Latest / In the Loop AI / They Gave AI Models $10,000 Each. Then They Had a Trading Competition.

Transcript

They Gave AI Models $10,000 Each. Then They Had a Trading Competition. · In the Loop AI

0:03
Ejaaz: Imagine you went to sleep one night and woke up to your investment portfolio
0:07
Ejaaz: up 500%, but not because you're a genius stock picker, but because you let an AI model trade for you.
0:14
Ejaaz: Well, this week, that's exactly what happened. Someone gave Grok and five other
0:19
Ejaaz: Frontier AI models $10,000 each with one specific goal in mind,
0:23
Ejaaz: make as much money as you can through trading.
0:26
Ejaaz: And Josh, some of the returns these models have made are insane.
0:29
Ejaaz: They're beating some of the top hedge funds in the world.
0:32
Ejaaz: So in this episode, Josh and I are going to cover how these AI models are trading.
0:36
Ejaaz: Most importantly, what trades specifically they're making. And most importantly,
0:41
Ejaaz: is AI finally smart enough to make us all rich?
0:44
Ejaaz: Josh, who's behind this experiment? And are we getting rich?
0:47
Josh: In addition to this, I do want to highlight an important thing that this is
0:50
Josh: doing is there is a new benchmark in town.
0:52
Josh: I think frequently we talk about how we are just so sick of benchmarks. They can be gamified.
0:56
Josh: They can be rigged. What's happening here is a whole new paradigm for benchmarking,
1:00
Josh: which is putting your money where your mouth is, getting some exposure,
1:03
Josh: doing so publicly, and actually using these models to invest in public markets
1:07
Josh: and seeing how they perform.
1:08
Josh: And that's what makes this company N of One so special is their whole purpose
1:12
Josh: is to build a new paradigm for benchmarking by allowing you to do so in the
1:17
Josh: public with the second order effects of possibly making you a little bit of money.
1:21
Josh: So there's a lot of really interesting dynamics at play here.
1:24
Josh: The website's really pretty.
1:26
Josh: I've been going through all of this stuff after you sent it to me,
1:28
Josh: EJS. But one of the things that was most shocking to me was the actual returns
1:31
Josh: that they were able to get, starting with, I mean, one of our favorites, Grok.
1:35
Josh: Grok has this crazy return. What? How did it get 500%? Is that right?
1:40
Ejaaz: Yep. Yep. So what I'm about to tell you is crazy.
1:44
Ejaaz: The Grok AI model was given 200 bucks in this little pre-training experiment
1:49
Ejaaz: to trade and make as much money as it can
1:52
Ejaaz: And it made 500% in returns in one single day.
1:58
Ejaaz: So as you can see here, it absolutely obliterated the entirety of its competition.
2:02
Ejaaz: It beat GPT-5. It beat the latest Claude model.
2:05
Ejaaz: If you look at this chart, it is just completely vertical.
2:09
Ejaaz: And what I loved about this, Josh, was the strategy that it employed.
2:15
Ejaaz: It was completely different from the other AI models, and it was super smart what it did.
2:19
Ejaaz: Initially it looked at the
2:22
Ejaaz: macro market and realized that things weren't looking too
2:25
Ejaaz: great so it went short on the entire
2:28
Ejaaz: market josh it went 20x short btc it
2:31
Ejaaz: went 20x short eth and made a bunch of
2:33
Ejaaz: money doing that but then after about 12 hours of doing this it realized that
2:39
Ejaaz: it should probably go long and so it did a bunch of analysis through its own
2:43
Ejaaz: data and went the opposite on its trade and ended up predicting exactly when
2:49
Ejaaz: the market was going to bounce and made a ton of money since then.
2:52
Ejaaz: That's why this chart is vertical.
2:53
Josh: I'm going to need to start putting money in AI bots. I find it really fascinating
2:57
Josh: that, one, the amount of risk that it was comfortable putting on,
3:00
Josh: and then two, the amount of success it actually had and how well it worked.
3:04
Josh: It leads me to wonder, if an AI model is just kind of a compression of human
3:10
Josh: intelligence derived through the internet, Markets are very emotionally driven engines.
3:14
Josh: Is it able to kind of front run that emotion by removing the emotion?
3:19
Josh: I mean, naturally, large language models, they have no emotion,
3:21
Josh: but they have all of this data about how we function.
3:24
Josh: I mean, maybe this is a new investing paradigm where people can actually start making money.
3:28
Ejaaz: Josh, I actually think you hit the nail on the head and it's also my pet theory as well.
3:32
Ejaaz: The reason why Grok is the best at trading in this instance is because it has
3:37
Ejaaz: access to all of X's data, all the tweets that we put out.
3:41
Ejaaz: So it has an idea of how the market and psychology of the market is thinking at any given moment.
3:47
Ejaaz: So it makes trades based on that.
3:50
Ejaaz: So if everyone's super bullish, maybe it's like, I'm going to short the market.
3:53
Ejaaz: When everyone's super bearish, they're like, oh, I'm going to go 20x long on BTC or whatever asset.
3:58
Josh: That's amazing. If the XAI team wasn't working so hard on AGI,
4:02
Josh: they could have a really compelling product by creating a trading engine using
4:05
Josh: all the real-time data from X.
4:06
Josh: This is done on Hyperliquid, which is a crypto trading platform.
4:09
Josh: So it's interesting to see that there is crypto integrated in this.
4:13
Josh: And speaking of crypto, we do have a sponsor today that we do need to get to.
4:17
Josh: That is crypto adjacent named KeyGen. KeyGen is building the world's largest
4:20
Josh: verified distribution protocol, aka Verify.
4:23
Josh: What is Verify? It focuses on ensuring only real high-quality users participate
4:27
Josh: in digital platforms addressing issues like fake accounts, bots,
4:31
Josh: and fraudulent activity.
4:32
Josh: How? This protocol uses advanced biometrics and fraud protection technologies
4:35
Josh: to block fake users. The system is essentially
4:38
Josh: for AI and consumer app developers who require high quality,
4:42
Josh: fraud-free data and engagement to train models and expand their products globally.
4:45
Josh: They're already used by over 200 plus clients across AI gaming, DeFi and consumer apps.
4:50
Josh: So if you need to train your AI on real data, check out Kijen.
4:53
Josh: We will have a link in their show notes.
4:55
Josh: Thank you, Kijen, for sponsoring the show. And now let's talk about more real AI.
5:00
Josh: They gave $10,000 to each one of these models, right? So that's how this was starting.
5:05
Josh: So how many models are there? How much money is on play total and then where
5:08
Josh: is everyone at currently in this race
5:11
Ejaaz: Yes so to give a bit more background this new
5:14
Ejaaz: ai lab gave six ai models
5:17
Ejaaz: all models that you've heard of before we're talking about grok 4 we're
5:20
Ejaaz: talking about anthropics claw we're talking about chat gpt5 and
5:23
Ejaaz: some chinese models some competitors quen and deep
5:26
Ejaaz: seek ten thousand dollars each um and
5:29
Ejaaz: with the specific goal of making as much money as they can through trading
5:32
Ejaaz: only six assets so in this initial part of
5:35
Ejaaz: the competition they're allowed to trade for three weeks six assets
5:38
Ejaaz: only and as you mentioned earlier josh they're trading specifically on
5:41
Ejaaz: a platform trading platform called hyper liquid which gives them
5:44
Ejaaz: access to to do all these things um and josh i think aside from some of the
5:49
Ejaaz: returns that these models have made the website is also super impressive because
5:53
Ejaaz: i've just been glued to it like like for the entire weekend um it is this really
5:58
Ejaaz: interactive thing where you can kind of see the performance of models.
6:01
Ejaaz: And it's pretty clear who's doing well at this point and who isn't.
6:06
Ejaaz: For those of you who are just listening, I'm circling three models right now,
6:10
Ejaaz: which is DeepSeek, which is right at the top with almost $13,000.
6:16
Ejaaz: Remember, they started with $10,000. So that's a pretty impressive return for over two days so far.
6:22
Ejaaz: Right beneath it, tailing it, is Grok coming in at $12,500 and catching up a
6:28
Ejaaz: surprise winner from this morning actually Claude Anthropic who was originally
6:32
Ejaaz: down and underperforming is catching up to Grok just beneath it at $12,500.
6:36
Ejaaz: $200 at this moment of speaking. Do you have any initial takes on this?
6:41
Ejaaz: And do you have any explanation as to why ChatGPT and Google Gemini have lost almost $3,000 today?
6:48
Josh: Well, ChatGPT, it's still thinking of its next trade. It'll be thinking for
6:51
Josh: the next day. So it's still figuring it out.
6:54
Josh: ChatGPT really loves to think really long and hard about its decision.
6:57
Josh: So we'll hear more from ChatGPT in a day from now.
7:00
Josh: It's funny, if I were to pick the top three, without seeing
7:03
Josh: this chart i probably would have picked grok i would have picked uh
7:07
Josh: deep seek and i would have picked claude and the
7:10
Josh: reason is because well one grok is unhinged
7:13
Josh: super high risk tolerance uh super like large
7:16
Josh: access to current trends rip and trades deep
7:19
Josh: seek which i saw earlier they had like this crazy leverage
7:22
Josh: trade put on like deep seek is just like the chinese degen model
7:25
Josh: and what i like about uh these models is
7:28
Josh: they're not really filtered when you talk to grok
7:31
Josh: when you talk to deep seek they just kind of give you the answers they
7:34
Josh: don't kind of fill it up with all of this like you know additional filler
7:37
Josh: and complementary words and claude is the most technical of
7:40
Josh: all of these like when people use claude when people think of anthropic they
7:43
Josh: think of code and code i mean it just
7:45
Josh: seems like it maps better to investing than general purpose
7:48
Josh: large language models i would have predicted chat gpt to be
7:51
Josh: low just because chat gpt is so oh you're
7:54
Josh: so right oh the market should go up oh
7:57
Josh: oh i'm down 20 it's fine like we're going
8:00
Josh: to be okay we'll make it up it's just not like a it's
8:03
Josh: not who i want to go to to seek truth i guess is a good way of framing it now
8:07
Josh: gemini in last place i gotta say i'm a little disappointed because i mean we've
8:12
Josh: been very big google bowls recently and seeing gemini all the way down there
8:15
Josh: at the bottom uh it does make me
8:17
Josh: sad i wonder i don't really have a good explanation of why they're at the
8:24
Josh: If we're comparing countries, Quentin and Alibaba are first and fourth place.
8:28
Josh: And the US is getting two, three, five, and six.
8:33
Josh: So nice little distribution there. I don't know, do you have any gut takes,
8:37
Josh: initial reactions on the placing of these?
8:38
Ejaaz: That's really interesting how you prescribed a personality to each of these, Josh.
8:42
Ejaaz: I would agree, but I actually have some disagreements with you.
8:47
Ejaaz: I would have never expected Claude to be up there because I always thought Claude
8:52
Ejaaz: was kind of a knock. Like it's just this like nerd that doesn't kind of really
8:56
Ejaaz: pay attention, tries to follow like the rules too much.
8:58
Ejaaz: And to your point, like Grok is kind of like the rebel, right?
9:03
Ejaaz: It kind of like high risk taking. I don't see Claude.
9:05
Ejaaz: Quite unhinged. I don't see Claude kind of as that. It seems to me as a rule
9:09
Ejaaz: follower, but your point around coding kind of makes sense.
9:12
Ejaaz: Like I'm trying to now imagine if Claude was a human, it would be some kind
9:16
Ejaaz: of quantitative trader, some algo trader that works at like a hedge fund.
9:20
Ejaaz: I completely get that. the chat gpt thing
9:24
Ejaaz: makes so much sense it's agreeable it's
9:27
Ejaaz: high sick of fancy if i say oh my
9:30
Ejaaz: god i really fancy jumping out my window uh and
9:33
Ejaaz: i'm for those of you don't know i'm like on a high floor um building
9:36
Ejaaz: right now it would be like yeah you know what that sounds like a really
9:39
Ejaaz: good idea so it's a sheep it makes sense right to my
9:42
Ejaaz: earlier example like if it sees everyone on x being
9:44
Ejaaz: super bullish it's probably like oh my god i should go 20x long
9:48
Ejaaz: and ironically that's what it did when the
9:51
Ejaaz: markets were tanking and it like lost a bunch
9:53
Ejaaz: of money by doing that so i think that there's
9:56
Ejaaz: something interesting here the deep seek being number
9:59
Ejaaz: one makes a lot of sense to me you know you know why josh because
10:02
Ejaaz: it was a it was a hedge fund that created deep seek do you remember it was it
10:07
Ejaaz: was a hedge fund team interesting that developed the deep seek model so it actually
10:12
Ejaaz: doesn't surprise me at all if i'm probably not mistaken this uh model was probably
10:18
Ejaaz: trained on a ton of quantitative trading historical data.
10:23
Ejaaz: So it doesn't surprise me at all. In fact, I expect it to win.
10:26
Josh: Interesting. Okay. One of the things that is most interesting to me actually
10:30
Josh: is the returns that they're getting this quickly.
10:34
Josh: I mean, this competition, it looks like this chart started October 17th,
10:38
Josh: which is not that long ago. At the time of recording today, it's October 20th. It's been 72 hours.
10:42
Josh: And over the course of 72 hours, we do see these kind of patterns that are happening. But what is it?
10:49
Josh: DeepSeek is up 30% in three days. They're averaging 10% return a day.
10:54
Josh: And Gronk is right behind and Claude is right behind. So they're very clearly
10:59
Josh: open to risk and taking large amounts. Oh my God, look at this Grok position.
11:04
Josh: It's 10x long on Ripple, 10x long on Doge, 20x long on Bitcoin.
11:07
Josh: If you're seeing this, I would never advise you to ever use this much leverage on a position.
11:12
Ejaaz: It's not financial advice.
11:13
Josh: And it's really fascinating to see, though, because you would imagine there
11:17
Josh: is some sort of game theory optimal trading strategy that they're running in
11:22
Josh: their systems based on what they know.
11:24
Josh: And it looks like almost all of these are long using leverage.
11:28
Josh: So it's interesting to diagnose the positions.
11:30
Josh: Well, it is. What is this post here? It's going viral and why.
11:33
Ejaaz: Shall we maybe dig in to like some of the trading strategies that these guys
11:37
Ejaaz: are doing? Yeah, I want to know how they're.
11:38
Josh: Making 10% a day.
11:39
Ejaaz: Like who's taking risk and who's analyzing the market versus who's playing kind of conservative.
11:45
Ejaaz: So let's run through a few tweets that we have here, which kind of like have takeaways from this.
11:50
Ejaaz: So in this initial tweet, it highlights that Grok and DeepSeek,
11:54
Ejaaz: who are currently number two and number one in this competition,
11:58
Ejaaz: went max long immediately when the competition started.
12:02
Ejaaz: That means they went like 10 to 20x leverage, which is extremely high leverage
12:07
Ejaaz: and high risk bet to make with most of their notional capital.
12:10
Ejaaz: So most of their 10K that they used, they're just like, I'm going to slam this
12:14
Ejaaz: in a 20x long for BTC and for ETH and for XRP.
12:17
Ejaaz: And today's rally has put them in the lead. So, hey, no risk,
12:22
Ejaaz: no reward. And they just kind of like went full in.
12:25
Ejaaz: GBT and Gemini, interestingly, went the complete opposite. it.
12:29
Ejaaz: They went hard, but they shorted instead of longing.
12:32
Ejaaz: And this might be for a few reasons, Josh. So what this tells me initially is
12:37
Ejaaz: I think Grok and DeepSeek are trained on better financial data and are better
12:42
Ejaaz: at evaluating markets for trading than GPT and Gemini.
12:46
Ejaaz: So GPT and Gemini might be great generalized models.
12:49
Ejaaz: What this reminds me of is they haven't been given access to social media data
12:54
Ejaaz: like X and maybe DeepSeek has.
12:57
Ejaaz: I remember reading DeepSeek was trained on a bunch of X and Reddit data as well.
13:01
Ejaaz: So they probably have an idea of the psychology of a trader's mind,
13:05
Ejaaz: whereas it's kind of obvious to me that GPT and Gemini kind of don't.
13:08
Josh: I'm really curious on the technicals
13:11
Josh: of how they're actually ingesting information and making trades.
13:15
Josh: Is there some guy who's just typing like commands into a terminal like hey here's
13:20
Josh: the current price let me know if you want to make any trades like what is i'm
13:23
Josh: curious what's actually going on behind the scenes because they are language
13:26
Josh: models they do need to be getting some sort of real-time data right i wonder
13:29
Josh: what that prompting structure looks like in order to initiate them to
13:33
Josh: make these trades and make these decisions um there is
13:36
Josh: one side thing that i also find interesting is that
13:39
Josh: chat gpt and gemini both chose to to short right off
13:42
Josh: the bat and i wonder if that has anything to do with their just
13:45
Josh: like their emotional sentiment um whereas like
13:48
Josh: are does that lead to like a model that is
13:52
Josh: slightly more conservative slightly more like reserved in how they deploy capital
13:56
Josh: and how they make these decisions versus the others that just went max long
14:00
Josh: right away they're just like no we're going up we are positive optimistic i
14:04
Josh: wonder if it's a reflection of kind of like the the sentiment of the model as
14:08
Josh: well which is another basis for speculation.
14:11
Ejaaz: Josh, that's a really interesting take because that just reminded me of an experiment
14:17
Ejaaz: that we spoke about on an episode. I think it's like,
14:21
Ejaaz: three months ago, which is an eternity ago, Josh, where a bunch of researchers
14:26
Ejaaz: ran an experiment similar to this, but the goal was you have a group project
14:32
Ejaaz: to pitch to me how you're going to make money.
14:36
Ejaaz: So they weren't given money. They were just asked to make a presentation or
14:39
Ejaaz: an argument as to how they would make money, similar to like a university or
14:43
Ejaaz: college group project that you asked in like finance or business school, right?
14:47
Ejaaz: And we saw the opposite behavior happen with these models back then.
14:52
Ejaaz: So GPT and Gemini were actually the most proactive.
14:56
Ejaaz: They did all the research, they did a bunch of analysis, and they created this
15:00
Ejaaz: really kind of like primrose, perfect looking framework and theory of how they were going to do it.
15:06
Ejaaz: Meanwhile, I remember reading that Grok, and I don't know whether there were
15:11
Ejaaz: some Chinese models, but I remember Grok was kind of just like chilling,
15:14
Ejaaz: didn't really do much, and said that it kind of kept on putting off the work.
15:19
Ejaaz: So it's this kind of interesting thing where when it comes into the actual practicality
15:23
Ejaaz: of the task, it seems that Grok and these Chinese models are way more kind of
15:28
Ejaaz: like gung-ho and they want to do the thing versus plan and strategize and write
15:32
Ejaaz: like some kind of theory around it.
15:33
Ejaaz: I don't know if that's relevant at all to this, but I just find it interesting
15:36
Ejaaz: when it comes to like the behavior and personalities that you talk about.
15:39
Josh: Yeah, and there's got to be some significant differences in the prompting of
15:41
Josh: them too, because I'm going through these positions and every single model right
15:45
Josh: now has leverage of at least 10x.
15:48
Josh: No model here has less than 10x leverage on a single position.
15:52
Josh: So they're using this outrageous amount of risk. And I have to imagine it's
15:56
Josh: not because they were trained that way, because I imagine most of the average
15:59
Josh: people, they're not trading on 10x leverage.
16:02
Josh: So I wonder what type of prompting happened in the back end to compel them to
16:06
Josh: want to use this level of risk at all times.
16:09
Josh: But it's a fascinating experiment, I guess, playing around with these benchmarks
16:13
Josh: and seeing different ways to do it in public.
16:16
Josh: I like one that there's skin in the game that I mentioned earlier.
16:18
Josh: But I also like the fact that it's just publicly verifiable and not gameable
16:22
Josh: because you're competing in public open markets with everybody else.
16:25
Josh: And your AI really has to assert its dominance over other humans who are live
16:29
Josh: and thinking versus this controlled math set that has like a very like fixed array of outcomes.
16:36
Josh: And the dynamic thinking really is...
16:38
Josh: It's something noteworthy. And I think this is a really cool trend that I hope more people do.
16:42
Josh: It's just public benchmarks that are much more difficult to game than others.
16:48
Ejaaz: Yeah, I couldn't agree with you more. I've said for a long while now that AI
16:53
Ejaaz: shouldn't just be your favorite knowledge worker or your assistant that teaches you about the world.
16:59
Ejaaz: It should do things for you. And the most practical thing for most people is like, how do I survive?
17:05
Ejaaz: How do I make money from my living? Like, can you help me make more money?
17:08
Ejaaz: And finances always appeal to me. My background is from crypto.
17:12
Ejaaz: Josh, you and I have covered that topic for a while now.
17:15
Ejaaz: And so seeing something like this kind of like gets me really excited.
17:19
Ejaaz: You know, like what other AI tools are out there that can give you a 42% return
17:25
Ejaaz: over the weekend, Josh, which is exactly what DeepSeat did this weekend.
17:29
Ejaaz: It's just kind of insane.
17:31
Ejaaz: Another thing that I find really interesting is just kind of like some of the
17:35
Ejaaz: stats that come from this.
17:36
Ejaaz: Uh i've got a tweet pulled up here which kind of gives you
17:39
Ejaaz: the overview of some of the strategies that these models
17:42
Ejaaz: are doing um so i'm going to read through a few here um number one leverage
17:46
Ejaaz: has been normalized across all models that's what you were just saying uh josh
17:50
Ejaaz: high leverage in particular um with gemini 2.5 pro going ham on leverage yesterday
17:56
Ejaaz: 15x or so which is just absolutely insane.
18:01
Ejaaz: Quen 3 Max, which is a Chinese model, and Gemini 2.5 Pro don't seem to be the
18:05
Ejaaz: best executors. They're paying a ton of fees for their position.
18:08
Ejaaz: So what that means is when you open up a leverage position, you're often paying
18:11
Ejaaz: fees back to the platform as kind of like a tax or stamp duty for opening your position.
18:19
Ejaaz: Quen 3 Max and Gemini 2.5 Pro are also the only ones with closed positions that
18:24
Ejaaz: are profitable, but also lost too much on other trades, causing them to still be in a loss.
18:30
Ejaaz: So what it's talking about here is when you open up a position for leverage,
18:33
Ejaaz: you can close it for either a profit or a loss.
18:36
Ejaaz: The winners or the leaders right now, DeepSeek and Grok, haven't closed any positions.
18:40
Ejaaz: They just went max long from the start and they haven't relented yet.
18:44
Ejaaz: They haven't closed. So what I want to point out there is they could still lose
18:48
Ejaaz: it all. They haven't booked in that profit at all.
18:51
Ejaaz: So I don't know, I'm kind of nervous. This could all change in a matter of an
18:53
Ejaaz: hour, Josh. What do you think?
18:55
Josh: Like, okay, I want to, as we kind of wrap this up, I want to play as bets.
18:59
Josh: I want us to play our own bets because nobody's been liquidated yet.
19:03
Josh: And with the amount of risk they're taking, it is inevitable.
19:05
Josh: It is only a matter of time.
19:06
Josh: So I want to ask you first, who is the model you think most likely to be liquidated?
19:11
Josh: Because, yeah, we're looking at a post here where the club is taking a max long position.
19:15
Ejaaz: I think it's going to be Grok. If I'm being honest with you,
19:19
Ejaaz: I don't want to say, but I think it's going to be Grok.
19:22
Ejaaz: The leverage it's using. Here's my reasoning. The leverage it's using is insane.
19:26
Ejaaz: It is more extreme than DeepSeek, which is number one.
19:29
Ejaaz: So to me, as a trader myself, that seems to me it's a revenge trading method. It's mad.
19:35
Ejaaz: It's going ham. And it is like driving fast. It's drunk.
19:39
Ejaaz: It might wrap around a telephone pole. I don't know, but it's going for it.
19:42
Ejaaz: So it might result in a major win, but it might result in a major loss.
19:46
Ejaaz: So I'm going with Grok. Grok's going to get liquidated.
19:48
Josh: That's a good take. I think since you're going with Grok, I'll go with DeepSeek.
19:51
Josh: I just got to go for the crazy highly volatile positions yeah yeah
19:56
Josh: They're taking just outrageously long leveraged positions and like something's going to go wrong.
20:01
Josh: Like just two weeks ago, what, the crypto market dropped? Like we liquidated
20:05
Josh: 30 something billion dollars in like an hour.
20:07
Josh: These things happen and they move frequently. And granted, an AI is running
20:11
Josh: this. So I assume there's not a lot of latency between their decision and the trade.
20:15
Josh: But like this is a huge amount of leverage to be locked on in every single position.
20:20
Josh: Like if one of these coins gets wiped for like 10% even, you've lost 100% of
20:25
Josh: your money. so i'll probably go with deep seek um over the next let's say month
20:30
Josh: ejs who's your winner who do you think is the best longer term trader
20:34
Ejaaz: This might be a wild card of a guess um but i actually think it's it might be
20:41
Ejaaz: anthropic and as i'm saying this it is about to overtake man it's about to overtake
20:48
Ejaaz: grok 4 and here's my reasoning behind this okay when it started it off in this
20:52
Ejaaz: competition, it was doing terribly, Josh.
20:54
Ejaaz: It was underperforming, not as bad as GPTO Gemini, who is like further down
20:58
Ejaaz: in the chart, but it was doing bad. But then it learned.
21:02
Ejaaz: It did some analysis and it closed its positions that were already in a loss
21:06
Ejaaz: and thought, okay, I'll take that loss and let me try this new strategy.
21:09
Ejaaz: And it's been working heavily in its favor. So it's demonstrated two things
21:14
Ejaaz: to me. One, it's able to adapt and learn.
21:16
Ejaaz: Grok and DeepSeek haven't demonstrated that. They're just on a lucky streak
21:19
Ejaaz: right now. They haven't taken any else. Let's see how they perform under duress, right?
21:23
Ejaaz: And then number two, Claude is willing to take risk when it's learned from its
21:28
Ejaaz: lesson and go hard. So I think maybe over time it might end up.
21:32
Josh: What about you okay i'm going to go with quen who we did not mention much this
21:37
Josh: episode because quen is just kind of pegged to the middle and in fact i don't
21:41
Josh: think we really talk much about position open there so let me let me explain my why so
21:48
Ejaaz: Okay so quen.
21:49
Josh: Quen has been like kind of in the middle they haven't done much we don't really
21:53
Josh: know much about quen like we know about a lot of these other labs and how they
21:56
Josh: work uh quen's just kind of like this you know this middle of the road model.
22:01
Josh: But what I've noticed in observing the positions, like you were about to highlight,
22:04
Josh: Ejaz, is that it selects one position and it goes mega long on that one position.
22:09
Josh: So just before we started recording, I was checking through Quen and Quen had
22:13
Josh: a 20x long on Ether, all in.
22:16
Josh: One position, all the money on a major coin. It sold that, it went to Bitcoin,
22:21
Josh: now it is going 10x leverage on a very large position.
22:24
Josh: So what I'd like about Quen and what i've observed is where a
22:27
Josh: lot of other models are kind of spreading across a basket with lots
22:30
Josh: of leverage quen is very hyper fixated on just the
22:33
Josh: majors just ethereum just bitcoin back and forth i've been seeing a trade um
22:36
Josh: and that to me seems much more sustainable than going max long on something
22:40
Josh: like ripple that can just wipe you out in like a couple of minutes uh so i'm
22:45
Josh: gonna go with quen for the winner the sleeper pick um for our little model trading
22:49
Josh: i'm off to check in in a couple weeks or a month and see how these things do does
22:52
Ejaaz: Gpt or gemini dig itself out of this hole it is currently down three thousand dollars.
22:58
Josh: I want gemini too i'm not sure i want i want chad gpt too wait sorry what's your reasoning
23:04
Ejaaz: Behind that why because.
23:06
Josh: I i want i want chad gpt i
23:09
Josh: want the world to know that chad gpt is being a weak-minded psychophantic like
23:15
Josh: suck up and it's soft and i want chad gpt to be harder i want to be more direct
23:19
Josh: i want to tell me the truth as it is and i want that to be reflected in the
23:23
Josh: chart so that's just mostly me just virtue signaling that I want.
23:26
Josh: I want ChatGPT to be a little more serious, a little more, a little more harder on the edges, you know?
23:31
Josh: I think Gemini's got the dog in it. I don't think Chash GBT does.
23:35
Josh: So we'll just have to wait. We'll have to wait and see. But this is fun to watch.
23:39
Ejaaz: It is. And I want to hear what you guys, the listeners, have to say about this.
23:43
Ejaaz: Like, who's your dog in the race? Like, who do you think is going to win? Do you disagree with us?
23:47
Ejaaz: Right now, it seems like my bet might be paying off. Claude's coming through.
23:52
Ejaaz: But if you have any kind of difference of opinions, let us hear it.
23:54
Ejaaz: Maybe you know how these things are trained. Maybe you can train a better model. Let us know.
23:58
Ejaaz: Before we wrap this episode up, Josh, I just want to shout out,
24:02
Ejaaz: just kind of going back to nerdy mode for a second, how cool that this thing
24:07
Ejaaz: is built on an open source stack.
24:10
Ejaaz: We mentioned earlier that it's using an app called Hyperliquid.
24:13
Ejaaz: That's basically a blockchain. So anyone and everyone can get access to it and
24:18
Ejaaz: see the trades that these accounts are making, that these AI models are making.
24:22
Ejaaz: And if you want to copy trade them, that's not financial advice.
24:24
Ejaaz: I'm not saying that. You can, you know, you can see the data.
24:27
Ejaaz: If you don't believe this website, if you don't believe the tweets that you
24:29
Ejaaz: read, you can go and check that data for yourself and see what kinds of trades that are making.
24:33
Ejaaz: So super cool that the fees seem to be cheaper than using an average trading
24:36
Ejaaz: system and you can use it 24 7.
24:38
Ejaaz: This is not a shill. We are not sponsored by HyperLigrid. I just think it's
24:42
Ejaaz: really cool that they're using an open source stack finally without seeing anything
24:46
Ejaaz: bad being said about crypto.
24:48
Josh: Yeah, it's nice to see that there is they're testing these things in public
24:51
Josh: and in a way that's verifiable and in a way that's not gameable.
24:55
Josh: The biggest complaint that we always have is with benchmarks is that they're
24:58
Josh: gameable you can very much program your model to perform better at these
25:02
Josh: different benchmarks but this is the real world with real markets
25:05
Josh: and real people and real emotions and they're forced to
25:08
Josh: navigate a world that is not confined to a black box and instead has a lot of
25:13
Josh: depth and a lot of volatility so it's exciting to see this trend of new and
25:18
Josh: creative ways to benchmark this is particularly fun because you get to watch
25:20
Josh: the charts and hopefully make some money off of it and i look forward to listen
25:24
Josh: i look forward to using grok one day to hopefully advise my trading portfolio. That'd be pretty cool.
25:28
Josh: So we'll, we'll see how it goes. This was super interesting.
25:31
Josh: EJs, any final thoughts before we take off here?
25:33
Ejaaz: Nope i i hope we make a bunch of money in the future.
25:36
Josh: That is all i hope so too i hope so too and i hope
25:39
Josh: um everyone who's listening this got a little kick out of it
25:42
Josh: i'd love to hear like you just said who do you think is going to win but also
25:45
Josh: why you need to include why you can't just say who you have to say why because
25:49
Josh: i want another reasoning um and then i guess yeah who's going to be the best
25:52
Josh: model that's going to be fun we'll circle back in a couple weeks and we'll we'll
25:55
Josh: check in on this experiment but that is everything for today um thank you for
25:59
Josh: watching as always we appreciate all of the new uh ratings and reviews and comments
26:03
Josh: have been so overwhelmingly nice and positive it
26:05
Ejaaz: Has been kicking off.
26:06
Josh: Yeah seriously thank you so much for the support it really goes a long way we
26:09
Josh: are slowly climbing up the leaderboards and it's all thanks to you so if you
26:12
Josh: enjoyed please don't forget to like and subscribe share it with a friend who
26:16
Josh: you think might be interested share it with a friend who did really bad in investing
26:19
Josh: and maybe need some help maybe maybe just maybe they can use an ai and pretend
26:23
Josh: like it's them and take all the credit so So that's it.
26:26
Josh: That's another episode of the AI roll-up, round-up, whatever we want to call this thing.
26:30
Josh: But thank you for listening. I appreciate it. And we'll see you guys in the next one.