00:44:17 · Kris Bennatti, Brett Caughran
[00:44:17] S0: Agreed. And one of the things that's cool, we talked a little bit about cost and efficiency and if you try to run Hudson Labs risk scores out of the box, even for us, it costs hundreds of thousands of dollars to run the Hudson Labs forensic risk score over a ten year period, which is a sort of somewhat interesting element of AI. And one of the reasons that we're now finding it harder to support quant funds, which I think is a pretty, just because, you know, whenever you're updating something in that stack, if you have to rerun ten years of data, becomes harder to offer the state of the art functionality to purely quant funds who need that back testing, which I think is a pretty interesting development in AI where we've all been thinking about like, aren't LLM's amazing for quant funds? But we end up running into these cost constraints where we're really only able to offer state of the art functionality to people who don't need ten years of data. [00:45:35] S1: Yeah. How work? Does I mean, so I haven't tried this yet, this conversation has inspired me to try this. So I'm a healthcare investor. Can I create like a risk skill where it's like a three page risk checklist on a name, pulling in all of my underlying research on those companies, but also pulling in the Hudson Labs forensic risk score, pulling from the 52,000,000 data point vector database via MCP? Like, is that something where I could sort of press that button and get sort of a custom made risk report on any of my companies? [00:46:19] S0: We should chat. Let's chat about, try to get you set up.
00:46:19 · Kris Bennatti, Brett Caughran, Khe Hy
[00:46:19] S0: We should chat. Let's chat about, try to get you set up. [00:46:23] S1: Cool. Alright. Sounds great. Can I [00:46:27] S2: ask about, on this cost question, the, there's so much happening, right? So you have Fable that is 10x more expensive than You have all these, a lot of my clients have become heavy co work users. And guess what? That's not a cheap harness, turns out. Then you have Chinese models, open weight models. How are you seeing this kind of cost? And also, people care about cost. Like, people stopped caring about people didn't care about costs three months ago. And now a lot of people, maybe some of the mega funds excluded, don't care. What's your take on the shifting tokenomics landscape as it relates to investment decision making?
00:47:15 · Kris Bennatti
[00:47:15] S0: Yeah, it's going to be interesting. One thought I have is let's all max out our subscriptions in the short term because they're not going to stick around. One thing that we tell our customers, which is a funny thing to be telling customers, is Hudson Labs subscriptions are profitable for Hudson Labs. So you can, you know, trust that we're not going to randomly hike your prices. Yeah, it's going to be interesting and does come back to this, We're creating these really extraordinary outputs in part by using very inefficient ways of using AI, very inefficient search, very inefficient, like using a million calls instead of one in order to get something you want. And I don't think the fact that it's so expensive and we're not seeing the cost matters. It just means that when this starts to collapse on us and our prices start to go up, there's going to be a lot more, a lot of people are going to start thinking a lot more about AI infrastructure, AI architecture, and what's actually happening on the back end instead of just thinking about that top level LLM, the top level model. That's going to be the next push is we're going to move beyond the generation step and really think about how are we storing data, where are we getting data from. There's so many ways to make all of this more efficient, which is one of the reasons why Hudson Labs can run data sets using long running agents and it, you know, it only costs us a $100. But if you just went to an Opus 4.8 API, you're going to accidentally spend $13. You know, all of these problems are solvable. They just require, they require more.
00:49:34 · Khe Hy, Kris Bennatti, Brett Caughran
[00:49:34] S2: I wonder too, right? Because you mentioned like it's at the app layer, like this cost decision is going to have to be made at the individual contributor layer, the app layer, the model layer, the model selection layer. Like, do you think one I mean, you have a horse in the race, but do you think, like, one of the levers matters most on, like, optimizing against cost or optimizing for cost? [00:50:00] S0: There's a reason open source tools like OpenCode are having such a big moment right now. People who are thinking about cost don't want to get locked in to one model provider. It doesn't make sense. OpenCode a good friend of ours and they're getting huge because developers are frustrated with, oh, if you're using quant code, only getting to use Anthropic models. Anthropic probably won't be king forever. So, yeah, if you're thinking about cost, it's a good idea to make sure that you can switch providers. You're not locking in to one model provider in perpetuity. We know a lot of bigger banks, bigger enterprise customers that have built their own routers rather than relying on somebody else's MCP, so they do have more control. When we think about AI and where we're going, it's, yeah, it's all about being able to control your own outcomes and make your own choices. [00:51:04] S1: We've covered a lot of ground, Chris. Any other core debates that you're having with clients or peers or friends or sort of big questions that you think still need to be resolved that we should talk about?
00:51:04 · Brett Caughran, Kris Bennatti
[00:51:04] S1: We've covered a lot of ground, Chris. Any other core debates that you're having with clients or peers or friends or sort of big questions that you think still need to be resolved that we should talk about? [00:51:20] S0: One thing that's somewhat interesting is, have you tried using any of the big models for extracting soft guidance? [00:51:33] S1: No. Tell me about it. [00:51:34] S0: Or do it. Yeah. So we built a guidance model two years ago now that we thought would only have a moat for about nine months. The reason we bought... Built the guidance model is because LLMs are brilliant, but when they think about tense, they focus... When they think about whether or not something's future or past, they focus on the tense of the verb. So we talked about this before, Bret, where, you know, American Eagle frequently says CapEx is now expected to be. And if you run state of the art models and say, get me all management guidance, it's missing a bunch of guidance. And that appears to still be true, which I find absolutely fascinating, especially because so many of those big model providers have been investing in finance specific training. Now that I've said it on this podcast, it will probably be, fixed tomorrow. But [00:52:43] S1: No. All all 10 listeners to the Invest with AI. [00:52:46] S0: All 10 listeners. Yeah. Yeah. [00:52:47] S1: Although it's probably being scraped into many many agents, I'm sure. Probably. That's interesting. [00:52:55] S0: Yeah.
00:52:55 · Kris Bennatti, Brett Caughran
[00:52:55] S0: Yeah. [00:52:57] S1: Yeah. Fascinating. Okay, great. Yeah. Why do you think, like, I hear this a lot, like the Opus model of agents have more been more specifically finance trained than the ChatGPT. Do you have any sense of why that's been the case? Has that been your experience? [00:53:18] S0: Why it's more finance? [00:53:20] S1: Yeah. Is that just the intention of the training process? [00:53:25] S0: I think a lot of the big firms are now specifically focusing more on finance because A, they realized it was a major gap before and B, realizing it, that this is a consumer base actually willing to pay. [00:53:41] S1: Yeah. Yeah. I think it was a little bit of galvanizing moment when Anthropic did their finance day and showed that finance was the second largest vertical behind tech. [00:53:54] S0: Complexity as well. Yeah, obviously.