Sentiment extremes still work. Saylor went from cigar memes at the top to talking about selling MSTR on the front page of the Journal — then crypto bounced. Meta was cheap, hated, and legally radioactive when Muse shipped; a few days of actually using the agent was enough to buy it before the Street caught up. Now the real tell may be quieter: parents and boomers treating personal agents like a full-time assistant, and token usage going parabolic. That’s the setup for another AI leg through the daily doomer noise — and, later, the pruning that comes when the market finally prices the next wave
[00:00:01] It's been a long while, but I've been hard at it over here. Loving this time with so many things coming out and being curious and fleet-footed and willing to try stuff is the best time to be alive for that. Now, they're not putting it back in the bottle, so you may as well get on with it is the way I look at it, but it's easy for me because I like it and I have time all day to play with it, so I'm lucky that way.
[00:00:24] But I think things are shifting a little bit. I'm noticing, so I have a few things to share today. I'm interrupting myself as I think about it here. I want to talk about crypto briefly and the bottom we put in, and then I want to talk about the meta trade, and then I want to talk about agents and where it might be going and overall market and all the doomers and stuff like that.
[00:00:46] So start with crypto quickly. I love sentiment stuff. At the top, you have Michael Saylor basically sitting there. We have rap music playing to hit memes of him with cigars in his mouth, and everyone's an idiot if they don't own Bitcoin. And then over 100,000, they suck in the crowd, and then eventually it gets wiped out and goes lower. And I thought that was a great setup up there to stay out of trouble or sell or whatever, get short.
[00:01:12] And then on the bottom, front page of the Wall Street Journal, MSTR goes below 100, and now he's saying after never wanting to do anything but buy, he actually wants to sell MSTR. It's comical, right? So then they crush it, and it sits there for a couple weeks, and then now we're back to rallying again. It's just you couldn't – it's like everyone loves it. It can do no wrong. And at the bottom, he flips his tune. There's no more memes, no more interviews on CNBC, and then we go the other way. So I thought that was great, and I like that trade a lot down there.
[00:01:41] And then we also had tokenization really picking up with Robinhood and Ethereum, so it all kind of lined up. So that was great. And these things, they keep repeating over time, so if you miss them, you just keep getting this thinking process in your head. So the second one, similar, is the MetaTrade. So if you're trying new things, like when Muse came out, I got on it right away and tried it, and I had been building my own OpenClaw agent for like the last nine months or a year, so I had a lot of this stuff already working on my own processes on my server,
[00:02:07] but mine wasn't a consumer product architected to go and like return clothing items and buy things. So right away at Muse, I started giving them my email and my stuff that I don't really care about. But I'm just not super concerned about that. I know some people are, but I just, I want to know how much, I want to push it hard and see how good it can be, first and foremost, before I worry about that, because I'm in the market. So that's kind of my priority, is to test things. So I was very impressed by it, and then I've noticed it's screwed up some stuff since,
[00:02:34] but I'm assuming it's only going to get better. So Meta is on the front page of the journal, multiple days with legal problems from, you know, destroying society with social media and all that other stuff, and children addiction. So as they're settling their claims, the stock's super beat up, but it's actually quite cheap compared to, if you look at the ad business and how stable it is, and they're making ad gains versus Google, because people aren't manually searching as much. So it's a good setup down there. And then Muse comes out, and the stock's trading like,
[00:03:02] I don't know if it's like 630, 620, 640 maybe. So right away, I bought a bunch of Meta, because I played with Muse, and I was like, whoa, this is like, I know from having my own custom agent that people are going to love this, but now they've made it to be an easy consumer agent. And I actually use Muse pretty heavily as a personal agent now for buying stuff, for reminders, for tracking like prices of things, for trip planning. And then I have my open claw for all my business work processes with, you know,
[00:03:30] code on servers and all this kind of stuff. So, but I love like blow out the positioning. Everyone hates it, good fundamentals. And all of a sudden out of the hole, it comes with a new product. You're four or five days in front of Wall Street if you're playing with it. And then once it hits and becomes disseminated, it's at 780. And then now it's at like 730, and that's your trade. And I think there's probably a hold for a longer trade too. But everyone else will eventually come up with an agent, but they're going to have so many users. I mean, it's growing faster than ChatGPT.
[00:04:00] So again, it's like you just keep seeing these things, and we're going to keep seeing them over the years. So it's just study those things, and you get better and better at catching them. I used to see them and miss them and get in and get out and trade too big. And I feel like I'm getting better at getting them now. But staying curious is a huge part of it and just recognizing these shifts. And then the last thing I want to talk about is agents and where it's going. So for me, this has been just amazing.
[00:04:26] I've been doing better the last couple years, cleaner with my trades. I've had more time to do more of the same, actually, is optimizing, optimizing, and then not taking time to go. It's not really my style to go sit on the beach. Sure, a couple trips are fun, but I just like doing this stuff. I've always loved doing it, and I love studying it and getting better. So more time to do nothing usually means more time to play with this. But it's true. I don't have to work as much.
[00:04:53] I mean, I really don't have to do more than like 90 minutes or two hours or less a day if I don't want to be effective. And I used to not have that because I have my agent dropping all the stuff in my lap now that I've – and that was the hardest part is getting – how do I get this in to the system, get my experience into code? It's a whole job in itself. So I'm playing with this JEV, J-E-V, which is kind of a yes or no probability model, which is really cool. I think it's going to unlock judgment.
[00:05:21] It's because everyone's like, well, yeah, I got all this stuff, but how do I know what to focus on? Or what's the – like can it judge itself? And this thing is basically trained to do that. So if you have a lot of your own processes and data in there, you can start to ask key questions, and that's the whole crux of it, right? Can you ask good questions? Better question, better result about like do I deep dive into this, yes or no? Does this earnings report change the story, yes or no? Does this product expand the market and is it worth trading, yes or no? Or whatever it is.
[00:05:49] So I'm trying to figure out how to use that, integrate that into my process. I'm just starting that. So that's an idea of something new I'm doing. In terms of agents, like you have to assume now – like let me give you an example. My mom uses GPT+, which is pretty good for someone who's older, and my dad uses it too, and he's a little slower at it. But then they putter around at night and they fuck around on it, ask questions. I got my mom on Muse right away, and she just cannot believe it. She's all over it. Her token usage – I wish I had a chart. I'll try to find a chart and post one on the Substack or something.
[00:06:18] I have to go extract her data. But her token usage has got to be parabolic from what it was on GPT+, which I'm sure she still uses that too. But she's on Muse constantly, like just for everything. She's used it as her total personal assistant now. And I can only imagine that slowly, especially the baby boomer generation, they're going to figure this out and get on these things, and they're going to use so many more tokens.
[00:06:43] So a great thesis would be that the market has another big leg up in the AI stocks into all this fear and doomer bullshit that you see every day about us all dying. And then what's going to happen is – I mean, I'm not trying to predict the market. I'm just – that's a foolhardy move. But what I'm trying to say is I think we could have another big leg up, and then after that, then we could have trouble.
[00:07:07] When they price in the massive – when they validate all this expansion that we're having and they crank these names up again, then it's like, okay, now we need some destruction. We need some pruning to wipe out all the slop and get everything based out and figure out, like, okay, the physical AI, the robots are coming. Like, let's get ready for the next trade. And there's going to be some damage, destruction, and maybe a bear market or something. I don't know.
[00:07:32] As the technology keeps moving up, maybe it's just so, so fast that it just doesn't matter and it keeps going and going. I don't know, but I could see the market getting ahead of itself because if we burst through this doomerism on the agent stuff, then it's probably going to be a pause for a while until we get ready for the next wave. So that's kind of what I'm bracing for right now. So it's a tough time right now in some ways, but it's also, I think – I think it's a good shot we have another leg. But we'll see.
[00:08:00] I mean, protect myself if not, right? Like strong opinions, light loosely held or whatever they say. But, yeah, that's kind of all I have to say today. But I'll try to come on more. I do like talking, but I've just been doing this and have a substat going. But I've got to get on that too. It's just – I'm so bad at producing content. But I wish I could just have something over my shoulder and just dump everything I do all day into content because I'm looking at so much cool stuff, but then I just have to sit and talk about it.
[00:08:30] Anyways, hopefully people are doing similar stuff and enjoying it. And you can always email me or figure out how to connect however you want. And I love hearing about it from everybody else. Take care.