Invest with AI
00:56:16 · Robert Yang
[00:56:16] S0: Um, I think this is this is definitely like um I guess we talked a lot about like spineless, and I think that's that's a profound issue with um with today's language models, right? They're trained to be uh aligned, and but the specific ways that they're aligned is essentially to kind of just like more or less agree with you, right? Now, of course, if you throw them something that they know are not facts, right? If you tell them like, look, United States is actually full of like lizard, right? Something that's it's then they'd be they'd be like, no, no, no, that's wrong, right? But but for things that are require opinion, right, they don't take they don't take stance, right? And that's that's fundamentally for for the big labs, they have to do it, right? So like imagine like a language model actually taking a political stance, right? It'd be really bad. Like and it will piss off half of half of Americans if it takes any stance, right? Um so so you can't do that, but but of course, a great investor, they have some stance, right? They have some belief in the world, right? Um, and that's how you can hold on to like a good stock for a long time, right? Without like just like, oh, like it's up, it's up like 60% year on year. Let me just like sell it, right? No, no, no. If you believe that this can go like if you if you get NVIDIA right 10 years ago, right, and you're like, I believe in this, I believe in AI, right? You can hold on it forever, right? You can still hedge by like short other things, but um, but you're like, no, NVIDIA, I'm gonna hold on on to it. So that I think requires the agents to really have uh to be opinionated, right, in a way that is grounded in real facts. Um and it takes human influence into account, but fundamentally it's about like being grounded in facts, being grounded in the market, right? And the real like finance of a company, of the industry and what it believes the the future will be, right? Instead of being just like so easily moved by whatever the last prompt from a human they got, right? So getting to agents that can do that, I think is something that that as you know, I'm very excited. Um, and that that is a human quality that I don't think it's we're gonna get it, like at least not in production within the next 12 months.
00:58:51 · Robert Yang, Brett Caughran
But I'm pretty hopeful that in the next two to three years we will start to see agents that not only have opinions, because it's easy to have opinions, you're gonna have them have like a bad opinion very quickly, but right, but it's like its opinions are actually grounded in like real world and it's able to adjust its opinion based on real feedback, but not based on just like any like random human influence. [00:59:20] S1: How do you do that technically? Is that sort of a skills pack? Is that starting to post-train the model weights, like what you know? Uh like what's the what's the pathway? Like, how do you take this generic agent, consumer agent structure and make that useful for the highly specific vertical, yeah, vertical process of the case?
00:59:42 · Robert Yang
[00:59:42] S0: Yeah, absolutely. Absolutely. I think this is a staged process, right? So um so the first thing is is like good skills, right? You and good skills would go such a long way already, right? You just they don't like again, right now they're kind of easily influenced by whatever is in their prompt, right? So just give them the right prompt, right? Just give them the right context, right? You can be like a particular firm with a particular view, and or you can be a particular like a team, right, and a multi-manager um uh hedge fund and and your team has a particular view, right? So so kind of distill these views into a set of skills, right? And give it to the agent, and the agent would already become more opinionated in the right direction, right? Uh, and and you can then continually evolve these skills, um, and then they will be able to still shape their shape their opinion, right? So I think that's that's a first step. And that I think this will go a long way already, right? Lots of people are not doing this. So we're seeing this in our users. Um, I would say our users are already like really some of the best, a lot of our users are really some of the best, best um pros, right, in this business. Uh, and but even among these pros, right, some people are like like really like ahead in like how they use AI, right? They build very elaborate kind of like skill, like like uh collections of skill. And some people are still using it without skills, right? They're kind of just like like um just doing it raw. Uh now giving them like a starter set of skills, right? That's like that's like thoughtful and like professional, and then they can they can like tune it and adapt it to have their view of their thesis, right? I think that's a first step that's gonna get us a long way. Uh now, once you have done that, how do you go further, right? So the fundamentally we can we just gotta realize the the models, this kind of spineless behavior of the model is uh is built into the models through post-training, right? So the pre-training is like, okay, predict the next word, right, from like all the text we have in humanity, right? So that part is just about compressing, compressing the world, right? Like understanding the world. It's it's it doesn't have opinion, it doesn't have like specific kind of behavior yet, right?
01:02:18 · Robert Yang
And then human like the labs essentially post-train them to behave in a certain way, right? So it turns out you can actually have their early evidence for that, um, not just from ourselves, but from the community, that you can like further post-train them to unlearn a lot of these behaviors, right? Um, then you can actually get them to not be so like spineless, right? Um then they can actually have some like stronger opinion, right? So I think that's that would be the second step is like okay, post-training, um like post-training an open weight models to for it to write character-wise, right, like uh personality-wise to to be able to hold their opinion. Uh uh at the same time, knowledge-wise, they can really understand like a specific sector much, much better, right? Um, you can have like an agent that really dig into like like 20 companies, right? 50 companies, they understand everything that's out there about these companies, right? They understand the management, they understand the the personal history of the management, the the relationship between the management, all the related sectors, the historical like performance. They understand all these things and they keep reasoning about these things. And so, so then I think we get closer to uh these agents being really, really useful for investors to the point that they can probably suggest kind of long, short, you know, decisions that turn out to be pretty good, right? Like a human would probably still like take this and make discretionary like decisions, but they would increasingly realize, oh, um, it turns out the agents' suggestions were just like almost just right, right? There's not like I I need to fact check it, I need to think about it, and but but it turned out they're very good. Uh and I think the final stage, right? So there's a skills, there's a post-training. I think the final stages we'll get to a point where we might question the question the entire kind of like the the paradigm, right? The architecture, the way we do training. A great capability of human um is we're able to learn from very, very small amount of data, right? And this is this is opposite from the paradigm of like current AI, right? Current AI is train on and human, like it's just like incomprehensible large amount of data, right?
01:04:52 · Brett Caughran
[01:04:52] S1: You can say, right, like read the whole internet, and yeah.
01:04:55 · Robert Yang
[01:04:55] S0: Yeah, it's like hundred trillion tokens, right? So it's it's uh it's and uh and that's um that's just a very different paradigm. Whereas a human, like kind of from a baby to right now, don't really process that much data, right? Now some people are like, oh, but but how about the human civilization, right? We have processed data that went into our gene, et cetera, et cetera. But your gene doesn't really encode that much information, or at least it doesn't encode the specific information, right? Like your gene is not encoding like the stock price, right? So um so human, despite not having all this information, is able to do great reasoning about like the market, understanding about like the world. And that's a remarkable capability, right? If we have agents that can do that, right, we can imagine every single day the new information that comes in, they'll be able to absorb it and like think about how that should change their investment decisions. I think that would be very, very powerful, essentially having this like truly continual. Learning system. You can build a continual learning system today, right? But but have a truly continual learning system where it's able to see some one piece of information that's like really significant and change its entire like worldview almost, right? It's like, oh, like today is election day, right? I still remember 2016 election day. Essentially, like all the predictions were like Clinton is gonna win and then turn out Trump won't, right? So it's I think I think a good model would be able to take that one bit of information, right? It's literally a single bit, right? It's like a single bit of information, a binary decision. That one bit of information and really like, okay, like let me like reorganize all the knowledge I have processed in the past and change my worldview so that I can really understand what has happened here and really better understand how I should do investment in the future, how do I do decision making in the future? Right. So I think getting to that point is probably longer. It's like maybe five, five years, five to ten years.