduring value. It has to be something beyond just software.
00:29:24 · Aditya Agarwal
[00:29:24] S1: One of the things I was struck by, Prateek, in our conversation before this was just your willingness to learn, desire to wake up. And particularly right now, it's just an incredible moment that every single day, the pace of change means that if you enjoy learning, it's a little bit overwhelming but it's of it's awesome, right? You wake up every day, there's something new and it's easy to be both overwhelmed but also kind of like giddy with the progress. But what is your information regimen? How do you keep up to it? Are you a Twitter guy? Do you read more sub stacks? Do you like to talk to people, the smartest people that you know? How do you get alpha in terms of information that you ingest and then digest?