Price people out of a medication? Why would you do that? That's not a good business idea. You want to maximize the amount of people who could take a medicine. Did you go to college, Hibiscus? Are you a teenager? Should you even be here right now? Do you have your parents' permission? I did. Oh, it sounds like a terrible business move. Why would you make something so expensive that nobody could take it? What's your view on the price of Zolgensma? If you think Zolgensma's expensive, how do you view the gene therapy market? You must really dislike the AAV therapies, or perhaps you have a pharmacoeconomic view of the AAVs and that you believe those prices are justified. I'm really curious to know your philosophical viewpoint here. Do you believe in the value theory of pricing or the cost theory of pricing? What do you think about Vinod's work? Do you subscribe to the Chicago School of antitrust? I think you should talk to HHS. You can go work for the government with your insights and your vast knowledge.
This old dog can still hunt. That's how I feel about the Merck trade. Yeah, this old dog can still hunt. Well, I mean, if you're a pharma analyst your whole life and you can't pick out a couple moves in pharma, how do you like our Venus pet? Peter Thiel co-founded Palantir. I mean, Vertex is more or less cured CF for most people. Do not trade currencies now. We are getting foreign stocks on Godel like imminently. A lot, but if it's consistently going down, I also like that. So it's okay if it doesn't drop in a straight line. These things typically don't drop in a straight line. I haven't looked at Anaplasmosis. I thought they already had a bunch of binaries. I actually bought a million dollars of Bitcoin by accident. I wanted to buy 100K. I like buying Bitcoin when it gets beat up. Some interesting problems.
So I was kind of one of the reasons I was a little upset earlier is that for many years, in fact, I would say decades, there's a press release system. Okay, that's just sort of how news is disseminated. Okay, more and more companies are not using the press release system. It's very easy when a press release comes out to list it in the news section. That's what this section kind of looks like. So for Amgen, this is the Amgen press release, Amgen via PR Newswire. Now we have a deal with PR Newswire, and this came down on the newswire right at the time that the rest of the world got it, the exact second. Follow so far. Now Google doesn't deal with PR Newswire. They just make their own press release and post it on their website. Tesla does this as well. So there's a couple things that you could kind of do here, but our competitor does not have a really satisfactory solution here. And because traders want the news and they want to react to the news right away, like there's no... we're bottlenecked by developers. I mean, all of our developers are very busy. There's sort of two other avenues. It's not easy to just hire new developers. You know, a bad developer can turn your project into a nightmare, can turn a project within a company... a good developer can bring everything forward. You get 10 amazing people, you build a billion-dollar company almost guaranteed. You get 10 developers, three are good, three are okay, and three are bad. Company's going nowhere. You got to get rid of those three bad ones ASAP. I don't know what you do with the three medium. You probably got to get rid of them too. Yeah, so that's what I'm trying to figure out.
Bloomberg does something, but they're actually about a minute late, and the trader wants instant. So what they'll do is they either have an NLP thing or they have a manual. And knowing Bloomberg, I would be surprised if they're doing it manual, to be honest. Crawler is probably too slow. I think you need something... you really can't really download it from their site. I mean, there's a Selenium-style approach. So here's sort of the single source of truth, and you know, it's a custom URL with a UUID. It's like, could you have a Selenium crawler or some other guy downloading this HTML? You know, it's not hard for guys like us to make something unique to that site. Now, of course, they change the site, and you know, you're cooked. Program won't work. There's a couple of other ways to do it. Well, one of the things they do do is they put out... Alphabet or Google put out their 8-K right away. So they put out an 8-K, and these 8-Ks had this press release, but the press release is not distributed with the normal channels such as Business Wire, PR Newswire. Now look, I mean, they're saving $2,000 a quarter. So now there's no requirement that your 8-K SEC filing actually comes out beforehand. I think Tesla does it afterwards, for example. You know, it's a bit of a conundrum. I mean, I don't think it'd be too much latency to have to point it at. In this case, it came out first as an SEC filing. Another system has... they had Dow Jones report it, and I think that was a... you know, that's an automated system. But yeah, to be first, here's how Dow Jones did it. Dow Jones Newswires. Now we talked to Dow Jones. We have a bit of a beef with them, but there's no reason why we can't beat Dow Jones. But yeah, this is sort of a good start. It's basically like, how do you beat Bloomberg and Dow Jones for, you know, the fastest gun in the West? Yeah, the problem is that you're going to have guys that don't do the SEC filing. You know, one of our engineers, Vad, he's actually got a little bit of expertise on proxying scrapers, so that you could hit these sites a little bit and may be able to get... you know, I guess may be able to get to ping the sites, you know, once a second or something. But yeah, I mean, in a lot of ways these sites haven't changed very much in a long time, but it depends on the interface you make. Now Bloomberg also had the SEC filing come up in the feed, but you have to open the SEC filing through it, whereas investors just want to hear, this is the revenue, this is the... so I'll show you what Bloomberg looked like actually, since we're here. Here's what Bloomberg looked like, and you know, this is the gold standard. So you could see here that at 4:01, I don't have the second timestamp, but at 4:01 you had the SEC filing, Google files 8-K, and here's the exhibit. The problem is, does a trader know to open this exhibit? I mean, probably somebody does. And then at 4:01 you had the Bloomberg News notice. They don't have Dow Jones news here, but they have the Bloomberg News with YouTube ads revenue for 10.47 billion versus 10.22, Google Cloud revenue. But the important part, which is EPS, comes down here, and ad revenues over here. But you want the total revenue, which kind of shows up over here for some reason. You can see it's pretty wonky. You know, you have a lot of this that looks like it's a minute or two late. I mean, ideally this guy, EPS and revenue comes... I mean, they didn't have revenue. I mean, I don't doubt that they actually did get this all done within under the minute. We did not get it done at all, basically. We had sort of this, okay, backup, which is a different newswire. And again, because it's not an API, how ours came out. Ours came out here, 4:01, 42 seconds. Nope, 4:02, 28 seconds. And this is the actual press release. So you could not get that on their system basically at all because they don't... unless the press release is like on a newswire. Now you could have clicked this thing, but it's just a weird HTML thing. The problem with this guy is I'm not sure in this case the SEC filing predated the website update because I was refreshing the website. If some of you were watching that, I think we use NewsEdge. That's a Moody's product, right? Well, traders like to read news releases about the companies they care about. If you haven't noticed, that's a big thing in terms of machine-made decisions. For sure, I do think we want to provide that too. Like it would be kind of cool if, like, you know, here it says Amgen reports fourth quarter. You know, if we had that, you know, here we do have it for Google, but basically like just something like sell, you know, in big red, you know, red bold, just says Alphabet disappoints, sell now. And then the even stronger form of that to our system, the SEC filing is different from a news release, one. And then two, it's not packaged in a way that's really that readable right away. Because on Bloomberg, as I pointed out here, what's interesting is they sort of take any new SEC filing and put it in the news deck, which I think we're going to start doing. So that will be real time. I mean, that'll basically be instant. But I think that it's going to look something like this, which is 8-K and Google, GOOG, exhibit 99.1, Q4 2024.htm. This could probably be maybe convert the file name into a title. And then yeah, so the strongest form. So the weak form is you show the user the press release. The stronger form is you show the user the decision. The strongest form is you make an API for the decision, and that the user subs to the API and connects their trading systems to our API that says, okay, within a millisecond or whatever the number is, you can have an estimated impact. Say we were estimating Google will be down 4%, and you take that and you can make the trade immediately. So I think we'll make... we want to make all three of those tools available. Every, you know, people trade differently. Some people don't want... We're going to advertise Godel when it's ready because it'll probably lead to a 10x, 20x user increase, and we just want to be ready for that kind of volume and attention because ultimately, like, it's pretty important for me to get the product right because, you know, our product's still young. Once our product's really solid, like, we have international equities, we have futures now. We've been increasing our price metronomically. I mean, it was 20, then 40, then 60. Bloomberg's $2,500 a month. So I think, I don't know how fast bonds will come, probably slower than international equities, which is planned for this month. Yeah, anyone who buys today locks in for life. Just don't skip your bill. Awesome biotech with Bruno or what's up, Bruno? Maybe Logan, we'll see. I mean, Bloomberg's built a nice business and a great product over 40 years, so I'm not trying to say that they don't have a good product or they don't have a lot of data, which they do. But there are things we can do better. I mean, just pointing out this news thing, the little details really do matter. You know, traders care a lot about those details. You know, if you can really show the trader right away, the only reason to read the press release is because you want to know what's in there versus the market's expectations and what's going to happen to the stock. That reasoning level is already well within the ability of the LLMs. I mean, it was within the ability of LLMs last year. So instead of showing the user the press release, which is a means to the end, just give them the end. The end is Google disappoints, sell. You know, these... I mean, it really is as simple as sell the stock now. You have five seconds before this stock's going to be down a lot. Yeah, and I think you guys saw it earlier. You saw it. It was down. You guys are more perceptive than most. It's a wisdom of the crowd. You guys saw it start to tick down a little bit. It was down like 10 cents, 20 cents, 30 cents. Now it's down $10. You can make a fortune by helping people see that immediately. Yeah, I mean, you could theoretically do it without an LLM and do it with text parsing, and we used to do that on Wall Street all the time. First, we were human text parsers, then we would parse in and stuff like that. But, you know, I think you want beta. It's super alpha if you... you know, the problem is what you used to have to do is program a bunch of stuff. So you basically... let me show you kind of what you used to have to do. You used to have to sort of do this, like, you know, you know what, I could do this in code. You'd have to do something that looks a little like this. You know, if I don't know, you know, you'd write some kind of revenue. You'd get the revenue number however you see fit. And, you know, if you'd kind of almost do like a case where you'd say, I don't know, if revenue is more than 77, you could say buy. But you could also do something like this: revenue is greater than 75 and less than 77. You might say, but remember, there's price sensitivity to all this too. You could also say, okay, expected... this is a little bit more sophisticated of a way to sort of do it. You might say expected return is this, expected return is that, expected return is that. Okay, now here's the problem: is the LLM could do something more complicated than this because imagine it's also if EPS... okay, and now how about if guidance... well, what's guidance? Guidance is a text I have to parse again. So, you know, you got a bunch of stuff to do here that's really tough. And at the end, you sort of combine, you know, all this to make a decision. Now, I will say the SEC website's notoriously slow and notoriously bad downtime. So I would say keep that in mind. You know, we should be able to download it and cache it. Yeah, the subtleties are important. And look, the LLM can zip through that in five, ten seconds. Now, could you do it faster? Maybe. You know, could you do a parser faster? Yeah. And I think you got to maybe do both. I mean, I think the LLM could have made the decision within a second, and you probably could have made $50 million if you had that set up right. I mean, how crazy is that? Maybe five, ten, twenty million. I mean, yeah, let's look. When did the market react? You have that data. Even if you take the one-minute candle, you could have gotten off, assuming you could have gotten off. So about 30 seconds, it dropped $9. Interesting. Be back. My farts here. Okay, they won't let me leverage, so I have to go to the max position I'm allowed to have on Binance is $3 million with 1x leverage and $300K with 6x leverage and $600K with 5x leverage. So it's basically like a cash market. It's hilarious. I run into the max position limits in Binance. Yeah, it's kind of a hopeless disease to target. Make sure you guys sign up for Godel Terminal if you like our stream. Godelterminal.com. Well, I'm cataloging every trial from Arvinis. I don't think they really did a solid Phase 2 with which I can predict, and they also didn't publish preclinical data, so I have to troll through some patents or something. They're doing a lot of DDI studies, which doesn't... you can't really read into it, but it's... yeah, I'm not liking out of the shorts too much. I just added to a couple because I want to maintain exposure. The account value goes up, the value of the position goes down. Now, of course, you know, you worry about a snapback rally. I'm not sure Fcoin is going to snapback rally someday. You know, dog with hat, these are exactly dangerous shorts. I understand in the past they've rallied. It feels like something's kind of permanently changed, especially with Bitcoin, you know, below 100K SMA. Most people, most patients die at one or two years old. Now with these drugs, they live much longer. It's great. Yeah, I helped develop one of the first drugs ever for neurogenic orthostatic hypotension, which is related to POTS. There aren't too many medicines for POTS. I thought Roblox was a little overdone, but kind of missed the bottom there. Didn't get a chance to buy it. You know, tried to develop some other products, and we had some mixed success. Dr. Gupta actually was fairly good. Our TTS product was fairly good, but we had a lot of success with Godel. You know, so at some point you got to invest in the product that's doing the best. I think you got to look at the fully diluted. I mean, if they're about to drop another 10 zillion tokens, then and there's no way that they're not going to drop it, you might as well count all the tokens now. Well, I mean, the DCs mostly own all the Bera, and you know, I think some of them will be patient. They're not that greedy, but I think they're probably locking in a 10x, 20x. So Paps, I short IONQ. I think that's my best, my best one. Yeah, I think that instability, Andy, is a sign of something, probably a much bigger move up or a much bigger move down. Ail, I don't know how that's supposed to help, but there is actually something I'm going to do to the crypto market, and it's downright diabolical. Let's just say by the end of February. Oh, well, let's just say the cycle's going to end. Shut up, Getty. You can't stop it. You can't stop it. Even in my sleep, you can't stop it. Plus, you're supposed to be on my side. We're going to stack the tuna into the ceiling. Might even get you a little cat tree, cat house. I wish, Andy. I used to be that guy. I used to be. I know Aelis, of course. I remember meeting the founder of Aelis. I tried to buy the whole company when it was private. So I am very familiar with many, many companies. But yeah, I used to be... there wasn't any company I didn't know. I don't have that kind of time anymore running Godel. I don't think I've peaked yet. I mean, the nice thing about investing is you can get better over time. You know, I think that experience helps a lot. When I was first starting, I mean, a lot of the stuff I did was kind of stupid, and as I met more people and worked with more people, you know, I got better. So I think I get better every day, and a lot of times you guys help me get better too. Oh my God, look at S&P nosediving. Not the S&P 500. I thought you were talking about the SP, S&P 500, the actual, the token SPX. Like, who's going to build on Bera Chain? For what reason? There's no reason to build anything in crypto on any chain. Why would I build something on Bera Chain? I mean, if I'm going to try to build something in crypto, I'd do it on Ethereum or Solana or something. I'm not going to. And again, I'm a crypto fan. Chain isn't about having a L1, isn't necessarily about the best performance and the best. Then we throw out Bitcoin every day, we throw out Ethereum the next day, and throw out Solana the next day. It doesn't make sense. At some point, you know, it's about user base, and there's a lot of users. And yeah, you could see every transaction, but they're theoretically anonymous. You can't de-anonymize them all. Right. I covered a little Trump, praying for a bounce here. He usually bounces. I mean, this Trump doesn't go straight down. It tends to have a lot of bounces, but it could. I think it's going to end up at five. Like, I don't know if I should short Hyperliquid. Probably. What do you think? The question was, do you not think Trump or others have authority or some say to push his coin or others into circulation? No, I don't think we're going to be using Trump coin. Could they do things to push it up? Yeah, they could. Yeah, Hyperliquid has a big market cap. It's not listed on the EY exchange. If I had to open a crypto short this second, oh boy, maybe Bon, Bera, Trump, SPX, Popcat, WIF, dog with hat. Check those out. You like WIF as a short? Yeah, but it makes us feel silly to some extent, shorting Lays and shorting Retti, because there's actual human beings that work at these companies and they're trying to make something work. It's still overvalued, but you know that nobody works at these cryptos. Now, people work at Bera Chain, and I don't want to disparage them because I'm sure they're working very hard. It's not about that, though. It's about the valuation. Yeah, ION's a good one. I haven't looked at that in a while. The price has dropped a lot.
