Inspired with Alexa von Tobel
00:14:39 · Kevin Ryan, Alexa von Tobel
[00:14:39] S0: I don't think my advice is very helpful because the thing about building a startup is it's more about execution. It's a bit like a... If you said, why is LeBron James good? I know what he does. He just... He drives and he shoots and he makes it. So there's no secret. He just does it better than other people. So if I said to you, know, make sure that that senior team is incredibly good, every company you and I see, they think their team is good. And the truth is they're not that great. Right? You can't... Not everyone... Not every CTO is in the top 2% of CTOs. There is a big, big difference. Everyone thinks, oh, I know this guy. He's an engineer. He can be the CTO. Is he... Like, when we started Mongo, you know, the... Our CTO, Eliot Horowitz, who was quite young, was the most brilliant engineer out of the 700 engineers at DoubleClick. And so that's credible. You know? But if I ask you a a... Someone coming in, is your guy really the best out of 700 in a large company? Generally, the answer is no. They're good. They're not that good. [00:15:48] S2: We talk about this a lot, which is like people don't really know what point 1% talent looks like. [00:15:52] S0: Yeah. [00:15:53] S2: You can say that's point... That's top 3% talent, but you're not looking for top 3%. You're looking for top 1%.
00:15:53 · Alexa von Tobel, Kevin Ryan
[00:15:53] S2: You can say that's point... That's top 3% talent, but you're not looking for top 3%. You're looking for top 1%. [00:15:59] S0: And there aren't that many, And you realize it when you launch a company that... And it has to do with talent, to do with drive, it has to do with some intangibles. They have to be able to recruit. I have a company right now that is hiring senior AI talent, and they just had two people turn them down because they said, look, I make $10,000,000 a year at OpenAI or Google, which is a staggering number. So I'm not gonna take your salary offer for 500,000. You know, being able to persuade people to take big pay cuts and to believe in the vision is hard. It's hard. And the best CEOs can do that. [00:16:38] S2: Yeah. You're right. I always say the best CEOs will success into existence Mhmm. Even when they don't necessarily have the ingredients. Yeah. Kevin, I'm so excited to talk about the future with you because you have been somebody in my life where I'd always come back, hey, Kevin, I wanna get your opinion on this. What do you think? There is so much happening in the world right now, and we'll talk more about New York City tech specifically. But I would love just to step back and say, what are the things you're most excited about? What are you seeing? And if you had to give us like a category or two that you're really diving deep into, what are they?
00:17:17 · Kevin Ryan
[00:17:17] S0: I think the two categories would be vertical AI applications. So, you know, the competition for big LLMs is playing out, but, you know, none of us are gonna get involved with that. So you have OpenAI. You have, you know, Facebook. You have everyone, Gemini. But the applications that are in investment banking or, you know, using for pathology or for supply chain logistics in hospital systems, those are being developed right now, and those are big opportunities. They are going to work. And so there are more and there are more to be done there. It's going to just replace all the SaaS software that was implemented in the last twenty years. Think of the market it's going after because it's really going after labor. How big is that market? And that's why AI is valued as high as it is. If soon, if I'm looking out the window right here at an insurance company building that has 50 stories, a lot of people in that building are taking care of claims. They're doing rote work, and the truth is those jobs are gonna be replaced by AI, and so AI will be adding tremendous value. I have a company right now that, you know, just quietly let go of 35 customer support people because 75% of our 5,000 inquiries a day are being handled by AI now. That's gonna ripple through this entire country in so many different ways. The other area that I'm I'm still excited about is in the deep science area. You know? If we step back for a second, fifteen years ago, when you and I met, ecommerce was an interesting space, and Gilt and Zola were companies that did very well there. I think that area is completely mature, and I haven't been involved for seven or eight years even though I know the area well, Because I don't... What need do you have right now? I can order anything I want online. It'll be here in twenty seconds. It's inexpensively priced. I can return it. I don't know how to do better. Whereas we are seeing breakthroughs in robotics. We are seeing breakthroughs in nuclear. We are seeing breakthroughs in psychedelics. These are all science based things that through computing or other factors are changing the world. So the AI and basically real world applications are the ones that changing the future.
00:19:38 · Alexa von Tobel
[00:19:38] S2: By the way, on the AI, we have sort of a point of view, which is when the elephants dance, don't be an ant. So, you know, Microsoft has a billion users. Google has 3,000,000,000 users. Stay the heck out of the things that they're gonna focus on. Yep. I wanna double click. No pun intended on vertical AI. What in vertical AI when you're like how what what makes you comfortable? That whatever you're going after or some, you know, some of the many things that you backed has enough defensibility. Yeah. How do you think about drawing that line?
00:20:13 · Kevin Ryan
[00:20:13] S0: I'll use a good example. I was the first investor in a company called Rogo that is creating a product that investment banks use to supplement or replace the investment banking analyst job that many of us did a long time ago. And so bankers at Lazard and half of the top 30 banks are already in pilot or have rolled it out. It's a company that's probably worth close to a billion dollars now, it's four years old. So what do I look for? I look for the fact that they're using certain proprietary data that isn't even accessible to other people. I look for the fact that most of those companies don't wanna share their data with the large LLMs, because they have their own proprietary data. And the third thing I look for is, are there hundreds of small features that that sector needs that need to be done? In theory, someone can do it, but of course the reality is they don't do They're not gonna get to it. You know, if you... Right now, Sam Altman is worried about one thing. Is his general model better than Gemini or not? And right now, it's not. Yep. He has announced, for example, that he's gonna go after investment banking. But, you know, is that what he's focused on? Is he personally calling on the big banks? Of course not. He's not gonna get to it. You know, the way that DoubleClick beat Google, and Google had to buy DoubleClick, you know, twenty years ago, is because they just... They weren't... They didn't have a 500 people focused on this one area that were from the CEO on down on every single element of the business. You know, we just did a better job. And so that's what I that's what I look for there. I mean, open evidence is a good example. They have done a better job than all these large LMs because they focus on that. Only that.
00:21:51 · Alexa von Tobel, Kevin Ryan
[00:21:51] S2: We're very much mind melded to you, which is very, very specific areas that happen to be big enough surface area, but they're edge enough. We did an autism LLM. Yeah. It's one in thirty four kids. So it actually really matters, but it's not something that OpenAI is gonna come after tomorrow. Yeah. But that also isn't gonna go away, and and it's where the world needs advancement. So basic base... Basically big enough, but specific enough that it's still on the outer edges of the war between the LLMs. [00:22:26] S0: Exactly. And look. Do I think that some of the large LLMs will buy some of these companies? Yes. Yep. You know, if I were Google, would I be thinking about OpenEvidence? If I were Google or Microsoft, would I be thinking about a rogo over time if I think I'm not gonna win? That could happen. You know, I don't I don't have our companies thinking about that. They just need to build the biggest, best, and most profitable company they can over time, and maybe they go public. Maybe they sell. We'll think about that at a later date. Focus on the product. [00:22:57] S2: Yeah. I I I... We're we're mind melded there. Quickly on... Also similarly mind melded too on sort of what we're calling, like, the acceleration of science, which is is really sort of another huge surface area thesis. How do you think about that more specifically in in places where your antenna goes up?