Tomorrow - DeepSeek Has Joined the Chat

Episode Date: February 3, 2025

This week, Josh and Rani discuss tech earnings across the board, as well as DeepSeek's big upset to OpenAI's Research and Development heads. Also, we will be taking a hiatus for the time being, see you on the other side, Tomorrowers! Learn more about your ad choices. Visit megaphone.fm/adchoices

Hey and welcome to tomorrow. I'm your host, Joshua Topolsky. And I'm your other host, Ronnie Moller. And boy oh boy, do we have a week of news to talk about in technology and beyond, but only slightly beyond. Uh, we were just running, I were actually just talking as we started this and we're like, well, we got to talk about deep seek the new Chinese AI that you're like, that happened like a month ago. I'm like, that was like last week. That's kind of old news now, isn't it?

And then we did a little bit of sleuthing, including using this extremely handy Sherwood.News post, which is a timeline of the DeepSeek market freak. And that was like Monday is like when the DeepSeek stuff started happening, or Sunday night maybe. Yeah. I mean, the stock market imploded on Monday, but like the model, like the testing of the model was last week, but it only started really freaking everyone out over the weekend.

And then everyone's trillions of dollars out of the stock market. Rodney, could you, for the listener, explain the DeepSeek? Just explain DeepSeek a little bit for someone who maybe, I don't know who's listening to this podcast who doesn't know, but I think it'd be worth, with a little bit of the backstory of what DeepSeek is. Okay. who doesn't know, but I think it'd be worth sort of with a little bit of the backstory of what DeepSeek is. Okay, so DeepSeek is a Chinese AI startup and they were able to make a chatbot, an AI chatbot. They were able to train it very very cheaply and China, there's some bans on selling chips, these GPUs to China. These advanced Nvidia chips. So they were they made

this model this new chatbot using not really advanced chips not that much money to train it and it turns out to be just as smart as like chat GPT. Right. So like they spent like six million dollars or something. Right on the training of it. Obviously that's not all they spend and there's some questions as to whether like is that all they really spent. But chat GPT is like talking in the billions, billions of dollars. Right. And like each training time is like a hundred billion. And yeah, it's a lot more. Yeah. And the sort of the bottom fell out of the stock market because it sort of undermined everything we thought

we knew in technology. For a long time, all the biggest tech companies have been spending billions and billions and billions of dollars on CapEx to build out these AI data centers and do everything AI. And the idea was, if we don't do this, we're going to be left behind. Spend, spend, spend.

We've got to spend in order to get ahead. The R&D projects were just growing and growing, right? And a lot of the valuations, like the kind of like growing valuations of both these private companies, but then also the market caps of like the NVIDIAs and the Microsofts was all kind of predicated on this AI boom that we were in the middle of that was very costly, but would yield huge value on the tail end of it. But it was very expensive to get going, right? So it was like the market was like, wow, the R&D costs are going up.

It must mean, that must be because there's so much incredible innovation we're unlocking through these like AI. Exactly. Big GPUs and these, getting compute time and power and all that. And like the funny thing is like, spending all this money was supposed to like secure American AI dominance, right?

Right. And the whole idea of like banning China from having these, these chips was like, okay, we're going to, you know, we don't want, we want to get ahead of China. And then it doing so actually made them more creative, made them more resourceful, made them do more with less. And so it totally undermined the whole idea. I mean, this is honestly, it's so funny because it is so often the story of innovation is this exactly,

which is not... I mean, it's actually like Apple and IBM. I mean, if you look at the story of how that company came to carve out its own space and eventually in many ways dominate at least Mindshare and certainly the trend lines of technology. Even Microsoft, to some degree, started as these scrappy, hard-won... Originally, they were not the dominant richest, you know, people in the game. Right. And it's like they had to be scrappy and creative and use like all kinds of weird ingenious tricks and shortcuts to get where they were going. And like I think, you know, there is

something to be said for, for the now look, there's a bit of a myth in Silicon Valley of like, Oh, the, you know, Jeff Bezos in his garage. It's like, well, Jeff Bezos with a bunch of like, oh, Jeff Bezos in his garage. It's like, well, Jeff Bezos with a bunch of money. Right. There's always the bunch of money or the thing. Yeah. But yeah. But there is a little bit of, I think,

in terms of general ingenuity historically, there is a little bit of, there's something to be said for having the cards, the deck stacked against you, I guess is the expression, right? Or having the cards not be in your favor Yeah, and it looks like it does look like to some degree Like open AI and all these other folks look really bloated by comparison. They look really not like they're not the most ingenious creators of this type of technology and and if it can happen so rapidly that they get lapped in this way using much

lower powered GPUs spending much less money, what does that mean for the entire industry? I guess that's the freak out, right? It's like we've completely misjudged. And I think that's probably true. Right. There's always so much money you could just pour into, endless billions you could pour into. I think Microsoft had it on, they said they were going to spend like $80 billion on CapEx this year, like last week or sometime this month, Metta was saying that they were going to spend $65 billion.

The numbers are just so huge. Yeah, I see over and over again, we have this, in the very last decade especially, this trend of these huge valuation businesses with these, I understand that to some degree all innovation and all of these new emerging tech businesses require a leap of faith. And yet I think that we're further and further away from the proof points of the leap that you need to take the leap of faith. I think Theranos is a great example. Theranos is a great example of, well, the proof points, there wasn't really independent study done on that stuff. It was really just like, they believe what the founders said,

and the founders were like, it's going to be really important. They were like, man, I really believe in this mission, and it sounds like they've got the science, so let's just put like a billion dollars into that. It's like quibi, you know? They're like, well, these guys... Yeah, you know when they're talking it up so much before it's there. But I would argue that there was a bit of a proof point. Like, ChatGPT came out and it was like, even if it was a bit of a parlor trick, it was very impressive. Yeah, but it wasn't... They were just looking for a use case.

But the best one I'm saying is, what would have happened if we had said to the open AIs of the world, we need to see something a little bit more concrete here in terms of your speculation that it's going to be X, Y, and Z, and that you're the expert, and we need all this compute power. You might have, for instance, said, can, can you do this in a more energy efficient way? Because it seems like there might be a more energy efficient way to do it. And now meanwhile, fast forward, what, one or two years, and someone else is doing it in some ways lapping their innovation. I just, look, the point is, it's not to say that OpenAI hasn't done things that are impressive.

It's possible that the lack of scrutiny and the lack of scrappiness at all that was sort of demanded of a business like OpenAI has allowed them to be more complacent and more bloated. And I think there is something to be said for trying to do. I mean, in this case, especially when you're dealing with huge energy implications, is there a way that you don't need a billion dollars of investment to actually make a product that people want? And I do think the product people want,

and it's like use cases. I'm not saying that there are not. I was actually reading, Ashley Vance has a new publication that he's doing, Ashley Vance, who used to be at Bloomberg. And I think it's called Core Memory is the name. And he did a story about this guy who built a fusion reactor, a miniature fusion reactor, a fuser, I believe they're called, in his bedroom using like getting like one of the AIs or multiple AIs to kind of help him figure out how to do it and it talks a little bit about

Sorry, I swear to God this is connected to what we're talking about. I know you're like what is this? But he talks a little bit about how this the who did it is this kind of AI native, and Ashley is talking about how he's seeing a different type of computing. So I do think that there are, not everybody who is particularly, not everybody has yet to see the way that AI, or what we're calling AI in these large language models, can be used in all sorts of different applications. So there are real use cases. But I think the question that it raises for all

of these companies, from OpenAI to NVIDIA and all along, is do you know as much as you say you know about this technology? That's kind of the big question. I think one of the reasons the market freaked out is because it's like, man, if this Chinese company can do this with much less money and much less tech, and on the other hand, we're being told by OpenAI and Nvidia that you need the bigger and bigger compute power and more GPUs and more energy. And they're not yielding results that are as impressive.

It's like, well, does anybody even know? Do these people even know what they're talking about? One funny knock-on effect of this was like, for OpenAI at least, first off, Sam Altman, the CEO, he tweeted, he was like, you know, it's really impressive what DeepSeek did, especially for the amount of money. I'm really excited that we now have a real competitor, blah, blah, blah. And then a day later, DeepSeek's accusing them of training their models on chat GPT, which is just so rich because like, you know, You mean they were, sorry, they were accusing DeepSeek of training their models.

Yeah, on chat GPT, like, of training its models on chat. Yes, they, OpenAI was accusing DeepSeek then, like after he was like, oh, good job. He's like, actually they might have trained their models on chat GPT, but like, what's rich is that chat GPT or OpenAI has like been training on like everyone else's intellectual property for years and years and years and has all those lawsuits. It's like turtles all the way down or whatever it is. It's like, well, it is at the end of the day. One more thing.

And then finally, today, because of this, because of DeepSeek, OpenAI actually is releasing a new, cheaper model. So it's actually causing more innovation now that they have competition. Causing innovation. That's one way to put it. Or opening eyes like, okay, actually, we can't actually do it cheaper and faster. Sorry. If you stop writing, just open big blank checks. We just wanted to make it seem really complicated. There's an episode of Seinfeld where George

Costanza talks about how he, at work, he just piles up all of these papers on his desk and looks stressed out. Anytime anybody walks by, he looks really stressed out and people leave him alone and don't ask him to do anything because they think he's embroiled in some complicated project. And really important, yeah, I like that. That's like that, there's a meme, it's that,

and just as open AI, open AI R&D. That's a free meme for somebody, for a creator, for an online creator who'd like to capitalize on this moment. I'm going to pop that into one of the image generators. Please do. Please, I beg of you. Anyway, I think it's interesting. I mean, one thing that has become,

I have definitely, over the course of the last several months, gone from being a, I'm convinced that there are real and valuable and exciting applications for these models that are being created. I don't doubt for a second that there is a future where, and I'm really honestly loathe to say, I think like Satya Nadella said,

the CEO of Microsoft said a few months ago, he's like artificial intelligence is the wrong word for it. And I think about that a lot because I do think it's like we're using the wrong word and it gives us this kind of set of beliefs about the product that are kind of not aligned with … Harjes Kinnisar, Ph.D., Ph.D. It's like full self-driving with Tesla and you're like, you can say like drive assist. You're like, oh, that's cool.

Josh Birkhoff, Ph.D. Right, right. Drive assist and full self-driving, dramatically different nomenclature around the feature. You know, like same feature. And different expectations. Yeah. Different expectations, right. And I think what he called it like other intelligence or assistant intelligence or something like

that. I do think like there is a huge opportunity for this stuff to be deployed. I think what is sad and annoying, and this is going to segue into something I know you want to talk about. What is sad and annoying is that so much of the early versions of this have not really been focused around ways it's going to legitimately do things that are useful for human beings, but instead the things that I guess appear to tech companies as the most exciting tricks that it can pull off. One of them is like natural language chat or

whatever. One would be image generation. In Apple's case they are advertising, they literally have billboards for genmoji, which is their AI generated emoji. What's the expression when you're a hammer like everything's a nail? Is that the expression? That sounds right. But Apple, meanwhile, to pivot a little bit away from the deep seek story, which is fascinating and undoubtedly will, how about Apple's earnings? There's the segue. But the thing is, the stock was trading up. Let me look where it is. Where is it now? Not that that means anything anymore. Okay, now down 1%.

See? See, you can't rely on your memory. So like, from the earnings yesterday, like I thought it was like pretty mixed earnings, right? Like overall they're like record earnings, blah, blah, blah, but like- They always say that. They always say that because it's always, the number goes up and to the right. I have a piece about this today and like iPhone sales went down and they're like, but it's record install base and record upgrades. It's always going to be record install base. It's always going to be record upgrades because it's the law of numbers. If you have this many people with an iPhone and you don't completely bomb, every

time they need an upgrade, they're not, you know with Apple, it's not like Apple doesn't have like tragic quarters, you know? Not really. They're still like pulling money hand over fist. They're still Apple, but you know, it's definitely a mixed, I would say it's mixed. Is this related?

Yeah, except the stock was up like 4%. So I understand overall earnings look good, but the iPhone dip is certainly gotta be a concern for them. And then, you know, they've predicated the kind of latest push for iPhones. It's like, and frankly across a lot of their products is Apple intelligence. And man, Apple, I don't want to go because we've heard us, the people who listen to this

podcast have definitely heard Ronnie and I talk about our disdain for Apple intelligence and our confusion about it. But it didn't move units for them. Didn't move units, right? It was not a hit product. Well, it's funny because they kind of, he tried to make it sound like it moved units, right? It was not a hit product. Well, it's funny because they kind of, he tried to make it sound like it moved units. Like what he said, he's like, if you look at the iPhone 16 compared to the 15, like the iPhone 16 outperformed the 15. But like if you look last year, they said the 15 outperformed the 14. Outperformed in terms of what?

Like the number they were able to sell in that time. But I'm thinking, again, that's the law of numbers thing, as far as new models. He wasn't especially clear. But then he's also like, and I think you can conclude from this, that there are compelling reasons to upgrade. But the new model always sells more than the last model. They said the same thing last year.

And he was like, and the markets where we launched Apple Intelligence, they outperformed the markets where they did not. But like, no shit. Like, why would you buy a phone for Apple Intelligence if it didn't have Apple Intelligence? He's saying like in terms of like a ratio in markets where Apple Intelligence was available,

they sold more new phones than in markets where Apple Intelligence wasn't available. Where the flagship product that you're selling works. You're right. You're going to sell fewer things if they don't have the new thing that you're touting to everybody else. Like for sure people are going to hesitate and wait. I guess the point here is more to the point. It's that Apple kind of hinged a lot of its latest efforts on AI. This is getting back to what I was saying about AI and the result was mixed

at best. And you do not see, well, I mean, look, they also had, I don't know, Vision Pro has been like a whiff. I mean, we've talked about this a lot, but the Apple intelligence thing is like, man, Apple doesn't have a whiff of anything that feels like truly interesting. I mean, you know, if you look out amongst what's going on, if you look out around the landscape and technology about what's going on right now, you know, you could say what you will about AI, you can, you know, be a naysayer and be like, it doesn't matter. But we talked about this last week, the underpinning technology that Nvidia has created

and that underlying technology they've created, which is serving an entire generation of new PC gaming and other types of high math, intensive math operations on computers, Nvidia has cornered the market in some ways. of high math, sort of intensive math operations on computers. Nvidia has cornered the market in some ways. And no, they're not cornered, but they are occupying a large space on the block.

And somebody was like, oh, they're bringing Cyberpunk to Macs or something. And I'm like, OK, cool. But show me the graphics performance on a Mac that is equal to what a PC with a good graphics card can do. And I don't think it, you know, I think it's, I think it's a real challenge for Apple. And it just says like, they're missing some component of, of this moment, right, which is again, I said this last week, it's not

about the hardware. It's in the sense of the stuff up top, it's not about the operating system, which I still think they've lagged an innovation on. They've lagged an innovation on their OS, I think, and then their general sort of the way they tie all their devices together. Though nobody's done that much better, to be honest. I mean, it's like Windows is crushing it across platforms. And Apple knows that it's fucked this up, right? There was a reporting in Bloomberg last week that

was like they had listed this veteran executive at Apple to fix AI in Siri. And there's a new version of Siri coming out in April and another one coming out in 2026. They know it's broken. As much as they say everyone loves this and is using it, they're not. I mean, I'm sure people are calling up Siri for things.

And I'm sure there are plenty of- I'm yelling at it because it's wrong. Right. No, I mean, the funny thing about all of this is the one place where we were promised for a long time, and I think we said this last week, the one place that we were consistently promised there would be a great experience, and we even started to seemingly approach it with things like Siri and Alexa and Google Assistant or whatever, is that it would get smarter and better and more capable and more clear on these assistants would be able to handle these

kinds of mundane intermediary tasks that we do with our phones, like conversion or whatever. And playing music or whatever the stupid things are. It's like making appointments. And Apple demonstrated quite a bit when they had the AI event and talked about the new phones. They demonstrated quite a bit these amazing things that it would do. I think we talked about it.

And we're like, wow. If they can pull this off, it would be really cool. Yeah, if this works, it would be great. They did not. Not yet. It doesn't do any of that stuff. And it's like, why even bother? I mean, this is actually, for Apple,

it's very uncharacteristic to show features and then ship something. It's very uncharacteristic to show. I think that's a testament to like, just like ChatGBT coming out and everyone being like rushing, even Apple rush. Right, but that's, but Apple's whole thing was, I mean, Apple sat out MP3 players for generations.

Until they could do a better one. Until they had a thing, they were like, this is gonna be better, more cohesive, like the iTunes situation, all that. They had a plan that was much more cohesive and much more sophisticated, and then they launched it, and then they obsolete it all. It's the same with the smartphone. Apple could have stepped into the smartphone market in a hundred places, but they waited

until they had something that seemed like truly a unique product, and then they dropped that in and they completely changed the market. I think, and they did something similar, to be honest, with the personal computer, but it took, I think the change there took a lot longer and was a lot more subtle, though at the end of the day, a lot of their ideas were right. They were right, and it changed the industry. I just think, go ahead. I was just thinking that we actually have like sort of proof that Apple really screwed up with AI at least so far in that like when there's this whole giant like tech route where

you know Nvidia lost like half a trillion dollars in market cap, all these tech companies were just down down down this week. Apple was actually trading up. I thought that was sort of like a testament to the fact that like maybe they're not an AI company after after all. They didn't spend as much as other companies. They've outsourced it and they did a bad job. Their stock's not going to go down when the other AI stocks crash. I think that's actually a really good observation, which is they didn't put all their eggs in the AI basket. Smart. I'm like, that's smart. That's good. That's classic Apple. It's also a smart move for a tech company to not be like, we're going all in on AI, because they didn't do that. Right. They don't have the capex levels of the other ones.

However, but they kind of half-stepped into it. And their half-step sucked. And so, you get a lot of the negative effects of that, which is like, I think there's a general dissatisfaction with their innovation, general dissatisfaction with the product. But it's true, you get less of a, right, if you're not all in on AI, if you haven't like doubled down, definitely. Yeah, their exposure is going to be less. When there's turbulence, it's going to be a very different thing. But yeah, it's interesting. I mean, who else did earnings? Microsoft and earnings, they were down a little bit, right? Yeah. So like, so Microsoft's cloud business brought in like less money than expected. Cloud business brought in less money than expected. I think there was some adjacency to DeepSeek in general that didn't make people feel too

good like Microsoft spending all this money on CapEx and then DeepSeek just ate its lunch. There was also Meta and test learnings. Yeah. I mean, how did Meta do? They're crushing it. They're going to... I think they said they wanted to get like a billion AI users and then start monetizing them.

I don't know where they are in the journey there. I don't know either. But their net income is way up. Quarterly revenue is up. You know? It's way up. Yeah, they're... Quarterly revenue's up, you know? Everything's up and to the right. It's all good, it's all good. I mean, I think, but Metta's taken,

I mean, Metta has a very different business. I mean, at the end of the day, their business is not actually judged on their AI business. It's an ad business, yeah. It's an ad business, and it's been good for them this last quarter, I think it's been really good for them. Microsoft seems to have much more of a lasso to AI. And I think they've made a big play on that.

Yeah, big investment in OpenAI. Yeah. I mean, look, at the end of the day, there is a level of hype to all of this. There is a real level of hype that is just going to die down at some point. You've got to put a product out there that, if you're Apple or Microsoft or Meta, Meta less so, I think their business is a lot more straightforward in some ways. They can experiment with virtual reality and Ray-Ban glasses.

They've got this giant ad network. Metaverse and AI. It doesn't even matter if you're on their sites. Yeah, like at the end of the day, because what really matters is if like they're keeping people paying attention to like the ads on Facebook and Instagram and they've done a pretty good job of that. Or the rest of the internet, wherever their ad network is.

And not for nothing, but I'm sure the TikTok ban scare, it's, you know, miniature one-day ban and all the rest of the kind of panic around it probably helped their bottom line in this quarter because a lot of advertisers probably like, we better move our business over to somewhere stable, right? So I'm sure they saw some pick up there just on the fears of the TikTok ban. I just saw a headline with Meta and TikTok and it was like they missed Zuckerberg in some leaked meeting because everyone leaks all of his all hands.

It was like we missed out on TikTok because we didn't really consider it like social. What? That doesn't make any sense. Right. Mark Zuckerberg said that Meta missed TikTok's rise because it didn't seem quote social enough. I don't know. I mean, you mean they didn't see the social qualities of TikTok?

Or Meta's not a social company. Unclear what that means. Let's see what the quote is. When I look back on TikTok, I think part of the reason why we were slow to it is because we didn't think TikTok was social. We looked at it and we thought, oh, this is a little more like YouTube.

So they didn't get it. But YouTube is a social network. I mean, what's insane about that, I just want to be really crystal clear. What's insane about that statement is it suggests that Mark Zuckerberg does not consider YouTube to be a social network, which it 1000% is, and in fact lives or dies on the social aspect of its network. Because it is like, I mean, literally it's like creators who build followings, who get these like human beings to subscribe and then interact with those people and serve them a certain type of content.

And like it's all about like getting the mass of people into your personal little private YouTube network or whatever. I consider YouTube to be an absolute video social network and I think it's odd. They've built out features more and more that make it act like video social network. And I think it's odd. And they've built out features more and more that make it act like a social network, including YouTube Shorts, which are like TikTok. And also, they have a very robust commenting section and experience.

They have a very robust user profile sort of thing going on. It's weird. And I'm reading a Business Insider piece right now. And it's like, Zuckerberg explained at the meeting that Metta's traditional view of social interaction centered around friends posting content and commenting caused the company to initially misread TikTok's appeal. The company failed to recognize how users were sharing TikTok through private messages, which has become a crucial form of social interaction across Metta's platforms.

And then quote, because we were too dismissive upfront. It wasn't just about people commenting in the feed How myopic they're like if you're not doing exactly like we do it. It's not social. What's weird, but what's weird is tick-tock is like Follow it's like look at a feed and then follow. Okay, it's literally not just sharing it in texts Yeah, no, I mean, and I'm sure they have a good business of people sharing, you know Yeah, no, I mean and I'm sure they have a good business of people sharing, you know Tick-tock to each other but that's beside the point. I mean literally it is all about like oh this person has X amount of followers Because tick-tock is a social network Which is predicated on I mean, yes, there's the for you page, which is like this pro Just this algorithmic feed but like it's it's certainly people look at that

I mean truth is it has more in common with TV in some in some ways like that But like like normal TV Maybe they didn't take it seriously because he knew he would just be able to copy it later like he did He hasn't been exactly able to copy it. That's the funny thing is that there is a very distinct feeling from using reels Versus tick-tock. Yeah, the vibes are off. The vibes are off. And in some ways, like with threads, which I have basically come to totally ignore at this point. Yeah, I stopped it.

The mashing together your people from one social experience into a totally different one actually doesn't work that well. Because it turns out what I want to see in short form video is really different than what I want to see from my friends or people I'm interested in. Yeah, I don't want to tweet with my friends from high school. Yeah, right? But I think Instagram, when I think about what I want to look at on Instagram, historically has been either people I know, who I because like we're friends or family or whatever, or like interesting, like visual, either like artists or creators or like curators. There's

like a lot of really great accounts on, on Instagram that are like what, what Tumblr used to be where it was like people who would curate interesting images and stuff like that. And that translation from that to like what Tik TikTok does, totally different space, totally different feel, totally different expectation of what you get in the feed there. And yeah, anyway, and then threads is like that. Threads, I go on threads and I'm like, oh, geez. The people who follow me here and who I'm following

and the people who respond when I post on here are really not the kind of audience that I would expect to have would like. Yeah it's like you're at the wrong party. Yeah it's like Twitter back in the day which is why Blue Sky actually is the closest thing I've experienced to Twitter back in the day. It feels like you're having a conversation with the people who are want to communicate in that like form and format it's not meant for other it's not like you took them and displaced them for somewhere.

Why are there no pictures here? Right, right. And it's like people going viral for posting bad takes that are like, it's just a whole different, I mean certainly people go viral for bad takes everywhere. But anyhow, I don't know how we got onto this topic, but oh, Metta's earnings, Metta's business. So I mean they're doing great. They pay topic, but oh, Metta's earnings, right? Metta's business. Yeah. So, I mean, they're doing great.

They pay Trump some money so he'd leave them alone or some like bribe money or I don't know what you can, I don't know what they're doing. Everybody's like paying Trump now. Paying him in different ways. Yeah. Just giving him money. I think it's a good gig if you can become president and get people to give you money directly. I think it's like, and it's a good gig if you know, 25 million is like a rounding error for you. And you're like, oh, not a big deal. I bought my way in. It's very good. It's very good. Anyhow. But, um, but Zuckerberg is crushing it. And there's a, I guess there's a leaked, I didn't see this, but there's a leaked, you mentioned it, some transcript from a meeting they had.

Yeah. Where he's like, um, well, first off, he, in this all hands or whatever, he says, you know, they've sold a million Ray-Bans, that's the first time that number was out. But then he was also being real hard on the employees, saying it's gonna be a real hard year. And then he was saying that he doesn't really talk about stuff because everything gets leaked and it sucks. And I was like, oh.

And then there was a New York Times reporter who was like, after he was like, stop leaking, she got a bunch more leaks. So, worked out. So, it worked out. People really respect and love Mark Zuckerberg. I'll be honest with you, and I'm not saying like, I do think it says something about the type of environment you're building when people feel compelled to leak. I'm not saying that's always true. I'm not saying there's a uniformity to that, that we should be like, oh well.

No, but on the scale of Facebook. It does feel like there's certain companies that, or certain things that seem to get leaked more. There's certain people who are getting leaked, you know, and the leaks are like not a super flattering view. I also think like he's been like weirdly antagonistic to the people that work for him, which is one way to go. I mean, I understand the hard-ass boss model, and I do think there's a fear in Silicon Valley that the workers have had too much power, which is funny because I don't think there's a single union in all of Silicon Valley. But like...

Yeah, there are too many women. Not the masculine energy. The masculine energy thing. I mean, just if I were like a not a man who worked at Facebook, you know, if I were, if I were, I would just be like, huh, this is like a weird, aggressive way to talk to people who are probably really good employees. Right. Like, I'm sure there are, I mean, it goes without saying there are plenty of employees

at Metta who employs, What is there? What is there? What is the Metta employee count? Do we know? At a glance, let me take a quick look here Metta employee count 76,000 a day. Let me find out actual things I don't know it says as of December 31st, 2023. This is all it had 67,000 That's 23. I don't know how as of December 31st, 2023, this is old, had 67,000. That's 23, I don't know how recent that is. There's investor relations.

Yeah, I think that's probably close to right. I don't know, these numbers are, whatever. It's like tens of. So head count as of December 31st, 2024 was 74,000. Okay, 74,000 people work for Metta. A lot of people. All over the world. And very smart people.

And it's really can you imagine? I mean, it's hard to imagine. I mean, we work at a company that has fewer employees, and I'm trying to think of what it would feel like to hear like our CEO say something like that. It wouldn't feel good, I think for a lot of employees I think anybody who works at any business when you're like you're somebody singling out Some other in some other part of the group of people who work there

It's just a very weird thing to do like it's very unusual and it's like maybe it's for an audience of one But it's still like he's got to like go back and I get you. I get day to day basis and hope that you're Zuckerberg. Yeah, right. Yeah. I mean, maybe if you're Zuckerberg, like you can be kind of sort of act more like a like a dictator or whatever, more like a hard ass and given the the the success of Metta. However, it's no surprise that there are leaks.

Yeah, no surprise there are leaks. It's no surprise there are leaks because I think once you start to position yourself as adversarial towards the people that you work for, there's always a split between bosses and employees. It's a tension. As the boss, you're often going to seem like the bad guy. That's just how capitalism works, unfortunately. And but but there are there are. You could be that you know, you can have like a you can have bad days with your employees and still like

have them not like leak the conversations or whatever. And it's like, I think I mean, you know, I think. But I'm not I don't know, because I don't know what it's like to be. I think he's doing the you know he's doing the Elon Musk playbook who Tesla also had earnings. Oh yeah how are they doing? They also reported what I thought were poor earnings and then you know the stock went up like it's the... You and Goldman were surprised apparently. Yeah. One of the big guys was like... They're divorced from fundamentals. Well're divorced from fundamentals. Yeah. Divorced from fundamentals. Like, okay, have you seen the numbers at all?

Yesterday. Oh my God. JP Morgan can't understand why in the world Tesla would be up after earnings. Yeah. Cause like the thing that they sell kind of like Apple, you know, the main thing they sell is cars, cars that people drive. And they're selling fewer of them now than they used to. And what was it? Their net income went down like 53%. Like a huge drop, their profit dropped 50%.

And then the stocks were trading up like 5% higher. Like it's out of this world. And like I listened to the earnings call and it know, if you, it's like a certain kind of, it feels like just like being at a party or no, it feels like being with like alone in a room with a crazy dude who's just smoking weed and just like talking to you. That's like the earnings calls are like with-

Sounds fun actually. I like the sound. I think more earnings calls should be like that. Yeah, well that's what it's like listening to Elon Musk. And when you know, he's just like. He's got all these... And this is what people are investing in, the idea that he's got these big plans and everything's all over the place. But he's definitely directing people all over, everywhere except the cars. It's all about the robo-taxis or the solar roofs or robotics, like the optimist or whatever, like he just like can't like autonomous ride

handling. He can't talk about the cars. Like he spent like two seconds being like, oh yeah, cheaper car might be coming out the first half of the year, blah, blah, blah. But like no details about it. Nothing there. But then like goes into like the nitty gritty of like energy storage or like solar roofs. Right. Right. Right. They are, I mean, it's getting cornered by somebody at a party, right? Yeah. And they're really like on different drugs than you. Yeah, I think he's on different drugs than you. I think that's for sure. And a lot more of them. But like, what was it again? Here that. Oh, yeah. So like, but like, at the base of it, they sold way fewer cars. They like and sold them for cheaper. You know, they cut prices even to move what they did. And then they like their their profit went down and like the stock goes up. But like part of that is because like he's selling this

dream and he he he does like one of the there was good news. Like they were he was like, OK, in in June, which is like a month this year and not, you know, some many years ahead because he's always making up timelines. He's like, we're going to have like, you could get, you could hail a ride hailing, you know, robo taxi in Austin, like, or you could get an autonomously driven Tesla in Austin in June. I mean, so that's good. He also says they're on track for the cheap car. You know, there's a lot of things in the future that are going to happen. But like also Waymo, you know, Google's thing is already in Austin and a bunch of places like is already doing this stuff. It just felt a lot of like, like he was pointing

every which way other than like the thing directly in front of him, which is like they're not selling as many cars. No. Well, I mean, I think it's again, it's like if Tesla is a I mean, it's so strange to me because you go back, you wind the clock back five years and Tesla is feels very like I mean, they're just like kind of somewhat unassailable in terms of their like dominance in the electric vehicle space, you know, and and Musk is not the Musk we know today. I mean, there are certainly some signs or whatever but he's just much more interesting and innovative than he is like political and sort of polarizing right and

and so the Looking at this progression. I understand why there's the Tesla audience That's like buying into the vision the broader vision of Elon Musk and Tesla. Right, as an AI company or robotics company. It's like the lightning rod. It's like the lightning rod, yeah. But it is at some point that there has to, I mean, I don't want to say it, but rubber might hit road.

I only used like two puns in my article. It's tough. It's really tough. But, you know, and there's a question of like is Is this a sustainable and I don't know I mean look to be honest with you I don't we're in sort of a very strange place with the markets and with these companies particularly the ones that are attached to The political the political world because now like I would have said like This seems like the wrong strategy.

Feroz-Doté For an electric vehicle company. Lyle-Tombs Like buying Twitter, the whole thing, buying Twitter and all this self-driving stuff that was a disaster and then not really focusing on the company and slipping in sales. I would have said all of this stuff is pointing to something's going wrong with Tesla and Elon Musk by proxy. But now it seems like it's all worked out pretty well, at least in the near term. But also his position in his ability to control the market, to control perhaps corner of the

market in the U.S., which obviously is very valuable. Though worldwide, it's not the same. He doesn't have the same opportunities. But, you know, even the market in the U.S., though, they used to be like they were the only company and obviously competition has come in, but every quarter that percentage of like they still sell the most electric cars in the U.S., but if you ask people what they want to buy, it's like a Toyota. If you ask electric vehicle buyers what they want to buy, like you gov did, they say, I want to buy a Toyota.

They're losing that head start. Right. And I think that, yeah, that's, I mean, they're losing the head start, but it's really like now, I think it's so not lassoed to the reality of the market. Right. Divorce from fundamentals. Right. Divorce from fundamentals that it's kind of like, does it matter so much if they lose,

if they're slipping further and further, as long as Elon Musk continues to outperform culturally, I guess, overperform culturally? And I guess, I don't know, but is this the same? Is ultimately the playbook here the one that Musk is running, like the German, like everybody's like, oh, he's like, you know, siding with the far right party in Germany and trying to get those people elected or whatever. And it's like, is it just because it presents

for Europe a similar opportunity for his businesses as he has in the US now? Like, or is Elon Musk legitimately concerned with politics? Is he really worried about the state of politics? I guess I don't know the answer to this question. AMT – One thing I think about a bit is maybe he just has such an unfathomable amount of money that he could do whatever he wants. He could make these bad decisions. He could have his foot in all the wrong doors or whatever. He doesn't really, he's not beholden

to making mistakes with money. Just do whatever he wants. Right, but I'm just saying, is this all just like, is it just, should I look at it as just more of a, I mean, I'm trying to figure out if this is really just a cynical, how do I position my companies for success type of situation and make more money in the process? Or is it like, the man has a legitimate love of the political process and wants to have

a big impact? Or does he have so much money that he can force outcomes no matter what? Yeah. I mean, I don't know. It's really... All right, what else? Anything else in the news? Absolutely zero, nothing.

No? No, I don't know. There's a lot we talked about. That's it. Trying to think of anything we missed that we should have covered. What else anything else in the news absolutely zero nothing? No, no, I don't know Trying to think of anything we missed that we should have covered It's been such a I mean, it's been such an insanely long week It's hard to believe that the deep-seek situation was only five days ago and when it feels like 500 days ago It feels like so long ago There's the Javon's paradox like that's what all the people who are like all the companies

that are big into AI were like, well, even if everything gets cheaper, people will just use more of it. What is this argument about the energy use? The argument is about Deepsea coming in and making AI much cheaper potentially. And like the, the, the, Javon's paradox is like, okay, but that'll actually increase demand. It's not going to, it's not like demand is like finite. If it's more available and cheaper, that'll just mean that like the people who have all the big chips and all the big AI will do even better. Like it's more is more.

Is this the, this is literally the, um, risey tides lift all boats. Is this the this is literally the Rises tides lift all boats. Is that the that's exactly what I wrote. They came up with a fancy name for it I think but apparently it's from like coal times Okay, so it might I don't know boats are coal which one never I suppose. Well, uh, you could say that but You know, there's winners and losers You could say that, but there's winners and losers. You could definitely be like, hey, this is good for everybody. The DeepSeek is kicking ass or whatever, but it's not good if somebody's like,

hey, I did an OpenAI, but it's more efficient and cheaper and more energy efficient and faster and smarter. And I should use my model because it actually makes a lot more sense. And everybody's like, okay, open eye, nice going, but you had your shot. We gave you... And they're raising around $40 billion or something? Aaron Powell They're assuming that because they are so much more powerful, even if you're able to do it cheaper, they could do so much more

with what they have. I don't know. Kline What do you think Deepsea could do with $40 billion? If they're really untethered from that kind of money and they could get it, like what would they do? Or if they had like the biggest chips, like if they were playing with like, you know, the smaller ones, I mean, it does raise a question. I was also like at some point, I mean, I don't understand, like aren't many of these chips fabricated in China? Are they all, are they not? Like aren't many of these chips fabricated in China? Are they all or they are not? Like aren't many of these components fabricated in China?

Am I crazy to think that? I don't know. If they're not being fabricated in America, I know that. They're not making it. I mean, right. That is not where they make them. They were trying to make them in Taiwan. Taiwan is one of them.

They're trying to bring chip companies to the US. Where are chips made? Where are chips made? Where are chips made? I know Samsung makes a lot of chips, but they're not, are they providing, I don't think they're providing Nvidia is making its own chips. They fabricate them somewhere. Let's see if Google's AI overview is correct. Asia, the United States, Ireland?

Asia is broad. Yeah, Asia is broad. And then being like Ireland, Israel and Germany. Okay. Can we get more specific about Asia? Because like it kind of matters. It's a weird overview. Like about Asia? Because like, it matters. It's a weird overview of Asia. It's like, okay, that's a continent. This article says production is primarily located in Asia in hotspots such as Taiwan and South Korea, and also the United

States. Okay. So Taiwan, which at any moment now could be invaded by China. Yeah, China, yes. Any moment. And in fact, we have to assume that it would be very easy for China to get things out of Taiwan if they wanted. Right? Like, again, I'm not talking about pure geopolitics. I mean like proximity. I think if they really wanted to like sneak out some plans for these chips, I think we're all in for a very, I think this concept of American dominance, especially given our current political climate, if American dominance can only be expressed in retaking the Panama Canal or whatever and we kind of don't actually- And renaming the Gulf of Mexico?

Yeah, the Gulf of Mexico to Gulf of America. And I don't mean this in a political way. I mean, it's simply to say there's a lot of incentive for countries that we are being adversarial to, to be adversarial to us. And I think the more adversarial we are, the tougher it's going to be to create healthy, profitable relationships. And I think, not to say that China and America have necessarily been best of friends, but there have been periods where things were really working pretty well.

It's funny, actually, I remember thinking a decade ago, like, wow, we've really embraced China and don't seem to have as many problems with them as like a country as we used to like we're still kind of like Russia sucks, but there was a period in America where it felt like We were kind of like China, you know, yeah, they do some stuff. It's kind of messed up. But overall There was some like diplomacy happening. I mean diplomacy, but also I mean, maybe there's also a little bit of like I mean And I'm not you know, I'm not trying to call out any political groups, but there's probably also a little bit of like being there's probably some, some amount of like, I don't know, like, you know, at the end of the day,

like Americans, I mean, I grew up in a country where people were always talking about like spreading democracy, you know, and then it's like, when you see America like partnering or like working really well with a country that's like, it has a completely different political system, one that is like very repressive in a lot of ways and kind of like bad in a lot of ways. That is certainly not a democratic political system. But now, of course, in America, I'm like, well, we what do we what can we really say about it? Like, to be honest, like, like, who are we to talk? Who are we to talk? We are also a sensitive country who like wasn't't that the Google thing, like they had

to list the US as like, kind of like they do with China and like the Indian border, like things like that. Anyway. I'm sure, I'm sure there's some designation for us at this point, but what I'm saying is again, yeah, like we can't really be like, well, democracy though. Anyhow, we're not doing that. We're not even trying that. I just think we are taking like a more adversarial stance like to the world at large. And you know, some places that make sense

and then some places like with with China, like just they're just going to go off on their own with this, you know, and I don't know, man, they're pretty good at technology. Like China has proven to be pretty darn great when it comes to creating. Was there like a daily show joke? They're like, oh my God, who would have thought like China was going to make it cheaper? I mean, I don't know if there was a joke about that, but there certainly it certainly is like the reality, which is like, they're really good at taking things apart and understanding how they work.

The Chinese tech industry has been extremely innovative, especially in fabrication and production. Anyway, it'll be interesting to see how it all unfolds. There's going to be a lot of global competition I see happening. Is this the new Cold War predicated on AI dominance? Something tells me no. Flippenberg No. It's Web 3 and NFTs. Kline Web 3 and NFTs are back, and China's cornered the market on them. We've got to wrap up. One thing I want to say is, we're actually going to put tomorrow on a bit of a hiatus. There's some work that

we're going to be doing that I need to do and really focus on. As much as I love doing the podcast, we're going to take a break. And then probably at some point down the road, get back at it. But as much as I love doing this, there is a lot going on right now. And so we will not be back next week with more tomorrow. Yeah. Thanks for listening. Yeah. It's, you know, Ronnie, I'm going to miss our weekly conversations here, but, but, you know, the show is called tomorrow for a reason. And when I figure out what that reason is, I'm going to start making new episodes. Anyhow, that's our show for this week. And we wish you and your family the very best. Thanks for watching!