Factually! with Adam Conover - The A.I. Bubble is Bursting with Ed Zitron

Episode Date: July 3, 2024

Big tech is betting tens of billions of dollars on AI being the next big thing, but what if it isn't? ChatGPT burns obscene amounts of cash daily with little return, Google's AI dispenses useless and sometimes dangerous advice, and a recent study showed that tech companies will soon run out of new training data to improve their AI models. If AI is really so costly, unreliable, and limited, what happens to the industry that has bet so big on it? This week, Adam talks with journalist and influential tech critic Ed Zitron of wheresyoured.at to discuss the impending burst of the AI bubble, the hubris of Silicon Valley, and how we suffer under big tech's "Rot Economy."SUPPORT THE SHOW ON PATREON: https://www.patreon.com/adamconoverSEE ADAM ON TOUR: https://www.adamconover.net/tourdates/SUBSCRIBE to and RATE Factually! on:» Apple Podcasts: https://podcasts.apple.com/us/podcast/factually-with-adam-conover/id1463460577» Spotify: https://open.spotify.com/show/0fK8WJw4ffMc2NWydBlDyJAbout Headgum: Headgum is an LA & NY-based podcast network creating premium podcasts with the funniest, most engaging voices in comedy to achieve one goal: Making our audience and ourselves laugh. Listen to our shows at https://www.headgum.com.» SUBSCRIBE to Headgum: https://www.youtube.com/c/HeadGum?sub_confirmation=1» FOLLOW us on Twitter: http://twitter.com/headgum» FOLLOW us on Instagram: https://instagram.com/headgum/» FOLLOW us on TikTok: https://www.tiktok.com/@headgum» Advertise on Factually! via Gumball.fmSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

This is a HeadGum Podcast. I don't know the truth. I don't know the way. I don't know what to think. I don't know what to say. Yeah, but that's all right. That's okay. I don't know anything. Hello and welcome to Factually, I'm Adam Conover.

Thank you so much for joining me on the show again. Artificial intelligence is supposed to revolutionize the world and our economy. The major platforms are spending tens of billions of dollars on it just this year alone to gain an advantage. It's one of the largest infusions of cash into a specific technology in the history of Silicon Valley. But while we're seeing all this investment, what we're not seeing yet is anything approaching a revolution.

The companies tell us that AI is about to get massively better, that this is just the beginning, and that a mega-powerful artificial general intelligence that will transform work and our lives is just around the corner. But there's a big problem. Large language models require material written or generated by humans to train on. And a recent study found that they might already run out of training data in just the next couple years, which means those models won't be able to get much better.

And in the meantime, all of the AI companies are burning cash. ChatGPT loses $700,000 every day that it's running. OpenAI's recent release of GPT-40 lets users interact with audio or video, but the underlying model is the same. It's not actually any smarter than it used to be. It's just a little bit slicker now. And it's not going that much better

for the other companies either. Apple just did a big reveal of their new Apple intelligence, which will let you use AI to turn a photo of yourself into an emoji, which, okay, I mean, who gives a shit? And meanwhile, Google has had to deal with the humiliation of people posting examples of their AI giving insane answers to boring normal questions. No, Google AI, you should not put glue in your pizza sauce,

and it's kind of weird that your brand new product told me that I should. So there's the suspicion growing that AI is nothing but marketing hype, that it's the same as crypto, web3, and the metaverse, just a new way for these tech companies to prop up their stock prices because they don't have any other real ideas. So is that the case? And if so, what is gonna happen when the bubble bursts? Well, on the show today,

we have an absolutely incredible guest to answer those questions for us. But before we get into it, I just wanna remind you that if you wanna support this show and all the conversations we bring you, you can do so on Patreon. Head to patreon.com slash Adam Conover. Five bucks a month gets you every episode support this show and all the conversations we bring you, you can do so on Patreon. Head to patreon.com slash Adam Conover.

Five bucks a month gets you every episode of the show ad free. We've got a lot of wonderful community features as well. We'd love to have you join us. And just to remind you, if you want to come see me do standup comedy on the road, head to adamconover.net for tickets and tour dates. Coming up soon, I'm heading to Phoenix, Arizona and Toronto, Canada, adamconover.net for those tickets.

And now let's get to this week's episode. To dissect the waves of AI hype being pushed by some of the most powerful companies on earth. My guest today is Ed Zitron. He's a writer, he's a PR guy, and he's also one of the smartest, fiercest, and most influential critics of tech out there, which he puts in his influential newsletter,

Where's Your Ed At? Please welcome Ed Zitron. Ed, thank you so much for being on the show. Thanks for having me. You have a wonderful newsletter called Where's Your Ed At? Correct. Where you are documenting more than almost anybody else,

what exactly is happening in the economy, especially in tech. You use this phrase, the rot economy. What do you mean by that? So the rot economy is this growth at all costs ecosystem that's kind of eaten all the markets, and it's most prevalent in tech. Where it's not about better, it's not even about making more money per se, it's being able to show that you will grow forever.

That you will show 10-20% growth per quarter, forever, perpetually. And what this means is, you see products like facebook especially facebook best example which are made worse they are harder to use the remoraid stay interrupt you wanna get in your way more the interrupt. You trying to do the thing that the website is meant to be for connect with friends for example so that they can show will straight up with growing, friends, for example, so that they can show Wall Street, oh, we're growing. We're not only increasing revenue, but number go up. And it really is the number go up mission. Facebook also hilariously, as of last quarter, no longer reports their actual user numbers. They report daily active family people, which is a really confusing metric. I often feel walking around, oh, I am a daily active family person. Yeah, that's how I feel today. But what's great is when you go and look back, they've mentioned this metric before and it's

just completely different to the other metric, it just makes no sense. Yeah. They're playing with the numbers, it's a fake number. Yeah, it's... I wouldn't say it's Enron-level, but it's Enron-adjacent. But you see it with other companies like Google. You just did your video on Google. Google wanted to increase the number of queries on Google.

Which doesn't mean Google is good. It means that more people are searching for more, which usually suggests that Google is not doing its job right. Yeah, because the more you have to search, the less good search results you're getting. You're trying three search results when previously you only needed one. Exactly. And the problem is, for the user there, is Google is worse, so you use it more, but that's a solution for Google, because you're seeing more ads. And also, number of queries goes up.

Yeah. Great. And this is really the raw economy writ large. It's when products have kind of escaped the pen of utility. There was usually, it used to not be quite so obvious a con, it used to be, oh, this is a free product, we make so obvious a con, it used to be, oh, this is a free product, we make so much more money, like, we should be paying you, as well as this being free. But we're all happy here because the product does what it's meant to. Not really the case anymore, in a lot of cases at least.

ALICE You make the argument that Facebook is dying, in what way? JUSTIN Yes. So, traffic to Facebook has been dropping off for like three or four years. I mean, I deleted my Facebook account totally four or five years ago, and I've never once missed it. I'm still, I'm an Instagram addict, and I kind of have to be for my work. As a comedian, literally people ask me to do shows on there, that's where I see other comedians, it's like that is the town square for what I am doing, for better or worse.

Facebook never comes up. It just has disappeared from the social world I inhabit. ALICE So what's insane is, billions of people allegedly use Facebook. LIAM Yeah. ALICE Now, you go in there, and it is kind of like Instagram, it is thing you've followed, carousel of people you might know, add, add, carousel of videos you can't, that show, it's like

a three second loop, add person you follow, add person you follow, or sponsored content. What's also crazy is it's full of AI content now. Just weird, like Shrimp Jesus is the, when everyone sees it, like it's just a shrimpy AI Jesus. It's Jesus but he's like made of shrimp. Made of shrimp. Or like, someone. Not even like cocktail shrimp, like crustacean, like in the sea shrimp.

It's so- I have seen both. Okay. And as a connoisseur of the shrimp, um, yeah, I just wanted to be clear, shrimp does not refer to the form of Jesus, because he's actually usually quite buff. What's also weird is, you see a lot of these ones where it's like, no one will share this on a poster with a million shares. Yeah. And it will be a disfigured child begging for money on the street, and people responding

just like, wow, this is so terrible, I will support this. It isn't obvious whether these are real people or not, but Facebook as a product has just deteriorated and the traffic's gone down, web visits are down something ridiculous, like 20%, 25% over the last few years, along with most of the major internet platforms, and actually the product itself functions so poorly. One of my favorite things I found recently was, if you type Facebook complaints into the search bar, you can find six, seven different groups with thousands of people all complaining to what they think is Facebook.

But when you go in there, it's just the most obvious scam artist saying, like, oh yeah, DM me, I'll help you get back in there. And then the occasional Judas goat with someone responding who's another scammer saying, oh, this person really helped me. And it's so strange because this website makes billions and billions of dollars, but the actual product is broken broke. You log in, you're like, there's nobody here. It's a dead mall.

There's sort of like deluded boomers walking around, like people who don't have their wits about them, sort of stumbling about the place, but no like actual activity of the type that we were used to 10 years ago, or the type that would indicate there's a thriving product happening. I wouldn't go on there to talk to a friend, which kind of feels like the point of the type that we were used to ten years ago or the type that would indicate there's a thriving product happening I wouldn't go on there to talk to a friend which kind of feels like the point of the thing

Yeah, and even Instagram the only reason I keep Instagram is I have three or four friends on there who for whatever reason Maybe they hate me is they like they're not great with text. They're not great with email They're amazing on Instagram DMs. Yeah, and I guess it might just be the lanes of brains that people go into. Instagram is, right now, pretty bad too, as you well know. In the sense that you do get algorithm, algorithm, person's thing, algorithm, algorithm. But I think it's going to get so much worse. I think once Facebook really falls apart, they're gonna go, no, we need to make Instagram even more profitable. Add aeri at the helm, just, nimbly screaming at people that we must show more ads. Did you see a few years ago when, with Kylie Jenner I think it was, when she complained,

the funniest thing there was, she was saying, we need to stop, and her and... Yeah, she posted, uh, we just want to see photos of our friends, like, what are all these videos, like, Instagram was just for photos of our friends. Like what are all these videos? Like Instagram was just for photos of our friends. What happened? And I love Adam Masseri. I hate him. I think he's disgusting. He's the guy who runs Instagram under Facebook.

He used to be the head of Facebook's newsfeed as well, which is just the Michael Jordan of breaking the internet. And he responded to that by saying, we're actually going to show you more videos, which honestly I have to give him credit for. Just like one of the most well-known celebrities in the world. And he's like, actually it's gonna get worse, what you gonna do? Use TikTok? No, don't use TikTok. Don't. Don't use TikTok. Well, but that is striking because celebrities like her used to be the main currency of Instagram.

This was sort of one of the points of social media was, oh, hey, this celebrity is actually on there, actually posting, you can comment, they might see it. Yeah, it's the imagined proximity. Yeah, well, and in some cases, like real proximity, people used to have those sort of casual interactions with celebrities or people they were interested in. And then there's the effect of, you know, the old effect of Twitter of this being sort of a large,

casual social network for lots of different professions, you know, everyone's sort of visible to'know, the old effect of Twitter of this being this sort of large, casual social network for lots of different professions, y'know. Everyone's sort of visible to each other, everyone's posting things. And having generally a good time, and then you see ads as well, and that's what you get for it being free, and it seems to have disappeared, that version of social media. ALICE Yeah. And Twitter is a particularly weird one one because it wasn't that bad before Elon took over. But Elon Musk- JUSTIN It wasn't growing.

ALICE It wasn't growing. But I think users were going, and it was not profitable, but had been profitable. It was profitable, Jason. It could have trundled along, it would have been fine. Then Musk comes along and says, hey everyone, you've never seen me run a company before publicly, it's going to be so epic and based. And it was just... He has ruined it. Mostly because he fired the trust and safety people. I don't think people realise how important those people were, because it wasn't just them stopping people saying slurs, which has of course increased,

it was them stopping spam. And the spam on Twitter is so bad, now, you just get, you will post a normal thing and there is just, hole. Someone posting hole in the reply. Just a bot, instantly. And I don't mean that in a, oh you click through and it's, no it's just, that's the picture. I mean, that's become a meme now on Twitter in the way that, you know, whoa this blew up, go to my SoundCloud, it used to be a meme. Now it's a P***y and P***y and Bio. Let's bleep P***y, but yeah, P***y and Bio.

That is that is the new meme of what is in the reply of every single Twitter. It's escalated now where it's just naked. Yeah, just full hole. And it's it's crazy. But that was Musk attempting to grow the website. He's not a good CEO. He's very bad at this. That's why it went so poorly, but that was his attempt to go, okay, I will take Twitter and I will just make it grow, I'll make the

money come out of Twitter. And it doesn't work like that, because social networks are very precious creatures. You touch one thing, a lot of things break. And he touched a lot of things, and a lot of things are broken now. But you're kind of seeing, with Instagram and Facebook, what happens when someone better at this does it. And Mark Zuckerberg's a horrible creep, but he knows what he's doing, and also everyone that works for him, they're all part...

Javier Olivaén, the COO, Boz, the CTO... Boz the CTO, by the way, did an internal letter called The Ugly back in 2016, 2017. And within that he said all things are justified for growth. It's a crazy thing to write. He talked about a terrorist planning an attack and killing people because of Facebook, using Facebook. And then he said, but we connected people. But that is what Facebook is.

Jesus Christ. I know, it's really horrible when you look back and you see these things and they said them out loud. But when you see what's happening today you kind of say, oh, that makes sense. And a lot of what I write is just me looking at old stuff and going, oh, no, I don't like that. That makes more sense than it should. And then you kind of look to them to be better, you try and look even optimistically for a

second. It isn't anything, it's not, they're not even trying anymore. They're not even trying to pretend that they care anymore, and that's really what frustrates me. Well, what would it even mean to make Facebook grow in a way that helped people, right? Or in a positive way, because they already had a third of humanity signed up for this service,

which is, I mean, that's gotta be, you gotta be pretty close to the ceiling there, considering how much of humanity lives in dire poverty currently, you know, of the seven or eight billion people on earth, how many is it, whatever. And so if they're showing growth, if they need to show growth, the only way to do that is to continue to squeeze the juice out of the existing users. You can't continue selling the same thing you were selling back in 2010 to more and

more people. ALICE That's the problem. So there is data I've seen that suggests that only 100 million people, more people, got online between 2022 and 2023. So, as opposed to like hundreds of billions of people getting online. So the amount of people getting online is going down. It's not going down, it's slowing. Right, the low-hanging fruit, almost everybody who has the means to get it now has it, or we're reaching the end of that process. With the big problem being that the global south is not getting online, there are serious issues there

that are just not being solved. One might think, and Mark Zuckerberg actually flirted with this idea before about the third world and trying to get people online, and he just didn't do it. Right, there was stuff about having a really low bandwidth, low internet Facebook, like you're going to be able to get Facebook on a phone that can work out in the middle of nowhere with almost no battery, that kind of thing. Just as an attempt to, we've got to keep the numbers

going up. I feel like in India there was some sort of legislative thing, but nevertheless, that's a classic Facebook solution to a different problem. The actual problem there would be sink hundreds of millions of dollars into internet infrastructure. If you want to actually do this, fix the problem. Don't make a new Facebook to fix. And that's a huge problem that the internet companies, I think worldwide have never grappled with.

The fact that internet infrastructure like limits what they do. This is a little bit of a separate issue, but it's just something I think about all the time. How Google a couple of years ago tried to launch, you know, a video game service called Stadia, where you would, where they, where their, their premise was. I'm sorry, this is off topic, but it boggles my mind to think about it all the time. Their premise was computers are too expensive.

Instead, people should rent the computer from us. We'll have it in a data center and you'll just like send the frames back and forth over the internet. Your, your controller data will go to the server and it'll send back frames. So you'll play Call of Duty running on our computer, but sent over the internet. I was like, that's completely fucking backwards. Computing is very cheap. People are able to get like incredibly powerful computers at a very low cost right now.

No one has good internet. People in the US are paying hundreds of dollars a month for internet that's worse than anywhere in Europe because our internet infrastructure is so bad. And Google, I think at some point had a fiber optic program to like try to bring fiber to people, that failed. And like that should be, if they wanna grow the internet or if they wanna grow their products in the US,

that should be their main focus is making sure that like everybody has fiber and they just didn't do it, and they're facing the consequences to a certain extent. So the irony of all this... don't know if it's irony... there is a connecting fabric, though. And all of this is, no one invests for actual... they want growth, but they don't invest for it. Google invested with Google Fiber, and it grew in some places, it was fine, but instead of investing on the time horizon you should with something like this, which is twenty years, they gave it five, ten, they were like, not really making us a bunch of money, and we don't really think of the future,

so we won't. So they didn't. And they cancelled it just like Google Read or Google Wave, all those other things, just, it's another little side project while they keep raking in money off of ads. Yes. It's frustrating, because Google Stadia, for example, there is precedent of this working, an Xbox Game Pass, a PlayStation remote. There are internet streaming things like that already. Now it's streaming from people's consoles. People do like the idea of streaming gaming, but like you said, the internet infrastructure

isn't there. And also, it's not good enough yet, even if it is, it's just, there's something missing. And they're not really thinking, why don't we make it really good, then we'll do it. They just went, well, we got it, what was it, 2021 they did that, it was 2020, they just did it, at the one time it looked like it would work, and then they just gave up on it. What they should have... They tried it for a year, and then they just, ah, nevermind. A whole entire year as well.

It also was a sucky program. You had to buy the games, but you didn't. You bought the games only on Stadia. And then they ended up refunding everyone, which is nice, but also, why do this at all? Regardless, with Facebook, with Google, with all this, they don't think of the future. They don't get to think of the future if they want to please the markets. Everything has to be now, now, now. And if it's the future, it needs to be the vaguest possible version so that they can

roll it back if it doesn't work. So that they can change it so that Jim Cramer screams at them in the right way. And it's frustrating because, like you said, Facebook is running out of people to sell to. And Facebook knows that the moment the market realizes, the market is going to punish them. Punish them hard. Now Mark Zuckerberg can never be fired.

He has, they have this labyrinthine class A, class B thing where Zuckerberg controls everything. Thank you Sean Parker, by the way. It's all his fault. He is the one that negotiate with Peter Thiel to have the third board seat go to Zuck. This is the guy who was played by, I think, Justin Timberlake in the movie?

Way more charming. Whenever I hear his name I'm like, Justin Timberlake is who we're talking about? Yeah, yeah. But that's the thing, Justin Timberlake, way more charming, way cooler. But Mark Zuckerberg could do anything, which is why he's sunk 30 billion dollars into the metaverse. But he could just say, screw you, I'm gonna have this company work sustainably, it's gonna be a great company, still gonna print money, but he likes money, he likes power.

And he's always been like this. Back in 2008 they were buying Google ads to get Facebook holdouts onto the site. Company's been rotten since the beginning. But there is no reason it has to be like this other than the fact they're scared of the markets. They're scared of what the markets will do, but also, they don't care. None of the people running these companies actually give a shit about technology. They don't build technology.

Sundar Pichai of Google, CEO of Google, Alphabet as well, McKinsey Man. Mark Zuckerberg, Agile Coder, but also, I had a bunch of growth freaks. The Javier Levin, CEO as well. All of these people are growth freaks. And what sucks is, it really does not have to be this way. It just is because it's super profitable for now. But I see everything slowing down. Everything is slowing down and they don't have a next big thing to sell to the markets.

When the markets realize that, they've got nothing else. It's not like they have a great product they can point to and go, oh, people love this. Look up how much everyone loves Google right now. They love, oh wait, they don't because we put AI search in there to show the markets how cool we are. They're no longer building products for people like you and me. They're building things, symbolic capital, I call them, to show off to the markets and say, look, we'll grow forever. We're so innovative and we'll be innovative forever. It's very frustrating.

It feels like we've turned this corner where the tech companies used to sell us things that were genuinely almost magical products. The iPhone, Google, Facebook, these were things that you showed the promise of them to people and every single person said, I want that right now. I need it in my life.

It'll make my life much better today. I remember when I got an iPhone and suddenly I can look at a map wherever I'm going. It changed my life. Your voicemails, that was the big thing. Visual voicemails. Yes. Oh my God, that saves so much time.

But that was all the low-hanging fruit. And now they are in order to keep up that promise to Wall Street, to keep up that promise to Wall Street, to keep up the promise of we are going to keep growing like that forever because that's what we did for the last 10 years. For the next 10 years, we've got crypto, AI, VR, and the public to all those things is saying, what? Yeah. Like, why would we want that? What is that for?

But the companies are like pretending as though everybody wants it, they're making concept videos, and they're tricking Wall Street to some extent, but it's all like a fake version of what they used to do. Yeah, so, call this the Rock-Com bubble, where for about 20 years, they have found all of the hyper-growth markets on the enterprise level like cloud. We've gone from a point when computers were relatively decentralized in the sense that getting a file from you to me was quite difficult to ubiquitous enterprise and consumer clouds. So you can share a file with someone quickly and it's kind of seamless. You don't realize how much is going on behind the background.

We have really good batteries. We have incredibly fast smartphones. We have apps on our smartphones that connect with each other. We have all of these things. All of these, we can now rent a property in a few clicks. We can do all these things in a few clicks that we couldn't before. These were all hyper-growth markets. These were ways of building billion, hundred billion, trillion dollar industries. We're out. I think we're actually tapped out of them.

Except everyone is still acting like there is one. And so, VR's a weird one because, been around for a while, and anyone who'd used it was like, yeah, this is still niche. Up until now. I have friends who love VR who said, this is, my mate Casey says, yeah, VR is niche, and he loves VR. I love VR, I think it's really cool. When it works, which is a load bearing when. Yeah. But I mean, I tried it, you know, like I have a quest too.

I played, I'm a gamer. I played the Valve's Half-Life Alex game, you know, the best quality game from one of the best studios in a thing I loved. The game was really fun. I stopped playing it after a couple of days because I couldn't handle being nauseous for three hours after playing for 30 minutes.

Like, it just wasn't worth the experience. And nothing has motivated me to put on that or any other headset since. And no one has come up with a compelling pitch for what it's for, including Apple, whose only pitch seems to be, what if all of your emails were in the room around you? Well, who gives a shit? Who does that solve a problem for? So what's funny with the Vision Pro is I have one. And I watched half of Dune on a plane before getting a migraine What was really annoying was?

As much of the movie as I could take So cool. Yeah, giant screen. I've used like the desktop. It's cool. Yeah about two years too early They should have kept it in the oven, but I can see what they're doing with it What's also interesting is Apple's been surprisingly, for their new thing, surprisingly quiet. They've not been too pushy with it. They did the whole Apple thing, well, this will be your walk around an office with this. Just a very funny idea. They haven't talked about it in like six months.

It's not like they're making a big deal about it still. But then you go to like Mark Zuckerberg in the Metaverse for example. So just a complete lie. How the hell is that not some sort of SEC problem? He went in this video and was going on about, oh yeah, I'll go into the Metaverse and I'll hit a button and I'll change how Mark Zuckerberg here looks. Just complete fan fiction. Nonsense.

The whole thing was bollocks. But what it is was that was Mark Zuckerberg realizing, I bet I could say something that sounds plausible and none of the people in Wall Street and many people in the media, like CBS this morning, they're not actually going to care. They're just going to go along with it. The problem is that he's stuck in his own bubble, as many of these CEOs are, where it's... When you don't build anything of use for a long time, when you really have no connection with a regular human being, you just believe that everything you say is smart. And I think he thought, hey, this'll work fine, they'll just buy Horizon Worlds even

though it sucks. And it didn't work. But it worked in the media, there was a large chunk of the media who liked it. Wall Street kinda liked it, and then they saw how much it was costing them, and went, no, no, no, Mark, Mark, you gotta get rid of that, Mark. But I do believe that that was the first time you saw one of these CEOs becoming aware of the problem in public. He wanted this to be the next internet, he said it would be the next generation of the mobile internet. I think Zuckerberg knew back then that the growth train was slowing, and he needed a

new thing. Except he chose the most insane thing. He chose virtual reality, and also online gaming. So unproven niche tech, and just one of the most expensive things to build in the world, from a company that is terrible at building things. Facebook's launched like 15 different things and cancelled. The best thing they built is the Facebook portal, which is just a video camera. And all their other products are things that they bought.

WhatsApp, Instagram, etc. They don't innovate, but also, they only build things for growth. So now that they're out of ideas, but the metaverse was a sexy big idea that sounded, if it was real, if it was the actual thing he promised, would have been. Except to do what he wanted to, based on Ready Player One, which is a dystopia, he would need extrasensory tech that does not exist. Very basic things like touch and feel and spatial awareness, senses within the human body with the brain, that just cannot exist maybe hundreds of years away.

And it's so weird that I'm still shocked he even pretended and I'm more shocked that everyone agreed with him. I mean when you even looked at what they built, like it was laughably bad. It was so bad. I mean I did a video about it, all this stuff about like we've added legs and things. And they didn't even add legs. That's what's, they never actually added legs. That's what I, I mean, it looked worse than second life, which was, you know,

as 20 years old as a product now, um, it had no, you know, people, you know, the actual service was a dead zone. People logged in and had nothing to do on it. Um, and the idea that people would hold meetings there, all this stuff. It was like, it was hot. Like Mark Zuckerberg, at least gives the appearance of being a smart guy, it was hard to imagine him believing his own bullshit in that case. ALICE I just think he's craven.

I think he doesn't care. I don't think he cares whether it was real or not, he was just like, bet these little pigs will buy this. Just complete... That man... So I did some Facebook reporting and just everything you read about Mark Zuckerberg and the Facebook crew and the people at Meta, you're just like, oh these people are just evil. Like, there's an evil to it. They have

this guy named Joel Kaplan, the head of US policy. So there was a time during Facebook's wonderful history where the algorithm kept pushing this video called PlanDemic, Conspiracy Theory. So Mr. Kaplan, I believe, worked with the Bush administration. He pushed to not have it suppressed. He said, no, no, leave it alone. And the only reason that he ended up taking, Facebook ended up taking it down was because Kevin Ruse of the New York Times posted that it was the top video on Facebook. Wow.

And that was using a tool called CrowdTangle, which used to let journalists see inside the most famous things on Facebook. Yeah. And then Kevin kept posting that, hey, everything on the algorithm is just right-wing garbage to put on genome. Right. There was a lot of press about that, about, yeah, that, that. So, the top-rated Facebook things were all conspiracy theories. So Facebook's response to that, you'd think, would be, wow, we've got a real problem with

the right wing taking over. No, their response was to shut down CrowdTangle. So that no one could see. Alex Schultz, their CMO, real, real winner. And there's all of, these are the people running the company. Yeah. Facebook is unique in the sense that I think the other companies are just completely lost and run by management consultants. Facebook is just honestly the evil machine.

It's so funny how nakedly evil they are, and just like, I know it's dramatic to call things evil, but you look at what they're doing, you look at how they act, there's a great book by Jeff Horwitz called Broken Code that everyone should read. Read that book and come out of that and tell me whether Facebook is a nice company or a bad one, because you read it. Since 2008 they've been like this, and all the OG people who did all the original stuff with People You May Know, algorithmic thing that culture people you might know, used Google

ads to target people who weren't on Facebook. Outed sexed workers helped patients find psychiatrists who really didn't want to be found. Cashmahill did some great journalism around that. The people who worked on that, Natalie Gale, Javier Ollivan, Alex Schulz, they're all running the company now. It's the same people. And this is with my work, people have said, oh you seem angry. How can you not be when you read this stuff and you're just like, it's not even like there's

one person who said, yeah, this is bad though. They're just kind of like, eh, maybe we're abusing notifications a bit, you know, but that's business baby. It's just so, and it's so weird, and it's so, it's so weird. Some of the stuff you read in this, just, you're like, this is almost cliche in how deliberately bad it is. Yeah. Especially coming from an industry that, that for so many years presented how good it was, how virtuous, how this is a new way of doing business.

We're not like those old companies in the past. The Sheryl Sandberg, Lean In, this is a new thing, where, you know, this is the utopian future. Sheryl Sandberg is Mackenzie as well. Yeah. Yeah. They're all there. Yeah.

Oh, God. Well, talking about these, you know, fads from, I mean, the metaverse is almost two years old now as an idea. And no one has said the word metaverse to me nor have I read it in like a Bloomberg article in like 18 months. It's gone. It's dead. The end.

I think everyone has conveniently forgotten about, you know, Mark's little attempt to juice what was happening with his company. So true. Same with crypto. Crypto is basically, I mean, you know, whatever, the BitTorrent freaks on Reddit are still hanging out saying HODL, but you know, it's no longer Super Bowl ads

or, you know, something that is, you know, the next big thing in business. Now the next big thing is AI. It seems like everyone has forgotten about those previous past two fads. Yes. How much is AI, especially coming from Facebook, just another one of those? Like, is that just Mark trying to do the same thing again? So Facebook and AI is almost an entirely separate thing,

because who even knows what they're doing there? AI as a general fad has a bit more validity, but not much more. Jason Yeah, because large language models are, I mean, the first time I played with one I was like, oh, this is cool. And they're still pretty cool. Sometimes I go play with one every once in a while. They do something. Jason They do something. And that is also the problem, because AI as a concept has been around a while. Like decades. Jason It's a very loose term. It can be used to describe a lot of things.

Exactly, which is great for companies like Microsoft. So, Satya Nadella, 2021, says that he could not overstate the breakthrough that the metaverse was. Two years before that, he said that HoloLens 2 was a breakthrough. Their AR goggle thing. Yes, which have now been... They exist, but they've been... He's fired most of the team. And Satya is now a major beneficiary of the open AI boom. They've put, what, $13 billion in, they've invested early, and they invested $10 billion, mostly in Azure credits, so they're cloud services.

The thing is with AI is, it smells like the future. It has a feeling of the future. And when you say AI, people will fill in the gaps. Autonomous agents that do things for you. Oh, a super smart friend who knows everything about that. Right. It has the advantage of it's at least been around in science fiction enough that the public knows what it means. You had to explain what crypto was and what the metaverse was unless someone had read like Snow Crash or Ready Player One. But, artificial intelligence, oh, I know what that is, that's the computer that talks to

me from every movie I've ever seen. From her. Yeah. From 2001, from the movie AI... What I love is when they quote these movies, like, yeah, great things for AI. Like, 2001, the movie that classically ends with the computer helping the guy. But with AI, what's interesting is, it has both the smell of the future, but none of the taste.

Because the idea of what AI could be, this autonomous thing that does things for you, to quote Samuel, when a super smart person who knows everything about you, all these things sound plausible, and you can, with with chat GPT, approximate something that sort of feels like it, kind of. But the problem they have is all these companies are promising a thing that is not there. So Super Bowl commercial for Microsoft's copilot. During it, there's a bit where it says, write the code for my 3D open world game, and it hits it, and then it cuts away before you actually see what would come out. So the problem with that is that is not what comes out. I've typed this in many times, because I'm a little bitch, and it just gives you a guide

for doing it. And it's like an SEO article, like, first you code the game. Microsoft can't even show you why you should give a shit about this thing on their Super Ball commercial. They can't actually explain what it does today, so they have to talk about the future. Sam Altman, worst of them, talks about what it could be, what it will do, it'll be like a virtual brain, all these things. And the thing is, the people in the markets, the Jim Cramers of the world, the CNBCs of the world, much less Bloomberg. Bloomberg has done

some excellent journalism about AI. Not so much with CNBC, those in world, much less Bloomberg. Bloomberg has done some excellent journalism about AI. Not so much with CNBC. Those in Doc Hadenfield there is great. But nevertheless, the markets people just go, and artificial intelligence is the future. And this is good. And thus they kind of buy the bullshit train. But then the consumers get these things. They get chat GPT. They get co- Copilot, what have you. And then they say, now what? What's this meant to do? What's it do?

And I'm sure some person will now email me and say, oh, well actually it helps me with my productivity. And every time you ask them to explain it, they say, well first of all, I sat there and I did all this stuff. It's like, no, no, no. Okay, shut up, shut up. I don't care. Because we're being told, we're being told one thing and presented another.

We're being told our autonomous AI future is here, and this is why these companies would grow forever. Google, Meta, all Microsoft, they're all connected to the AI boom. Yet, the reality is, the actual products themselves, LLMs, don't really do enough. They do something. You can use them. They are kind of like the next generation of internal automation. They're like robotic process automation, like UI path. It's stuff that kind of exists, but a bit better. There are some cool things. Yeah, I mean I used on the phrase on the show before, I forget where I got it, but the idea that it's a it's a word calculator.

It'll, there is a way that if you put in a certain combination of words, you might get some output that you find useful to some degree, but it's not... A lot of this stuff's things you could have found on Google already. A lot of it is something that if you put in a little bit of time, you could figure it out, you could do it yourself. And it always feels hollow because it is. It's just, it's calculated. It feels calculated, surprisingly enough. But the problem is, as well, it isn just, it's calculated. Like, it feels calculated, surprisingly enough.

But the problem is, as well, it isn't that it's useless, it's that it might only be worth three or four percent profit growth. Like, it's not, it may be the future, but it might not be the next big thing. But they're selling it as if it is. And what the big problem is, is, going back to your old iPhone example, you could tell a friend exactly why that thing was cool. They'd say it was too expensive. It was like $600, it was a lot of money, but they saw it and they went, wow, this looks great.

Like, oh, you can listen to your music and text people. You didn't have to onerously explain why people should care. With ChatGPT, genuinely gone to my head, I'm not sure how I'd describe what it does. I mean, I'd describe the process, but what is ChatGPT, genuinely gone to my head, I'm not sure how I'd describe what it does. I mean, I'd describe the process, but what is ChatGPT? It can write stuff, kind of, but it gets it wrong. It can generate pictures of stuff. So, right? Yeah, you can talk to it, sort of, and it'll give you a rather boilerplate response. You can talk to it sort of, and it'll give you a rather boilerplate response. Yeah.

It can, if you want it to, whenever I'm playing with it, which I do sometimes, I'm always like, okay, give me a recipe for a pizza with poop on it. And then it does that. And then I say, now write that as a limerick. And then it writes a little poem. I'm like, that was mildly amusing. Like I've never seen a computer program that can do that before. You know, there's stuff like I talk, I've talked, I think before on the show, but there's a game called AI dungeon that like works like a text adventure, an

old text adventure. And but it can do anything, right? You're like, I want to turn into a wizard and fly away. And it'll just generate some text that describes you doing that. So it's sort of like a, like a one person tabletop role playing thing where like an automated DM who can, you know, spit out a fantasy for you. That's like, yeah. Absolutely nothing wrong with that. Some people might pay for that, sure. But what's the...

It's not like Steve Jobs standing on stage going, it's a phone, it's an iPod, and you can... It's a revolutionary internet device. You can connect to it from... Connect to the internet from anywhere that you are. You hear that pitch and you're like, my God, I mean, that'll change my life. Exactly. Like... You hear that pitch and you're like, my god, I mean, that'll change my life. ALICE Exactly. Like, you look at what GPT does as well, and you look at Claude and Anthropix model, for example,

you look at Lama of Metas 1, and you go, wow, okay, you can do stuff. Now what? Now, what now? Seriously, what is the next step of this? GPT-40 just came out, the thing that got them embroiled in the whole scandal is Scarlett Johansson. So, that was an iterative update. That's the best they've had. It's the last time they did anything new. And it's so bizarre because I feel quite lonely in saying this, but what if this is all they've got?

What if this is all they've got? Yeah. What if this is it? Because once everyone works that out and is willing to say it, the tech industry has backed this so vociferously. They've really gone to town on this one. To the extent that they feel forced to. I mean, Apple looks like they're doing all this AI stuff, like with a gun pointed at their head. They're like, we've got AI emojis.

We, we, oh, it's, it's gonna, right? We've got Apple intelligence. And like, they give this whole presentation and I was like, these are the features that I give a shit about the least that I've ever seen in an Apple keynote. What I really liked about that though, was the chat GPT integration into the new iOS. They're like, yeah, we've got chat GPT. It will sometimes answer some things and you will have to give it permission every time. 100% just something they did for the markets. That's just a bit, yeah, we've got chat GPT in this bullshit, whatever, who cares. But when you look at what they're doing

with their own models, it's kind of interesting. Oh, you can do distinct actions across apps. That? That's cool. Don't know if it's the next big thing, but it's a cool new way for the phone. But when you look at that sort of feature, well that just looks like that's just a regular old feature they're calling AI now, you know, that it's a little bit of automation in your phone. It is actually AI. It is actually their own models, but it's their own models run on device.

It's stuff that people have been doing, so edge computing, so on device AI, but yes, like it's been around in very simple example, the one I give is like, you can Raspberry Pi, so these tiny computers, which by the way, significantly bigger innovation than anything generative AI has done. People invented Raspberry Pi should be like, no but, please... Yeah, it's like a $40 powerful computer that you can put anywhere. Powerful enough that you can attach... You can have that, and it can have edge AI running for like, a machine that can identify a strawberry that's rotten on a farm.

Genuinely useful industrial stuff. What Apple's doing is just what you can do when you have complete control over your operating system and no one else is really allowed to touch it. The SDK that can connect the model to different parts of different apps could be really cool. But also, surprisingly demure about it. Like, it was a big thing at Worldwide Developers Conference, but there's no big ads, there's no big... Like, they've done a media push, but it's not huge right there.

And also, they don't appear to be really... I'm sure that they have giant GPUs, so graphics processing units, which are used to train these models. Sure they have them, but they've not been very flashy about saying, we've just built hundreds of acres worth of service space for the future. It seems like they're hedging their bets, like they're dipping their toe in because they feel like they have to, but not going all in. Yeah, and Apple, to be clear, no company's particularly ethical, but Apple seems to also

realize we piss off all our users and make our products suck ass, maybe they'll stop using iPhones. That's bad for business. They don't want to do what Google just did. The Google AI search. That one's so good, though. That one really felt like, I began to disassociate when I saw it. Because I'd been watching this for a while, because it was on Google Labs.

There's an SEO expert called Lily Ray who's been watching it very close. And she's been posting for a while, just saying, this is bad, they should not launch this, these results are bonkers. And then they didn't, supposedly. ALICE And Bing did the same thing a year ago. Like, it released AI results in Bing, and people made fun of it because it did exactly the same thing that Google's did. JUSTIN Except they were very cautious with it, and then, yep, this is a side feature,

please don't trust this. Google, you gotta give him credit for just being like, yep, there you go. And then when people found the terrible results, they said, actually that's good. It's actually good that you found that. And also, those are on common queries anyway, so, cause no one makes pizza. But as though, but they've also said forever that like, what is it, 15% of all Google searches are like searches that they have never seen before are completely unique. Like, their bread and butter is unusual Google searches.

That's the point of Google to help you find things that are unusual. Oh, yeah. The interesting thing about ChatGPT 4.0 to me, okay, so the AI boosters, or a lot of the people I know who are still frightened of AI, you know, here in Hollywood, the thing that a lot of people were saying last year during the strike is, this is just the beginning. It's going to get radically better.

This is, we don't, you know, they've got stuff cooking up that we don't even know about yet. It's gonna be so huge, you know, in five years, people were literally saying to me last year, in five years, Hollywood won't exist as an industry, you know, because A.I. will be so radically great. And, you know, I would have to say have a little bit of skepticism because this is what Sam Altman is telling people.

He's saying my product is going to destroy the world. So you should do what I want. Yes. But where is the evidence that it's going to get radically better rather than that? Hey, they've invented this thing called a large language model that can do this one thing, but it might not have actually this like incredible growth curve ahead of it. And I think some evidence for that for me was chat GPT 4.0 was not a radical improvement in the quality of the large language model or its results. It was a better productization of chat GPT where it's like, OK, now chat GPT is in a little app and it can.

It has the front facing camera. Responds fast. It responds a little faster. It has a voice now. Didn't have a voice before, but it's still saying the same shit. Right. Like it still has the same cognizance. It's not smarter. It just can sort of look around a little bit better. We put the same dumb brain into a slightly more advanced robot dog. But it is still a dog. Yeah. And so what do you think the prospects are for the radical improvement of AI?

Because I've started to read these things. Oh, they're running out of training data, that sort of thing. So the big problem they have is that training data. Yeah. So there was an estimate in the Wall Street Journal, researcher, I forget his name. It was the GPT-5, so the next model of chat GPT or GPT, their large language model, will take five times the available training date or the five times the amount of training data that it took to train GPT for. So big problem there is they don't have much more.

They are running out, and they may have run out. Well, chat GPT-6 takes another five times. That's an exponential problem. And on top of that, they don't seem to be too good at making new models. They were meant to spin up one more efficient model for Microsoft called Arrakis. Gosh damn it. they were meant to spin up one more efficient model for Microsoft called Arrakis. Gosh damn it. Like when I read that I'm like, fucking, just, god damn it.

And then it didn't work out because it wasn't efficient enough, which is not a good sign. But also, these things are not getting, they're getting better, but they're not getting better. It's not like the jump from, I don't know, GPT-2 to GPT-3. On top of that, the ways they're getting better do not suggest they're going to get better in the ways that Sam Altman has been talking about. A super smart friend that knows everything about you suggests features that are not part of a large language model. Like memory is something they've kind of worked on, there are other people who've worked on memory within these things, but the fundamental thing of knowing requires knowledge, which requires actual intellect,

which GPT cannot have. It's mathematics. It's extremely complex. Yeah. But also all of these people talking about, oh, it's going to be so much better in the next year. History doesn't look great for you either, but also how? How just how?

Yeah. Why do I have to constantly explain to you why your product is good? Why do I have to make your argument for you? Why don't you have one? Other than it will be good. And that's the thing. Everyone wants it to be because it would be cool, and also if there's no big tech thing coming out,

what's the future look like for tech? It's a scary one. I just have to continue living in the present forever? Like, ah! I need the computer to distract me. But also, it's a situation where so much money has gone into it. So many billions of dollars, so much money into data centers, so much money into marketing efforts, so much venture capital. Everyone is so deep down the hole, the kind of who's going to wake up first rather than

it getting to some point. Sora, the video generating model from ChatGP. I can't wait to talk about this. Yeah. It's so good. It's so bad. Yeah. Everything it makes has hallucinations.

So in hallucinations in the written word, it will say something authoritatively that isn't true. In Sora, it just makes like a monkey with seven arms and it looks terrible. And every time you ask it to do anything it takes- I think if it's 20 seconds long it takes 10 minutes or 20 minutes to generate. And it- the prompts, no matter how precise, it usually makes up some extra stuff. And the amount of training data they'll need to make Sora work is insane. On top of all of this, all so horribly unprofitable, and it doesn't do anything! It does not do anything. It doesn't make anyone money. Like, this is the big thing. People talking about, oh, well it's going to get exponentially better. First of all, it hasn't been. But on top of that, it's still not profitable. It's still horribly unprofitable.

And it requires all this energy, and it requires all this training data. There's so much more evidence that this is kind of tapping out in the generative AI space. Yeah. I mean, when SORTA came out, and let's be clear, that product's not available to the public. It's, they released some tech demo videos. But again, working in Hollywood, a lot of people I know said, well, this to the public. It's, they released some tech demo videos. But again, working in Hollywood, a lot of people I know said, well, this is the end.

Like they're gonna, this is gonna, the future of moviemaking, they won't need to film anything. They'll just like have Sora output shots. And, you know, I'd have to walk through with people as like, okay, first of all, this is not available to the public. They've released like six videos that they say were made by this.

So how many bad videos did they need to make before they got these six that they showed us? How long did it take to make each one of these? What was, y'know, we were talking about, y'know, 30 frames per second of 4K footage, right? And... ALICE Not even 4K, by the way. Oh, really? So I spoke to someone who makes, who, a guy from a company called Shy Kids, who made a

video with Sora. They had to get it at 480p, because it took too long if you went higher resolution. Yeah. Because, I mean, rendering times are very serious just with traditional visual effects. And so, you're talking about a large language model generating something, or it's not a large language model, but a similar technology, right? A generative AI. In Sora's large language model generating something, or it's not a large language model, but a similar technology, right? A generative AI. It's a large language model.

Okay. But, so there's that piece of it. Then, so the amount of processing that it has to do compared to making a dolly image is like exponential, right? Then you don't have the ability to make a tiny tweak, right? That you do in, you know, when you're actually working with an actor, you can say, walk a little bit faster,

walk a little bit slower. If you're, I know, because I've, you know, if you ask Da Li, generate the same thing, but make it a little bit brighter. It gives you a whole new image, right? And so it would be like that as well. Plus if they, you know, they made a great image of a lady walking down

a Hong Kong street, right? With a shiny floor, right? With shine, with, with there was rain on the street. And the legs at one point go like that. Yeah, yeah. But it looked pretty okay, right? It looked plausible, yes. But the only reason that it knew what it looked like for a lady to walk down a Hong Kong street

was because it had ingested huge amounts of footage of people walking down Hong Kong streets. Like it needed the footage to begin with to generate the thing, so it can't generate new things and, you know, you won't be able to solve the problem that the executive will have at the film company where they're like, oh, I don't generate the things. So it can't generate new things. And, you know, you won't be able to solve the problem that the executive will have at the film company where they're like, oh, I don't like the take. Can we get another one? No, actually we can't because if we do that, it'll be a completely different person

in a completely different situation. So that problem is unsolved. And you're gonna ask, well, isn't it fucking, and it's gonna be massively expensive to compute all of this over and over and over again, generate all of this. So at what point is it cheaper just to send a fucking dude with a camera to Hong Kong

and get a lady on a rainy day? Like, what is the but it's a great tech demo. But people are getting rolled by seeing the tech demo and imagining what it could do rather than looking at the reality of the technology at all. And that is the entire AI hype bubble writ large. It is letting people so journalists, let's be honest, filling in the gaps for these companies. Sora, great example. So okay, theoretically if the computer could spit out an image of a woman walking down

the street at any time, that would be great. But like you said, it is not just a question, you can't even ask if you say woman with black hair, Asian woman with black hair, different Asian every time, probably pink hair at some point, because of the training data is trained on. Well, and because the technology is fundamentally unpredictable. That's how it works. You can't say I want exactly this but slightly different because it is generating it.

It's probabilistic. Yeah. And it's funny because look at Chad GPT. When people write about the future of open AI, they're not talking about this shit. They're talking about what Sam Altman has been beating off on on TV for a year. They're talking about Sam Altman saying, oh yeah, well, he said to New York magazine in 2016, I think it was New York magazine, New Yorker, that in it would be not unreasonable to think within 10 years, 20 years I think it was New York Magazine, New Yorker, that in, it would be not unreasonable to think within 10 years, 20 years I think it was, that he would have a computer version of his brain.

And he's very good at making these things where the gap between the predictions is long enough and no one's really going to check. Yeah. I will check Sam. I've got it in my flipping calendar. I do. But you look at this and he goes out there and he'll vaguely say that chat GPT is the future, AI is the future, we need to be scared of AI.

He'll say AI is not a creature. But what he wants you to think is, did someone think it's a creature? Because he wants this mystique. Because if he actually had to talk about the thing he's selling today, he'd go, well it's a calculator of sorts. Well, it doesn't always work, but when it does, it's sort of good. And I can understand in business process automation, I mentioned UiPath, they're a robotic process automation company.

They're worth billions because of them doing these very minute, very simple customer service things, but because you can scale that, that's useful. GPT and large language models fit into this kind of thing extremely well. These companies are growing because of that. That will help them. That ain't sexy and that ain't big. That's not trillions. That's hundreds of billions at best and that's going to take a minute. And also you can't sell that to the markets that nightly. It's just not that nice. And also RPA is profitable. GPT, large language models, absolutely not. But because Altman goes in there and says, is this thing that kind of looks like the

future? Look, this thing can do kind of realistic videos. People say they're scared. I'm scared of what this could be. And they fill in that white space and they say, well because this thing can approximate how a human writes, this thing will write in place of humans within five years. Oh, because this video kind of looks like something you could film of real people, naturally in five years this will be how entertainment is generated. And the problem is, that's the old old, that's the old world when things

used to, there is no Moore's law for this crap. This is not a system, a situation where we're simply going to get more power and this will fix itself. Even if we threw so much more power at this problem, it would still hallucinate because that's maths, baby. That's how it works. And it still wouldn't be able to give you that slight variation on the previous clip because it would still be probabilistic. That's the nature of the technology. It's not something that they can just like cram more transistors onto a smaller chip

like they do with Moore's Law. And the problem is, as well, is, notice that Sam Ortman has been crying about how he needs more power, he needs an energy breakthrough, he needs more server space, he's begging Microsoft for a hundred billion dollar supercomputer. Yeah. Which by the way is contingent on a meaningful improvement to GPT. So he's screwed. But he's asking for these things because there is no way this goes much further. Yeah. Based on what we have today. And the only company that's really having a great time with this is Nvidia because Jen Hwangs can just like, yeah, chips faster. You little pigs will buy this every

day, and he kind of, you watch any of his keynotes, Nvidia CEO, and it is very much him just completely waffling. There's a bit in the latest keynote where he says, yeah, you know, you'll have a super AI, which talks to smaller AI, which then speaks to other AI, and people in the audience are like, woo, yeah, baby! Just fucking idiots. Like, but he doesn't really care because he's the guy selling guns to the war.

He's just selling chips. And so it doesn't need to work for him. He's going to make a lot of money right now. It works. What he's selling works. The chip does the thing. Yeah. Like he's fine.

He is also just, he has jumped on trends before. Nvidia has had a very weird history. Yeah. Games Nexus has some great videos on Nvidia. He's been waffling this bullshit before, he did it with crypto. That guy's just a classic business. He's almost kind of folksy in how much of a classic tech, but he's just like, yeah, faster, bigger, who gives a shit?

When you look at everyone else with these LLMs, there's this air of desperation. Facebook, as we mentioned, great company, they've put AI in everything now. There's AI in Facebook, there's AI in Instagram, you can ask MetaAI anything, and it will give you a half... Yeah, I was fucking around with it at 1am last night when I couldn't sleep. And it's now responding on comments, it responded on a parenting group. Because you can activate it, it responded on a parenting group, a couple months back, saying that it had a gifted child, and then recommended a school.

I was just, when I saw that I'm like, this is so good, I hope there's more of this, I genuinely hope there, I wanna see this thing cook everywhere. But that's such a great... That, when I saw that I'm like, oh, maybe that's what Zuckerberg wants it to do? If you need people to interact with a product, you need them to engage, what better way than poking viciously with it. Hey look, you've got people responding to you, you have someone to talk to. I don't think it'll work, but I can imagine someone being like, yeah, it's a really good idea. But on top of that'll work, but I can imagine some being like, yes, really good idea. But

on top of that, again, how is that profitable? Every single one of these transactions, every time someone uses this thing, it's losing them money. Yeah, because it is very high, takes a lot of compute to make this stuff. So expensive. And also, there's no profit on the horizon. Where's the money coming from? Like, taking away all the cynicism, just, where's the money, honey? Like, where is it? Yeahicism, just, where's the money, honey? Like where is it? Yeah. Are they running ads in the middle of the chat GPT responses, or what are they doing?

Even then, do you think anyone's, like, they're now saying they are, 404 Media reported, that they are trialing doing injected ads into people's videos on Instagram, they are ripping out the wiring from the wall. They are desperate. And I think the AI stuff is just them going, ehh, ego, we're doing the thing now, who cares? Zuckerberg doesn't care at all, he can't be fired, like I mentioned, he's just doing it. What's really interesting is Microsoft. They're the ones where I'm like, you guys are really in the hole. I think they are actually really in trouble. So, of course, heavy investor in OpenAI.

However, OpenAI, a non-profit, sure buddy, for real. But heavily invested in OpenAI. Heavily saying it's the future, but also tied to Sam Altman, who is a nakedly corrupt and very strange figure, but on top of that, promising the world with AI, but probably more aware than anyone of how limited it is, because they have all the money in the chat, GPT, they've got their own models. They're paying Oracle billions to build more data centers. They build data centers too.

Why Oracle is big on build outs, I guess. And it's very weird because they must know. They must know this shit sucks. And they might let the copilot think, where they had the... They could not even explain why you should use it. They are probably looking at this, just counting down the days. They're just like, I... There's a company called Inflection, which made an empathetic large language model called

PoE, I think it was. Microsoft absorbed them in the most bizarre way. They acquired them, but also gave contracts to the... They took all of the people from the company who made the large language model, then made a deal with the remaining Inflection company people to keep running it. And now I think the Department of Justice is looking into it because it's extremely weird. It was also worth a billion dollars.

Jesus. What are they doing? Like, these are not the actions of someone doing well, these are like... This is like, crackhead style stuff. Or it almost seems like a Ponzi scheme, where where it's, I always need to find the new injection of money. I mean, look, the, the, the, the sky high valuations of tech companies, all the money that flooded

into the industry, my understanding is it came in for two reasons. A, the costs were extremely low. Yeah. Like for running an internet company, way cheaper than building a new car company, right? You don't need factories, you just need data centers. It's like the costs are low. And B, there was a potential to create entirely new businesses,

entirely new business models. You know, the, what is it? The blue ocean approach to business, right? You can create this like new world and be the first person to dominate it. And that is what they are, they need to continue to try to create those new blue oceans, those new vast universes, untapped wildernesses to go explore. And there might not actually be any left.

And so I guess my question is, what does this mean about the market? Like if the market is driving this, right? We've been talking about all the bad decisions that the tech companies are making, that are forcing them to turn a large language model from, hey, a kind of cool niche product that maybe would be kind of neat, like, into the wave of the future. Everything's going to be different,

where it's going to destroy the entire economy and we're going to rebuild around large language models. They have to make that argument because the market is demanding it of them. So what the fuck argument because the market is demanding it of them. So what the fuck happened to the market, where the market is demanding this, right? Because we are describing the dysfunctions of tech as being a symptom of the market. So what the fuck is up with the market?

So the market's addicted to growth. And the rot economy is much larger than tech. I just know my tech much better than I know wider economics. That's more fun to talk about too. Yeah. I also believe that trying not to... What's Dunning-Kerr-Kerr, the whole... I'm not gonna...

But nevertheless, there is a problem across the whole stock market where it's addicted to growth. When everything is always growing at all times, why would you not invest in the thing going up? Because everything, it's always gone up. You see it in games. Companies getting way overfunded, hiring, firing, hiring, firing. Journalism, a great example.

Yeah. People taking over journalism. Oh, well we'll just shove more money in it and we'll get so much more money out. Never works, but man, do they feel good about it. Yeah. And what it is, is that the markets are just disconnected from value. There has been an echo. I don't really know when it's died value. There has been an economic shift

away from sustainability. And in my case, it's a good fire is one that's sustainable, it keeps you warm. What they are doing now is throwing bits of furniture into the fire. They are throwing gasoline on it, getting burned and like, ah, they must make more warm. And the problem is companies that are sustainable, that are reliably growing or perhaps they're plateauing, but they are still making more money than they spend. They're happy employees, happy customers. Those aren't sexy to the markets. The markets must have growth. I'm not sure what started it. I really am not. Yeah. But it's everywhere. It's just more pronounced in tech because of what you mentioned, the Blue Ocean approach.

When you crave growth at all times, this tech is a sexy little beast. Ooh my, everything's growing all the time. You've got these little companies that grew with three people that are worth, during the Metaverse time. It was one that got funded, I think, like $100 million. They didn't have a product.

They just formed the company. I have no idea how they did it. Well, and every investor wants you got in early. You know, you were the first person in, there was this great big thing. Don't you want to hop on board? I mean, I hear this sometimes. People were telling me 10 years ago,

you should make something for VR, VR headsets are the next big thing. And I was like, I actually don't think that they are, but you know, that's what I was, that's what I was told. And I find it very interesting that, you know, we're having this debate constantly about like, is the economy good or bad, right? That's like the core issue for Biden right now.

Like people feel that it's bad, but the numbers are actually good, unemployment is down, you know, interest rates, blah, blah, blah. There's this whole thing. And one of the things you can point to is the stock market is doing very well. It's done extremely well since 2020. And then before 2020,

it was doing extremely well since the housing crash. It's done extremely well since 2020. And then before 2020, it was doing extremely well since the housing crash. It's been like this sustained upward growth. And you know, I've got, I have a retirement fund, you know? Like, hey, that's great for me. Look at the nice big number, right? But it puts this worry in the back of my mind. Well, it seems like the one thing

that the people who run the economy know how to do, or that the system is optimized to do, is to produce good stock market returns. That no matter what's happening, no matter what the fundamentals are, we can get the stock price to go up. But that might be by these raw economy methods, that it is by squeezing more juice out of the thing

while destroying it, like short term, short term, short term, shoring everything up, sticking two by fours underneath all the sagging ceilings. Yes. And that that is what's really going on, which would imply that there's perhaps a reckoning coming. I'm scared of that. Yeah. I am definitely scared of what's happening in tech, because once...

The overall thing you're talking about, I believe, like, why people feel bad when economy good, it's because things are more expensive. The cost of living is high everywhere. Rents have gone up, they rarely go down. Yeah. Wages are stagnant. It is harder to get a mortgage than it's ever been. Yeah. It's so difficult. Mortgage rates are high too. The basic costs of living one's life are harder and higher and nastier in a very tangible way. Yeah, there are articles saying economy is good. Yeah, we should all celebrate with what fucking money.

Yeah. Where am I? What am I celebrating? I'm I'm not complaining myself. Yeah, I'm lucky. I really am. Yeah. But also health care costs in this country as well. I grew up my father worked for the NHS. Like I have great.

The fact we don't have Medicare for all in this country is completely fucking insane. Insane. But also, that's why people feel bad. It's not about a CNN article about the economy being good. It's their life kind of feels shitty. And when they say this feels bad, everyone's reaction is, eh, eh. Things like overdraft fees, that's a good thing they can do. But there are more things. Credit scores should be illegal. As a concept, they're evil.

Made by evil people. Evil, evil, evil stuff. That's a great example of something. Make mortgages easier to get. That's a way of fixing things. But to the wider economy, yes, I'm scared of a reckoning and I'm especially scared of one in tech because when the markets realize that AI is not the next best thing, or the next biggest thing even, they're then going to say, so what is?

What do you got for me? Mark Zuckerberg is going to freak out? I don't know. I actually don't know what happens. I just had a Tom Doughton from the Wall Street Journal on the podcast and I asked him, what happens if all of this compute space or the GPUs get bought for AI and the demand's not there. Yeah. And he said, I don't know. Yeah. And I don't know either. And that's actually the scariest thing. That's a little bit like the housing bubble or something

like that. We bought a lot of something that we thought was going to be worth a lot of money and it is not anymore. And that's exactly why it was that. And it reminded me of the hiring in 2021 when all the tech companies hired tens of thousands of people they didn't need just to hoard them. But the biggest thing is there's a way of repurposing that stuff, but there isn't much. You can't just throw GPUs in something you can, but what's it going to be? And there's no other next big thing sitting around. If you think about it, there have been many directions we could have gone in the past.

You had Airbnb, you had Stripe, you had Uber, you had all of these different things that you could do. There's really none of that right now. There's some exciting companies, but there's not like a next big thing in process. And if there isn't, there's going to be a terrible downturn against tech. Look at the startup market now. Worst funding environment I've probably seen in my career. 15 years. And that is more of a bellwether than anything else. Like if it's a frothy investment market and the

public markets are a bit scared, okay that just means the public markets, the public companies aren't innovating. It's weird that it's like this. The startup market should tell everyone what it is because venture capitalists aren't spending their money because they have no easy wins. They know something's up. But all of the public companies and some of the private ones have gotten themselves tied up in a tizzy following a market. I think they all kind of realize this bullshit.

And when that falls apart, it's going to... I don't think one of the big tech companies will die, but I think you're gonna see a massive collapse in one of them. I think Metta is probably the biggest one, especially because you can't kick Zoc out. I mean, it's quite a prediction. If that happens, like, I mean, huge amount of the stock market as a whole,

like, in the U.S., is the tech industry. And that's, like, people's pension funds, it's their retirement accounts, a whole, like in the US, is the tech industry. And that's like people's pension funds, it's their retirement accounts, it's like that's a broader economic reckoning. Again, I mean, not making predictions here, but you know, 2008, that was housing, it rippled everywhere else, you know, there's like potentially a similar ripple here. There is. And I don't know if it'll be quite as ruinous because it will bring the stock market down, but the stock market down, but stock market will find a way to survive. I'm more worried about tech in general.

I'm more worried about society's view of tech. The thing that the tech industry, and I try and tell as many people as I can, tech people, especially the ones in power, so goddamn smug, they're so sure. Everyone in society loves tech. I don't think that's the case. I think things like Google's AI search are turning people against tech.

Yeah. And tech may say, who gives a shit about these people? I know they make us money, but they're little bastards. I don't care. They use my things. I make money off of them. Except when regulation comes to tech, they're going to really be in trouble. Yeah. On top of that, private investment in tech is going to go down even further. Yeah. No one's going to, people already don't trust these companies.

If they start collapsing, I mean, do you, like one of these big products is going to just, Facebook is probably it. It's the first one that's really collapsing in user numbers. If they're moving their numbers around to make it sound better, something's wrong. If Facebook falls apart or if Instagram goes into the red, that's when you're gonna see these companies freak out or maybe it's just gonna be in three quarters when the markets realize that LLMs are not the future

and then all of Microsoft, Meta, Amazon, they're all gonna get ripped through. And I would say sometimes people said, oh, wouldn't that be funny? Not for the people working at the companies who are victims of circumstance. Yeah. Well, especially after these companies have already destroyed so many other industries, like journalism, media, et cetera, have been, you know, just either destroyed or the industries

destroyed themselves trying to chase the tech companies as has happened in Hollywood, for example. It's pretty bleak. Is there a chance just as a way to bring us in for a landing? Or is there a chance that you're that you're wrong and things will go along? We'll continue to go along fine. Or I mean, do you have any positive view of the future? So, yes. So I don't think it's going to be a full scale apocalypse

where everything goes to zero. I do think that when this AI thing falls apart and the markets start getting a little mean, I think Sundar Pichai gets fired. I think he gets rolled. I'm surprised he didn't over Google AI search. I think they will have to start making things a little bit better. They could find a next big thing, sure. But at some when the because in the rock-clone bubble piece I wrote traffic to most of the major platforms other than like reddit LinkedIn and tik-tok is going down Yeah, these companies are losing users. There are less people getting online. I Do believe that there could be a shift where they go. Oh god, we need to fix our products

We need to be a bit more sustainable I believe we kind of heard a little bit of it in 2022, the year of efficiency and so on and so forth. I think that there will be that. The pushes things. I also think the EU is going to aggressively move against generative AI in a way that makes them move away from it anyway. Yeah.

I think that you will see these stocks get chopped up a bit and things will realign. But the future tech industry will have to be more sustainable. It's not about optimism, it's just the growth at all costs economy will always exist to an extent, but tech will have to realign because so much money is being wasted right now. Billions is being burnt and I just don't think the companies, personally talking to the startups I know, the companies that are getting funding right now are the ones that have this crazy idea, Adam, you know, Dunno, you heard about this. What if a company made more money than it spent?

These kids these days, they've got all sorts of ideas. I mean, hopefully, you know, there's a coming correction and hopefully there'll be enough people who see it coming and change course, ideally to blunt some of the impact. I will say one thing, please. The startups and the investor, the small scale seed investors I talk to, I talk to them all the time from a day job in PR.

They don't seem worried, but they also don't seem beguiled by large language models. They generally, everyone right now, the earliest stage people, they're talking about sustainability, they're talking about building things that will last and not growing too far, smaller teams. That gives me a shit ton of hope. That's great. When I talk to these people and it's usually like, and this is I'm aging myself at 38

here, but it's a lot of these Gen Z founders, they don't give a shit about big tech. They appreciate that they exist but their reaction to this thing is kind of like mine which is why are you acting like this? This is a bad company, this is a bad business. These LLMs are cool, but they're not doing the things you say. What if we built a business that people like that they paid for? And it's very pragmatic and it's cool. I actually, that gives me a lot of hope.

It's why I'm not full scale doomer about this. And honestly, I think big tech could use a punch in the face figuratively. I think they need to be taken down a few pegs. I think they need to realize that they've forgotten their customers and forgotten their purpose in life. I can't thank you enough for coming on here to talk to us about it. We'll have to have you back in a year or two to see whether or not your prediction... Because these are very strong predictions.

And I want to see how accurate they are. Because you are validating my own emotional truth, and you are making it very definitive. Well, if I'm wrong, I'll be happy to explain why. That sounds great. I love that attitude as well. Ed, thank you so much for being on the show. Thank you for having me. Oh, where can people find your work?

So you can find me at WhereIsYourEd.app, and that's my newsletter, and BetterOffline.com. That's my podcast, and I Heart Radio, and Cool Zone Media. Thank you so much for being here, Ed. Thanks to you. Well, thanks once again to Ed Zitron for coming on the show. I hope you loved that conversation as much as I did.

If you want to support the incredible conversations that we bring you week after week, once again, you can do so on Patreon. Head to patreon.com slash Adam Cahn over five bucks a month gets you every one of these episodes ad free and you can join our wonderful online community full of people just like you who love learning and complaining about how screwed up the world is.

For 15 bucks a month, I will read your name at the end of every podcast episode and put it in the credits of every single one of my video monologues. This week, I wanna thank Eden Welch, Patrick Flanagan, Kratone, Linslat, Andrew Empurafoy, Marcus Mitchell, Vanessa Russell, Josh Beol, and Mask When You Can Protect Your Community.

Thank you so much, Mask When You Can Protect Your Community for slipping that message in via a $15 a month donation. If you wanna slip a clever message in at the end of this podcast, once again, head to patreon.com slash Adam Conover to join all the wonderful people who make this show possible. Of course, I wanna thank my producers, Sam Roudman and Tony Wilson, everybody here at Head Gum who makes the show possible.

Head to adamconover.net for my tickets and tour dates. Thank you so much for listening, and we'll see you next time on Factually.