00:30:11 · Sridhar Ramaswamy
[00:30:11] S0: I mean, first of all, I think this is also a time where if you have a day job, especially like I do, you better be ready to put infinite amounts of time. The people that I see routinely have no clue about what is going on in the world of AI around them are the ones that I was saying this when we were chatting and Ruchi laughed, which is like, if you have an eight to five calendar and have a life, AI is not for you. You'll never learn anything anyway. First of all, it just requires a lot of time. I talk to a lot of people both one on one but also on Slack. I have direct connections with a lot of our most progressive engineers at Snowflake and we are constantly going back and forth about, hey, look at this or look at what is possible. I spend a lot of time tinkering with things hands on. I thought of various levels, yes, writing code but even figuring out new ways of doing things. It only took a little bit of tinkering to figure out that you know, Coco, our coding agent, mostly runs on Opus, but I wanted to generate an XKCD deck. I was like, okay. Just... You just need an OpenAI API to generate, like, really good cartoons. And out of that came a deck to teach people at Snowflake how to better use agents to write software. And so just keeping your mind open about what is possible. Yeah. I think it's exhausting, but it's also, I've had more fun at work in the last six months than probably the previous fifteen, twenty years put together. It's just an amazing time to be alive.
00:31:54 · Aditya Agarwal, Sridhar Ramaswamy
[00:31:54] S1: Yesterday was Father's Day. I have three young children. So Ruchi decided to interview my children with a couple of questions. And one of the questions was, What does your dad like to do? So one of them was like, My dad seems to be reading all the time. The other one was like, My dad likes to work. And there was a part of me, I was like, they think of me as a guy who reads and works. And then I was like, but right now is the time to do it because we're just living in these times where you of need to put it in. But it was kind of an interesting moment because it's hard to have the eight to five and then because I think we are just living in super accelerated times right now where you have to make some sacrifices. What is a book that most changed your trajectory? You asked [00:32:38] S0: me this, I don't have a good answer. I'll give you a geek answer. I think it was a trigonometry book my high school teacher gave me. I spent an entire summer solving all the problems in it. I'm ashamed to admit it but [00:32:52] S1: That's a great answer. I mean, now I kind of want to know the book. Okay. This is a good one actually. A belief about search that you held strongly that turned out [00:33:00] S0: to be wrong. People cared. I think there's more more about privacy, definitely about search, that people cared enough about things like the utility or the intrusiveness of ads to want them to actually do something. [00:33:14] S1: You're saying that most people actually
00:33:14 · Aditya Agarwal, Sridhar Ramaswamy
[00:33:14] S1: You're saying that most people actually [00:33:15] S0: Most people don't. Okay. Yeah. Okay. I think convenience trumped convenience accessibility trumped irritation. [00:33:24] S1: Completely agree. [00:33:25] S0: This is also what people say is not what they mean. [00:33:29] S1: Correct. Yeah. [00:33:29] S0: It's And like so like relying on any kind of interview, surveys, etcetera tells you nothing. There are a class of problems that you should not ask people about. You should see what they do instead. [00:33:42] S1: Like stated preferences versus revealed preferences. Absolutely. The thing that Google got right that most people I think still don't really quite grok or still misunderstand.
00:33:42 · Aditya Agarwal, Sridhar Ramaswamy
[00:33:42] S1: Like stated preferences versus revealed preferences. Absolutely. The thing that Google got right that most people I think still don't really quite grok or still misunderstand. [00:33:51] S0: Look, they accidentally landed in the most amazing whatever diamond mine in history. But I think Google got ad driven monetization, especially in the context of search that most people don't give it enough credit for. Yeah. I think piece to have early principles that did things like split value between our users and Google itself. There's a fifty fifty split rule. If you came up with a great new way to optimize ad revenue, Eric would always tell us you should devote 50% of that to improving the quality of the advertising system and 50% Google would take as revenue. In a weird ironic way, I was one of the witnesses in the DOJ trial. And so I got to see, this can often be a cringe worthy experience, but it wasn't thankfully. I saw emails that I had written for fifteen years and the replies. And what struck out was just the incredible thoughtfulness that went into monetization and how we approached it. The ferocious competitive nature, yes, that's also a part of them and me, but that ability to have a really high standard, think is pretty remarkable. And I can assure you, you can look at ads served by pretty much any other major players and they're all wildly more cringeworthy than what Google does even today.
00:33:51 · Sridhar Ramaswamy, Aditya Agarwal
[00:33:51] S0: Look, they accidentally landed in the most amazing whatever diamond mine in history. But I think Google got ad driven monetization, especially in the context of search that most people don't give it enough credit for. Yeah. I think piece to have early principles that did things like split value between our users and Google itself. There's a fifty fifty split rule. If you came up with a great new way to optimize ad revenue, Eric would always tell us you should devote 50% of that to improving the quality of the advertising system and 50% Google would take as revenue. In a weird ironic way, I was one of the witnesses in the DOJ trial. And so I got to see, this can often be a cringe worthy experience, but it wasn't thankfully. I saw emails that I had written for fifteen years and the replies. And what struck out was just the incredible thoughtfulness that went into monetization and how we approached it. The ferocious competitive nature, yes, that's also a part of them and me, but that ability to have a really high standard, think is pretty remarkable. And I can assure you, you can look at ads served by pretty much any other major players and they're all wildly more cringeworthy than what Google does even today. [00:35:20] S1: I want to say we're out of time. I'm going to get Sridhar out of here by five. Thank you for being so candid with your thoughts. I think we live in such interesting times that you coming in and sharing your story, think is very, very helpful. So thank you. [00:35:34] S0: Thank you so much.