Factually! with Adam Conover - The AI Hoax is Destroying America with Ed Zitron

Episode Date: February 26, 2025

The tech oligarchs who once pulled the strings from behind the scenes are now running the show right out in the open. Professional idiot Elon Musk is taking a sledgehammer to the federal government, and Silicon Valley’s sleaziest slimeballs are warming themselves in Trump’s fetid lap. Obviously, this is a disaster. To break down just how bad it is—and what it means for the future of our already struggling country—Adam sits down with friend of the show Ed Zitron, journalist and sharp-tongued tech critic behind betteroffline.com and wheresyoured.atSUPPORT 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. for a limited time. 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 OK. I don't know anything. Hey there, welcome to Factually. I'm Adam Conover.

Thank you for joining me on the show again. You know, it turns out that the tech industry is eating the world even faster than we thought it would. I've been a critic of big tech for a long time. I've criticized their monopolistic grip over huge parts of our economy, the way they've destroyed entire industries for their own gain, their willingness to push insane products on the public, whether crypto or AI, instead of the actual useful innovations they used to bring us.

And the way that billionaires have so much power in our economy, they actually threaten our democracy. I and many others have been making these points for years, and you know what? For a little bit, it looked like these arguments were finally starting to have purchase. We started to see pushback against big tech coming from places like the FTC, from Congress, even from the Biden administration.

But you know what happened? The super powerful billionaires who run the industry did not take that well. As soon as the least little bit of pushback started, they said, you know what? Fuck this. We just need to take over the country. And they did it in a matter of months. At Donald Trump's inauguration, we saw the tech overlords sitting behind him, almost puppeting him from behind the scenes. Not even behind the scenes, in front of the scenes. They were literally fucking sitting there.

And now Elon Musk, the richest man in the world, is practically running the federal government, dismantling it for parts, selling them off and replacing it with bullshit like grok AI. The people who have been screaming about this the entire time turned out to be right. So right, more and worse right than we could have possibly imagined. Well, on the show today, we have one of the very best critics of the industry returning to the show to help us remove the veil of bullshit and marketing from this industry and expose the rot at its core that is destroying not just our economy, but our very government.

Now before we get into that, I want to remind you, if you want to support this show, head to patreon.com slash Adam Conover. Five bucks a month gets you every episode of this show ad free. We also have a lot of other wonderful community features We'd love to have you. If you would like to come see me do stand-up comedy, live, in person, not mediated by any tech monopoly But right in front of your fucking face. Come see the Nihilism pivot tour. I'm touring all across the country from March 6th through 8th I'll be in Burlington, Vermont. On March 22nd, I'll be in London. On March 26th, I'll be in Amsterdam and the Netherlands. Then April 3rd through 5th, I'll be in Providence, Rhode Island, April 17th, Vancouver, British Columbia, April 18th and 19th, Eugene, Oregon.

After that, Charleston, South Carolina, Oklahoma City, Tulsa, Spokane, and Tacoma. Head to AdamConover.net for all those tickets and tour dates. And now let's get to this week's episode. The last time we had this man on the show, it was one of our biggest episodes of the year in which he talked about how AI is marketing, how it's a bubble that is soon to burst. But now it seems that that bubble has embedded itself

within our very government with potentially disastrous results. So here today to talk us through it is Ed Zitron. He's the host of the Better Offline podcast and the essential newsletter, Where's Your Ed At? Please welcome Ed Zitron. He's the host of the Better Offline podcast and the essential newsletter, Where's Your Ed At? Please welcome Ed Zitron. Ed, thanks so much for being on the show, man. Thank you for having me.

Yeah, it's great to have you back. Last time you were on, you talked about how AI is dying, how in your view, the AI industry is like running towards a cliff that is puffery, it's marketing, and that the whole thing is going to collapse and it's a bubble. Now, the people who created the bubble seem to be running the federal government. It seems like we're in a slightly different place vis-a-vis the tech industry than we were last time we spoke. How do you see it?

Well, I wouldn't say the people, the legion of beavers inside the government are connected to the people who inflated the bubble. That's more Sam Altman, Daria Amadea of Anthropic, and of course Sachin Della of Microsoft, Dhan Sundar Peshaw of Google, and so on and so forth. The people inside the government now are like the 19-22 year old gropers that Elon Musk could find with comp side degrees or bits of it to go and break things, I assume? To walk up to the... I really didn't think there was a computer that just had, like, Federal Reserve money taped onto it and you could just go and play around with it. That's the most worrying thing about all of this

for me. That there's just a room. I always thought there'd be, like, layers of security, and you couldn't just have, like, a Gen Z guy with the broccoli hair going in there and just messing around with the computers. I think the thing that also scares me is, we know there's a bunch of layoffs, but we don't know what else they've done other than that there's been an announcement that they're pushing code to the Federal Reserve. That doesn't sound good to me. That doesn't sound, especially based on how Elon Musk does not seem to understand anything

technical, it seems? Well, he's saying stuff on Twitter that makes it very clear that he either doesn't understand the numbers that he's looking at, or he is specifically trying to create a false impression of fraud or whatever. He'll pull out two statistics and be like, hey, this is massive fraud. And people are like, if you have a basic understanding of statistics, you would know that there's a million different explanations for these numbers that you're looking at Yeah But yeah, I mean well apart from the you know 19 year olds running amok which is worrying enough

The way it looks to me is that you know the the the billionaires who run the tech industry had this all this massive political power They've had it for years. I've been talking about it for years Yeah, then for the first time we actually, as a, as a country started pushing back on it a little bit. If you look at the stuff that was some of the good things that were happening under the Biden administration, Lena cons work, et cetera, ending mergers. And, and that sort of thing, or at least fighting back again for the first time ever, it seems like the tech industry said, well, hey, fuck that,

we just need to run the country. And then they took it over within six months. But the thing is, the tech industry isn't even running the country. It's just Elon. It really is. Google, Meta, we'll get to Meta, I imagine, in a bit. Microsoft, they're not, such an Adela isn't, they're not touching any of this. They're kind of off to the side being like, we all gave a million dollars to Trump's inauguration committee. Great, I guess? If you look at the way Mark Andreessen is talking, he's talking as though he's intimately

connected with the government and part of a political revolution. And this is a guy who's just like, oh, I'll just take the reins of power, why not? I have enough money to do so. And he has influence, don't get me wrong, but I don't know how much they're actually involved with this stuff. And if you really look at it, and Wyatt has done incredible journalism about this, it And yes, influence, don't get me wrong, but I don't know how much they're actually involved with this stuff, and if you really look at it, and Wyatt has done incredible journalism about this, it seems like it's very Elon-centric. And they're trying to do some legal bullshit with us, like, oh, actually, Elon does not

work here, there's another guy called Elon here, we don't know who Elon Musk is. But nevertheless, the tech industry, and in Andreessen's case, big Curtis Yavin fan reads apparently comprehensive biographies of Hitler, 2022, Andreessen said that. Always good to see this, but they're part of the- The problem isn't reading a biography of Hitler. The reason, the problem is why did you read the biography? What was your curiosity about Hitler? It's two books and it's why do they need to be comprehensive?

What do you curiosity about Hitler? It's two books and it's why do they need to be comprehensive? What do you need to know? Like paraphrase succession is like, did you miss some of the Easter eggs? But nevertheless, it's more that there are these kind of nebulous advisory councils, like David Sacks is on some AI council, these people who meet and they have lunch a few times and then they talk and they say AI is going to be big now, America should do lots of that. And then nothing appears to happen. And it is very scary having these people anywhere involved, and who knows what they may do next, because every week something new happens that is just so strange.

We are in the strangest part of history, and possibly one of the worst, who knows, we'll find out. But it isn't obvious how much Andreessen is doing because it isn't obvious how much Trump is listening, or how their Trump is. So it's this very weird thing where they're very immediate and they're all doing their touchdown dances, reading the 14 words as they do so. But it isn't obvious what it is they're actually doing. And maybe this is all they want. They just want to be the ultra-brords. The brunch lord Hitlers. I don't know. But they want to just do this thing where they can swan around the Mar-a-Lago and then do nothing? Because what, it is an obvious what Andreessen wants other than something

clearly right wing, something clearly bad. But it's when you actually look at the actual outputs and things they're saying, when? I mean, a lot of it is they wanna be left alone to do their shit, you know? Oh yeah. I mean, I was, you know, I went to an event a little while ago and I talked to some very normal VCs,

like venture capitalists who, you know, were involved in various tech investment funds. These are nice people. Right. And we were talking about politics. This was before the election. And we were talking about politics. This was before the election. And we were talking about Lena Kahn at the FTC. And she said, they said, oh, she went way too far.

She destroyed the M&A market, the mergers and acquisitions market. Now you can't start a company, right? And I looked at that and I said, well, you know, mergers destroyed my industry, like Hollywood. Like me and my friends are out of work as a result. So I think that we should be, you should be trying to stop some mergers. To them it was completely foreign.

I realized these people live in a world where the only thing that you do is you start a company and you sell it to Google or Amazon or one of the other big players. They can't even imagine living in a world in an economy where that isn't the case. But that was like their basic thinking, was we need to be able to do what we do

and anybody who tries to stop us is getting in the way. And again, those are normal people who maybe if I had talked to them for a couple hours, we could have come to some agreement. But if you're a guy like Andreessen, you're or any of these other monsters running around, think why should anybody ever tell me what to do? Like it was, a lot of them are just anybody ever tell me what to do? Like it was a lot

of them are just, you know, radical libertarians about it, right? Like let's just, let's put our guy in and then we'll be allowed to do whatever the fuck we want. And so what's funny with Andreessen in particular is he wrote this thing called software is eating the world 2011-2012. It was in the Wall Street Journal. It was this big, very influential piece. Software is eating the world. And you think with that, it'd be like, wow, so software is going to change everything. You actually read it, he specifically says that software companies should not be valued like regular companies. That they should be allowed to proliferate within industries. He refers to Pixar as a software company.

This is how people like Andreessen view the world, which is that everything must be penetrated by growth, and software is the best vehicle for growth ever. Andreessen and his ilk, and Andreessen is exceedingly inspirational to the venture capital people, even if they don't agree with his very obvious politics. They all want growth and they all believe that the best world is one where they can incubate money thingies that they can sell to other people or take public to get more money, and that that is how business works. It isn't so much about selling a service to a person that they like, or selling a thing in a store. It's about how can I make this expand and kind of touch every part of an industry and then get rid of it? Because I don't really care about the

value, I care about the penetration, I care about getting this out there and touching as many parts of industry as possible. Software is eating the world was extremely influential. But the truth is what you're describing with these venture capitalists is the world of growth. It is the growth obsessed people of Silicon Valley that they believe that software's job is to proliferate. It is not to solve. It is not to make things work for a consumer, be an enterprise consumer or a regular one. It's to dominate an industry with a product that is unavoidable. And that is what has darkened Silicon Valley. And that is what has ruined the tech industry as it stands.

And that seems to be what's infecting the federal government, because if you look at what Elon is doing, or at least talking about doing, or actively doing right now, is A, laying off massive numbers of people, which is part of the growth mindset. Hey, you can grow by shrinking, basically. You cut margins and you make more money that way. Number go up. Number go up. But also, they are talking about inserting AI

into every part of the government, including in like the FAA, like air traffic controllers, the money system. And to me, this looks like exactly what you're talking about, that they're taking a form of software that honestly has no obvious benefits, at least at this point in time, AI generally. You and I both agree that its utility is vastly overhyped.

And they're just saying, let's insert it into these places so that no one can get it out. So we just force it upon the American people. And I think that there's one abstraction higher that applies to everything, which is what growth never does is maintenance. Growth does not see value in anything other than what propagates more growth. So layoffs are a natural choice, right? Well, these people aren't performing at the level of this benchmark I set. And what do they do? Oh, they do a class with some people. Fuck that. We don't need that bullshit because these people also don't interact

with services or businesses or customers. They're one step above everything. Our economy is dominated by people that don't do real work, who don't understand real problems, and thus can't solve them and really don't see why they would in the first place. Inherently selfish, but also inherently cancerous. So in the markets you have services being sold to companies not based on whether they're good or not, but whether they're cheap enough to sell to a boss that isn't going to use them. Because you're buying 100,000 seats, that guy doesn't fucking care, he cares what's cheap and what makes him feel good and warm inside. Same deal with the federal government

and doggy or whatever the fuck I meant to call this bullshit. It's... They go in there and they say, well we have too many people, we need to save money. Delete, delete. That's how that works. That's how business works. They don't care whether the government works. They don't know whether this functions. They don't care that Yellowstone has like three people full time now. They don't think of these things because they think of everything in numbers because they

do not participate in reality. These people are, Elon Musk is divorced in many ways. He's definitely divorced in reality. He has more money than anyone ever should to the power of the bazillion. And on top of that, when was the last time he interacted with the civil service other than to complain about it? He never does. And these people working for him, they're probably being paid out the nose so that they can be thrown in jail eventually, I assume. I assume that they're going to be the sacrifices. But I think that all of this is symbolic of a bigger, bigger problem which is the economy is divorced from production entirely. Right now, the people

running the economy do not do real work, do not face real problems. So AI, generative AI, seems magical to them. All they do is go, they read emails, they send emails, they ignore emails, and they go to lunch. Right. Wow. A thing that can summarize the emails I don't want to read? Holy shit. I bet this thing will become God in two years. And it's like, yeah, the only problem in the world is emails. Exactly. And this thing can take care of some of your emails for you. Okay, great. That's all we need. And that and seriously, it can do large language models can do more. Sure. But the whole AI bubble in the markets, that's all it is. It's like, hey, can we staple fucking AI

to the side of it? Great. That will grow somehow. Is it working? No. Is it actually creating much revenue? No. But you know what? Everyone feels like they're part of the future, right. Is it working? No. Is it actually creating much revenue? No. But you know what, everyone feels like they're part of the future, right? Is it the future? No. But, I mean, I hope this doesn't- it's really bad, man. It's so bad. ALICE Well, let's talk about AI since the last time we spoke. You've been saying it's a bubble for, you know, the last year at least. The bubble's continued to grow. There are all these Super Bowl ads about AI, of course, which I saw you write some very scathing Blue Sky tweets about.

What is the status of these AI companies currently? So it's all very weird. So DeepSeek came along. We can get to DeepSeek in a bit. In the round, these companies, so it's been about seven months since I was last here, since then, Microsoft has announced that they have, wow, $13 billion of annual recurring revenue from AI. Now, just to be clear, Microsoft has no line item on their earnings about AI. So they've just taken every little bit to say $13 billion. So, what, $2.6, $2.7 billion a quarter? Not profit?

That's all that they can do? That's all that the biggest monopolist in software can do? That's all they can get? OpenAI, they hit about $4 billion of revenue last year. They burned $5 billion. $5 billion. That's after revenue, and that's not including stock-based compensation. Anthropic, Dario Amadei, history's greatest liar.

ALICE See his name again. Jason Cosper Dario Amadei. Jason Cosper Dario Amadei. Jason Cosper Yeah, Wario Amadei. I'm putting that one on the record. But nevertheless, Anthropic, compared to OpenAI, they made like 2.7 billion last year. They lost 5.6 billion. This is the future. And then you look at this and you go, well, Ed, the numbers are increasing, right? Well, surely they'd have a new product. Well, you'll be surprised to hear they haven't really got one. So they released something called reasoning models, which is just the models

check their own work and they think they compute longer. These things don't think, they think longer and everyone said, wow, oh, one's come out. Reasoning's come to large language models. What does this mean? And the answer is bugger all. There are scientific things, as ever with these large language models. There are use cases down here. And then there's all of this burn and all of this stealing and damaging our power grid and burning our environment so that we can generate bustier Garfields, I guess. But nevertheless, they do have some products. Adam, you're going to be shocked to hear how good these are. So imagine if you were in front of your computer and you think, I want to search TripAdvisor. Now you could go, I'm just going to type into Trip

Advisor. Why would you, you dink and poop, you imbecile. You could use Operator by OpenAI. Operator, what if instead of doing that, sometimes Operator could successfully search TripAdvisor for you in several minutes? And what if it was not impressed? It's also extremely expensive. And what's great is Operator is meant to be able to run your computer for you. It's meant to be able to do these distinct things. And it does so by taking screenshots as it scrolls down and going, what do I do what do I do here? What do I do here? And this is all part of a big thing called agents. They're saying AI agents are here. AI agents, what does that mean? Stop asking fucking questions. Stop it. No, don't ask me.

AI agents are meant to be, and they sell, much like everything in generative AI, based on the kind of ephemera around it. like helping people fill in the gaps. Agents are meant to be these autonomous things that go out and do things for you. In the case of OpenAI Operator and indeed Anthropix version, I think theirs is just called computer use, it fucks up like 90% of the time. It doesn't work because they're trying to use large language models to do distinct actions. Now you may think, well, that can't be the only product. And you're right, Adam. And it's called deep research. Now this is going

to blow you away. What if instead of looking something up, you could have OpenAI look it up and then they could give you 5,000 near incomprehensible words that cite SEO bait articles and Reddit posts and Hacker News. What if you could get a big useless report with hallucinations in it? And I know you're already excited, but what if it was also only available on the $200 a month service, coming to chat GPT plus, and was also insanely compute expensive to the point that they have to severely limit it. And we are two years into this bullshit, and this is the best they've got. Like what you would get for $10 from a guy in the global south on Fiverr, but worse, and it took longer.

And people in the press are just vacuuming this shit up. They're like, wow, look at this. It can do research. Yeah, it can do bad research that barely makes sense. And I must be clear, we're two years into this, there's nothing! They haven't got shit, and they're still unprofitable. And what's crazier is OpenAI's $200 a month pro service gives you unlimited access to varying asses that can shit in different volume. It's still unprofitable! The $200 a month subscription is unprofitable. They are not making money. Jesus Christ. It's so cool that this is the future. But on top of this, they keep being... Dario Amadei of

Anthropic, he said by 2027, or just after, it was like soon after 2027, we'll have AGI. That's also around the time 2027, the information reports that Anthropic will actually stop burning cash. They will also, apparently, they're projecting internally that they will make $12 to $30 billion in revenue. Now, I must be clear how much bullshit this is. This is all nonsense. Anthropic last year made like $900 million. These companies are just saying stuff, and I'm disgusted with the tech press for just walking around being like, yeah, makes sense to me. It doesn't make sense to anyone. This stuff's completely insane. If I went to the bank and I said, I have lost five billion dollars, they would not be like, don't worry, we've got some investors who

are... Jason Vale Well operating at a loss is a thing plenty of businesses do. You know, Amazon had no profit for like 10 years. Will Barron Nothing close, not even close to this. Not even slightly close to this. There is nothing. The worst year of Uber, I think, was 6.2 billion. That year was 2020 when Uber's business stopped working.

Now Uber burns cash by the boatload. But at least there's a service. You can go, I know why I'm using Uber. To this day. To this day. No one can really give a clean definition of chat GPT. Like what is this? No one can really, I've yet to meet someone who's fanatical about it because of the use cases. I've met plenty of people who are fanatical because they have a weird attachment

to it. They're like, I must protect, I must protect Sam Alton. I must protect the large language models or else I will die. And it's always the same kind of vague, like there are tons of use cases and they never respond when you ask for one. They're like science. Science. Jason Cosper I mean, the main one I hear is programming. Programmers seem to use it as a, you know, tool to increase the amount of code that they're able to write. And I've seen, I was just looking yesterday at some code editor, like a piece of editing, a literal text editor that like integrated AI in to like help you.

It's gonna predict what the next function you're gonna write is. And I could imagine that being a value add. And that's the really annoying thing. So large language models as an idea are not inherently bad. There is had Sam Alman, and this is a failure of the tech press and of Silicon Valley.

Had the tech press gone like, wait, wait, wait, wait, what is this? Why is this, why do we have to agree that this is big? Large language models as an idea, not a bad thing. Indeed there are on device, large language models that are actually doing cool stuff. I too know people, engineers, who like the coding editors. Nevertheless, coding editor is not the same thing as $1 trillion industry. There was a big report from Goldman Sachs that came out actually not long after I was on, where they were saying it doesn't solve any... Jim Cavello, lead equities analyst over at Goldman, said that everyone's saying

that generative AI is going to be this $1 trillion industry, but what $1 trillion problems does it solve? And it doesn't, and it still doesn't. What you're describing there is useful. It's useful. Is it worth burning our planet? Is it worth stealing from everyone? Is it worth every single transaction being unprofitable? No.

And indeed, there may be on-device models that can do exactly that and do that coding stuff And that might be cool. That might be fine. That might be fun. That is not a trillion dollar industry That is not a GI which is artificial general intelligence. It is not conscious and it is certainly not mass market Folks this week's episode is brought to you by Squarespace. You know, I used to design my own website. You know, I used to learn HTML, CSS. I could even write a little bit of PHP. All right. That was fun.

That was in the old days, the hobbyist internet when people were just logging on via Netscape and looking at your flat HTML file. Well, guess what? The internet is a lot more complicated these days and it is even more crucial to put your best foot forward. Whether you're running a business, growing a brand or just sharing what you love. The internet is a lot more complicated these days, and it is even more crucial to put your best foot forward, whether you're running a business, growing a brand, or just sharing what you love.

And you know what? You're too busy running that business to learn a bunch of brand-fangled new code. I definitely am. I'm too busy doing comedy to learn to code, okay? Shut up, don't tell me to learn to code. Instead, I'm gonna tell you to go use Squarespace. That's where Squarespace comes in.

That's right, you can use Squarespace to build a stunning website effortlessly. You can showcase your work, connect with your audience, and sell anything, products, content, even your time, all in one place, and completely on your terms. Squarespace lets you seamlessly integrate your social media posts and clips right into your site so your audience can find everything in one spot. Selling products? No problem! You can sync your catalog with Instagram, Facebook, YouTube, and Google in just a few clicks. And of course, Squarespace makes it incredibly easy to sell, whether it's physical products, memberships, exclusive blog content, videos, or tutorials.

Whatever you're offering, they've got you covered. So if you're ready to take your company or personal brand to the next level, head to Squarespace.com for a free trial. And when you're ready to launch, go to Squarespace.com slash factually to save 10% off your first purchase of a website or domain. That's Squarespace.com slash factually. This week's episode is brought to you by Alma. You know, isn't it wild how the things we often turn to for comfort end up making us

feel worse? I mean, I've definitely caught myself doom scrolling, getting stuck in endless social media loops that I think are making me feel better, but at the end of the day, I realize they're just leaving me feeling more disconnected from the world around me. In a time when we are stretched so thin between our physical and digital lives, you know what's really important? Building real human connection with actual people who can help. And that is why, if you are seeking some help in dark times, I recommend taking a look at Alma.

They're not one of these digital therapy services, okay? They make it easy to connect with an experienced therapist, a real person who can listen, understand, and support you through whatever challenges you're facing. I cannot tell you how important it has been for me to find my therapist who understands me, how much that actually helped me on my journey of mental health, and you can find your person too.

With Alma, you can browse their online directory and filter by what matters most to you, and then you can book a free 15-minute consultation with a therapist you are interested in. Unlike other online therapy platforms that just match you with whoever's available via some kind of chat service, that you're playing chat roulette with your mental health, Alma lets you choose someone you truly connect with because the right fit makes all the difference. With their help, you can start seeing real improvements

in your mental health. Better with people, better with Alma. So visit helloalma.com slash factually to get started and schedule a free consultation today. That's helloalma.com slash factually to get started and schedule a free consultation today. That's helloalma.com slash factually. It's a feature of a single software product. It's when like when Photoshop comes out with a new way to make an image quicker, right? And I look at it and go, oh, that's neat. That would be useful if you're a Photoshop guy.

It's cloud-based software. And that has been a healthy industry. Indeed, large language models might be a $15 billion industry if they had just not done this. Because you can look back, because I hate myself, I've looked back at all of the lawsuits

between Elon Musk and Sam Altman. From the early days, Sam Altman wanted this open AI, which was founded as a charity, who investigate AGI and to build responsible AI in literal response to DeepMind, to Google's DeepMind back in 2015. The whole time, Ilya Suitskeva and Greg Brockman, who were brought on to kind of run this thing with them, they're sending identical emails to Sam Altman and Elon Musk being like, Hey, why do you want to be CEO so bad? Why do you keep want to run this?

You seem to want unilateral control. What's that about? And Sam Altman the whole time was pushing for a for-profit company. Sam Altman wanted an initial coin offering. He wanted to raise money using crypto for open AI. Sam Altman from the beginning has been a little crooked. But one thing I'll tell you is Sam Altman was never interested in this being like just a fun software industry, which like a 15, 20 billion dollar industry. He wanted power and he wanted money. And Elon Musk fucking sucks.

But this lawsuit does pull out some interesting details and shows the narrative that it is a narrative, it is Elon Musk's narrative, that Sam Altman's whole plan with this was just to make a bunch of money and make himself rich. And he succeeded using whatever con he could. And there is no greater con, I think, than taking a relatively boring piece of cloud software. A really interesting, it's interesting to me, but like, it's cool. I mean, large language models are cool.

I continue to play around with them every once in a while and go, Hey, neat. But that's the thing. They're neat. Yeah, they're neat. They are not the future. Yeah. And had Altman done that, he wouldn't have been able to raise any money. No one would have given a shit. Everyone would have been like, why am I giving money? Oh, what, you want to make something useful? You want to make something? Oh, you want to do a large language model where people could do stuff with it? Fuck you, Sam. Where's God? Make God come out of the computer right now, you little bitch.

But he was already doing the lying there. But even from the beginning, they were pulling the AGI thing. It's just, it's frustrating because one of the classic gotchas people say is, ah, you know, there are actually people who use Chad GPT. There are people who use it as 300 million weekly users. First of all, those numbers are bullshit. It's 350 million weekly users. First of all, those numbers are bullshit. It's 350

million weekly now, they claim. A few things. Number one, 350 million weekly users, but similar web data I've got given says they've never had more than 150 million weekly unique visitors to chatgpt.com. You're telling me there's another 200 million just in the ether? Fuck that. It's complete nonsense. Putting that aside, weekly users just means they used it once in a weekly period. So it's like someone went on chat GPT and typed in like Garfield with Garnt. That is not... this isn't... That's literally what I do. I log into a large language model about once a week and I just ask it some dumb stuff and

then I stop. And that's nothing wrong with that. Other than the costs and all of that. That's not great. No, I know. I hate water. But eventually though, there are going to be on-device versions. And people might say, that's another gotcha.

No, it's not. If anything, it's really bad for Sam Altman, because as it stands, chat GPT as a service, and you've got 01, the reasoning mode, you've got 03, the one you can't touch that's too expensive, like $1000 to run for 10 minutes. I fucking love this shit so much. You've got Operator, you've got Deep Research. These aren't products. They are products, but they're not like industry leading. Deep Research has already been copied by Perplexity, which is barely a company. It already got copied by Grok. Fucking Grok's copying you. You're not a real company. You got, you got met by Grox.

Oh, I need a shower. But it's, it's all of this hubbub and all of this nonsense about, oh, this is automation. It doesn't even automate anything. It doesn't change our lives meaningfully other than making them more annoying and worse and filling search engines full of slop. And the irony is with deep research, by the way. So, um, a guy called Max works over at Buzzfeed theorized this, and I kind of stand by it, is the reason it can't, because it searches

the web, the reason it can't find really good primary sources is because robots.txt, that's on every website, is blocking all the AI. So it's just like, aha! Oh dear, you don't like that, do you? I hope that's what's happening, but I think it's also just as likely that because these not models don't know anything, because they have no knowledge, that they just don't know what a good source looks like. So go fuck it, what would support my point?

Forum? Forum? Fuck, who gives a shit? But it's all commoditized. Everything's commoditized. They don't have anything else, and they're nowhere near anything else. Two years in, and what? We've got like a bad search engine, and a kind of word calculator.

Sexy auto-correct? I mean, every product is the same, Anthropics Claude is the same as OpenAI's 4.0, it's the same as Grok 3.0. They all perform well on various benchmark tests that are all rigged so that large language models can do them, because large language models are not good at regular tests. They can do tests that have a defined answer, but they cannot really... there's no free form jazz for these things.

Yeah, they can do good at tests that have a defined answer because the answer is somewhere in the massive corpus of text that they have ingested, and occasionally they can pull it out at a high enough success rate, as could Google or something along those lines. But that doesn't, that's not reasoning. That's like, yeah, the answers to the state bar exam were somewhere in all of the world's texts that you sucked up. Congratulations. And also the answers from the bar exam that you fed into the training data as well. You can prep these things for these tests and people will say, oh, you can't rig them.

Sure you can't. How would you prove that they didn't rig? It's just, it's all so silly. And on top of that really simple point, which is if this was so futuristic, why is everyone trying to convince you all the time that it's futuristic? Yeah. Do you need to just point to the thing like the first iPhone, you weren't like, why is everyone trying to convince you all the time that it's futuristic? Yeah. Should you just point to the thing like the first iPhone, you weren't like, why is this new?

Yeah. Where, like with the iPhone, you, they did the pitch, then your friend got one and you were like, Oh, and then you got one and you were like, Oh, visual voicemail. And then there, yeah, there were new things that happened all the time. There were like active things that were actively changing your life. You know, Oh my God, a map in my pocket. The amount of, I've talked about this before on my Netflix show, the amount my life changed by having a map in my

pocket was like massive. The people I know who are using ChatGP, I do know people who say they use it in their lives. They're using it as Google. That's what they use. They're like, oh yeah, I use ChatGDP instead of Google. Well, you had Google before. It doesn't sound like a massive change

in your life or in your workflow. Yeah, it's a different kind of answer, I think arguably worse. But like, what's the, let me ask you this, because there's a lot of, we're talking about these very consumer facing large language models, generative AI. Occasionally I'll see a headline about,

AI could help solve a difficult problem in, name a science, right? In chemical engineering and in physics or something like that, because this sort of reasoning, when used by a scientist, know, scientists, right, could like really, you know, create a power multiplier for some actual difficult problems. And, and so I always want to make sure, you know, checking my own biases here, right, versus the products I'm using, is it possible that there is, you know, looking at AI behind the scenes, that there are

larger applications like that. So I think you start first thing to start with is has it actually done that? Anytime I see a headline that's like AI could I'm like yeah but has it done that? Yeah. That's all because everything in generative AI and the AI bubble that you see is sold off of exactly what you just said which is AI could and people go so AI will? That's the answer. Now yeah there are science examples like Alpha Fold which is not generative AI, which is not, it's a different thing. That's cool. Hoping in the sciences.

There are, I think there was some sort of protein, there's a protein related large language model thing. Like there are scientific applications. Yeah. There's like bio med stuff. And again, never saying that large language models are completely and utterly useless. Any statement I've made suggesting that I was a bit off, but I'll tell you the truth is the problem is the term large language model, because even in the sciences, it's specialized models.

But you can't really say small language model because those are so big compared to the past ones. But the thing is, just because those use cases exist, doesn't mean it's a trillion dollar industry. Like that's the big disconnect here. And also that what you're talking about there is not what OpenAI is selling off of, it's not what Anthropic or Google or Microsoft or any of these companies are selling off of. Amazon's ruthless AI is nothing to do with the sciences.

They're not selling off of that. They're trying to sell into the enterprise and the enterprise is saying, so what can this do? And they go, that's a good fucking question. Would you like to pay us so you could find out for us? And weirdly enough, there are some companies like, sure, I don't know what this does. You don't know what this does, but everyone's telling me to put AI in it. So k-dunk. This is where it's going. This is what's happening. This is again, the economy run by people that don't do work, sold to by people that don't

do work making a machine to sell them. Because if it really was this big scientific breakthrough machine, they could sell it. Like that would be the thing to sell it on and there would probably be something in two years, right? Something major. I don't mean a study where this happened once. I mean an actual breakthrough. Oh, it's the early days. Shut

the fuck up. No, it's not the early days. All the King's horses and all the King's men have been putting every fucking penny into this. Every bit of media attention into this. And this is what we've got. This, this, this nonsense. Like I would, you know what? I would eat crow if the whole thing was like, oh, we've actually cured a disease. I'd be like, shit, that's actually really cool. I would never be against it. That sounds great. I'd love to cure diseases. They are not doing that. They are working on curing the Earth of its life. It's just, it's so frustrating as well, because there are so many parts of this, I'm gonna pee off, I'm like, there's so many parts of this that are just really cynical marketing.

And it's like Sam Altman doing a partnership with, oh god, Ariana Huffington for some health AI app. And then the whole thing, there's a thing in Time magazine about it, where it's like, oh yeah, and then in the future this thing could take your data and it could do this and this. And it's like, in the future, a bunch of shit it doesn't do now. And also this thing doesn't exist. Time magazine. Sold. Great stuff. Agent Force. Salesforce's powerful AI. It's just a chatbot connected to a database. We had that shit in 2015, if not earlier.

Everything you look at, all of the big products, the magical products they have, are just shit we had already that sucks. Wow, it's Google, except sometimes it's right, but it's also not. And you don't really know whether it's wrong or right, because it doesn't know anything. It doesn't know what correct is. It probabilistically determines what the right thing might be. And it's pretty good. Pretty good is not good enough. Or even close.

Yeah. Well, what's the, what is the revolutionary use case, even if it's better than pretty good? I keep seeing this ad where I think it's for agent force, where Matthew McConaughey is outside at a restaurant where it's raining and he's like, my travel agent didn't have AI, so I'm here at this bad restaurant and it's raining on him. And then Woody Harrelson's across the street,

he goes, hey, Matthew McConaughey, like we're here at this dry restaurant because we have AI come over here. And I don't even understand what AI in the fiction of the ad was meant to be doing. How did AI, isn't booking a restaurant reservation something that we've been doing for a hundred years? Like what, why do I need AI to do it? So, first of all, I love the fact that, like, these commercials are not made for human beings.

So, that one, it's like, my AI agent didn't book my restaurant, but you can do that on Google, it's just, you don't open Tapermate, it takes a few clicks. You could even do it Matthew McConaughey. Yeah. What are you doing? You're fucking Lincoln, I don't know. I do it on McConaughey. Don't you fucking Lincoln? I don't know. I do it on the Yelp app. But worse still, the Super Bowl one is Matthew McConaughey in an airport, Heathrow airport. It's very clearly not Heathrow. And he's like, I didn't book, my AI agent wasn't booked using

the powerful AI from Salesforce. And it's like, it does not fucking make sense. Agent Force is a not a customer facing thing. It is cloud software for routing conversations, but he says, Oh, and I didn't, I didn't, so I don't have a guide to where my gate is again, not a feature of an AI agent feature of an app. Right. And then it says at the end, Heathrow runs an agent for, so I actually went and looked this up, I can find no evidence that this is a modern integration. I could find a thing from 2023.

And it was basically just like they've increased chatbot usage. You've managed to create usage of the thing that you increase that by 400%. It's new. So that zero is like, like one times four, like does that mean 400 people use it? It just, these commercials are made by people that don't do work for people that buy things For experiences they think happen people go to an airport, right? Well, they take planes there. I think yeah I mean we of course fly private look Go there and get these filthy diseases with these little pigs annoying winking around with their gates I mean this this is their attempt to show us how AI Might have an effect on your daily life, but it's clearly the effect of the point of the goal.

That's what most commercials do. Let's show you how the product would actually affect your life. But I watched the ad and I say, I don't, I am not making the connection. Well, because there's nothing for you to use. Yeah. The AI agent in question in the gate expectations. Fucking Christ. That's what it's called. I fucking hate everything.

It's called gate expectations. I hate it so much. But what's great is like, I didn't use the powerful AI from Salesforce. It's like, you wouldn't use that. Yeah. But on top of that, even if I buy this concussed premise,

you get to the point you're like, okay, so it's about the Heathrow app, right? Why wouldn't you? I mean, that app is free. And also pretty much every airline has an app, and usually they guide you to the gate. Where is the AI in this equation? Because the gate guidance thing wouldn't be AI. That would just be a point-to-point thing. A GPS location, I assume that they have some kind of beacons or something, so... And if they don't, the thing just

won't work. Nevertheless, it's like, the thing that guides you, the features he's discussing, are not AI at all. Indeed, he's clearly missing his flight, which is just a Heathrow experience everyone needs. I don't know where AI would come into that. You'd just be like, Chatbot, hey Chatbot, I'm gonna miss my flight. It doesn't make any sense, but I think it just shows contempt. That's all I get from these commercials. It's not just, Oh, they're trying to hype something. It's just like a, you little pigs will buy anything. You little pieces of shit.

Everything you watch the chat GPT commercial. Did you see that? Oh yeah. That's the next one I was going to bring up. It's just this vague promise of it's an animation of technologies progress through time. Uh, well we used to have fire in caves, then we had a wheel, and then we had the steam engine and now we have AI.

So you should use it. 47 seconds in. That's how long it takes them to get to anything about the product. And the examples they give are, help me ask for a raise, generate me a business plan, summarize this article, two things in languages I don't speak, but that's all you've got is the fucking Super Bowl, man. You're not got anything more impressive than summarize an article or prepare a business plan two years into this bullshit.

And I have to like, they're just like, Oh my God, I can, I can have articles summarized. I can do the thing I did a year ago. I can do that feature has been a bit. Yeah, it just, it is contempt. And I'm sorry, where is the market for summarizing articles? Did we need articles to be summarized so badly? I don't know.

You can read the first paragraph and the last paragraph. This was always possible. Skimming already existed. What is the, what is so great about summarizing articles? I've never understood. It's a product sold to people that don't care about knowledge. It's a product sold to people who are it's a cynical thing for people that want to sound smart without being smart, much like chat.

GPT is a product made for people that are smart imbeciles, for people that love learning facts without learning context. They love generating a list of stuff that they can memorize and go, hmm, yeah, I know all the dates when things happened. Oh, did you not hear about this? Hmm. They don't make arguments. They have arguments made for them that they then slurp up and spit out. It's gross, it's boring, and it's stupid. And while there may be times when you can go to a large language model and get an answer that's useful, ultimately, we are two years in, and the best they've got is summarizing articles, the best

they've got is like, what if you could ask a website a question? Gee whiz, did we ever have that? Did you see the Gemini Live one though? The Gemini... Oh, this one was good, so it's for the Pixel phone, and it's a guy, he's got his Gemini on a Pixel out, and it's like asking him questions for a job interview. And he's giving very bland answers like, oh, I dealt with challenging circumstances, and it's this, like, sad music, and it's him, like, taking his daughter to school, and at

one point he Googles, um, how to teach a child about bullying. I saw this one, yes. Which is so funny as well, because I didn't look up that search, but you know there's like eight SEO articles at the top, like, eight ways to tell your child about bullying. And then, he then gets to the end, and you're like, okay, so is this not a job interview? And then he joins the call, so it's like, this guy's like old old. Like he's like graying, his daughter has gone off to college, so it's like, in this dystopia, where you're preparing for a job interview, in this very vague way, like it's not obvious

what his answers were meant to get, or indeed what prompted this. But he's like a man in his 70s joining a job interview, like a remote job interview, he's like, oh god, the American safety net is gone. Social security has been ended by doggy. But the biggest thing I noticed about this commercial was that I do not know what the product is. I could not tell you, got into my head, maybe it's Gemini Live? And that I only know because it said it on the screen.

What if you could talk to the phone and the phone could sometimes say stuff back? Hmm? Is that good? Do you like this? What do you think? Please buy it. Well, it's free. And it loses us money every time you use it.

So use it, but maybe don't use it. But also you could use it in some way. How? Well, we just bought this $8 million commercial and we do not know. Enjoy it. But that's the thing. I get a lot of shit about the AI stuff. I get a lot of people saying, oh yeah, well, you're just a cynic. Oh, you just, you don't see the magic of it. And then we get to the Super Bowl. Even the crypto companies had more juice than this. ALICE You had Matt Damon being like, you fucking moron. Join me in the future, you pig.

RILEY Yeah, you can make some money. ALICE And you had Larry David go, ehhh! And then you... which is so funny. You had Tom Brady being like, yeah, well, I need you to do this now, if you don't do this. The implication with all of that was, you're a moron if you don't use this. The implication from the AI one was, ah shit, we didn't fucking come up with an idea. Oh, we didn't come up with a reason to use this. Oh Dave, we said we were meant to come up with a use case. I don't fucking have one. Summarize an article? Sure. I don't fucking care. We've been paid way too much money for this. Sold.

It's just, it's so bizarre. Because you've got the whole market screaming, all the media screaming about AI being the future. Look at all the things that AI can do. Well don't look. Don't look at what AI can do right now. Do not look at it. It's not good.

However, what if it was? Wouldn't that be good? Also it loses a bunch of money. That's bad. But it won't. How? Hmm? Wouldn't that be good? Also, it loses a bunch of money. That's bad. But it won't. How? I don't know.

There's a lot of just like weird fill-ins. I just don't get it. I really thought by now I would have something to argue with, but I have less. Just fantastical stuff, I guess. You know, I don't talk about this much on the show, but I love styling myself. Nothing gets me more jazzed up than finding a new wardrobe staple that reinvents all of my potential fits.

It's just a fun way to change up my life when things are getting stale. Now that used to mean breaking the bank, but with Quince, I get the high end versatile pieces at prices I can actually afford. Now I can upgrade my style by snagging killer luxury essentials that sync with my vibe and with my wallet. Quince has all the must-haves like Mongolian cashmere crewneck sweaters from $50,

iconic 100% leather jackets, and versatile flow knit activewear. The best part? All Quince items are priced at 50 to 80% less than similar brands. By partnering directly with top factories, 50 to 80% less than similar brands. By partnering directly with Top Factories, Quince cuts out the cost of the middleman and to keep high-end wardrobe staples affordable and accessible. Now, they sent me a couple items to try out. Check out these corduroy pants.

If you're not watching on YouTube, you can't see them, but oh my God, these are so sleek and comfortable. They're black. I can wear them on a plane. They're so comfortable. I can also wear them to a nice restaurant because they look sophisticated. They got a wonderful hand feel.

They just make me want to touch my own thighs, which is a nice feeling. Just rubbing my hands up and down my own thighs. Sometimes I'm like, I shouldn't wear these to the business meeting. I'm going to be rubbing my thighs too much. You know what I mean? So if you want to do that,

if you want to indulge in affordable luxury that makes you want to touch your own thighs over and over again, go to quince.com slash factually for free shipping on your order and 365 day returns. That's Q U I N C E dot com slash factually to get free shipping and 365 day returns quince.com slash factually. Folks, if you've listened to this show before, you've heard about our next sponsor because they are one of my favorites. This is a product that I've been using for years. And in fact, the only reason they sponsor this show is because they heard me mention them

on a different podcast because of how much I love the product. And then they said, Hey, let's reach out and do some ads. And I said, thank you very much because I can endorse your product wholeheartedly. The name of the product, of course, is delete me. You know, it's hard to face, but our data is out there on the internet. There is an entire thriving ecosystem of companies called data brokers. These are companies that collect, buy,

and sell our personal information. It's completely legal. They collect your name, your phone number, your home address, the names of your relatives and friends. All of them are out there and they put them all in a centralized database waiting to be snapped up by anyone with an internet connection and a couple bucks to pay for your data. And bad actors really use these data brokers to exploit and harass people they want to

hurt. A couple years ago, I started getting phone calls in the middle of the night from mysterious numbers that were saying, I don't even want to say what horrible things to me on the internet. It was awful. And I realized this felt invasive. It felt bad. It was awful. And I realized this felt invasive, it felt bad, it was scary, and that's when I turned to Delete Me. And as soon as I signed up for Delete Me, these calls stopped, because you know what they do?

They go to every single data broker and they file an opt out on your behalf, legally requiring that data broker to remove your information. Now, that's a lot of work for you to do as a person. That's why it is such a great service that delete me does it for you Online bullying and harassment have only been on the rise in recent years seems like every couple weeks I hear a story about a real-life incident a stalking incident that began as an argument or interaction

Between two strangers on the internet. They just got so heated one person said I've got to look up that person's real life and screw with them and you deserve to Feel secure and anonymous and not be constantly on the lookout for people trying to hurt you and get this you can get 20% off your delete me plan when you go to join delete me comm slash Adam and use promo code Adam at checkout That's join delete me comm slash Adam promo code Adam eliteme.com slash Adam promo code Adam. I do believe that we are going to see AI more and more in our lives, but I think what we're going to see is the very useless models that you're talking about being jammed into more and more systems that

we have to experience. And so our experience of AI is going to be arguing with AI assistant chat agent going fuck, fuck, fuck representative. Fuck, fuck you. Fuck you. Give me, give me a human. Give me a human. My pen broke. I got so mad. Adam, we already do this. We are already doing this today. This already exists. You can't even in this example you brought up to say how this will continue. It's still the thing that's already happening. I've been doing that for a decade. Me and the Verizon chat, we're not on speaking terms. Yeah. Like it's. But it's going to be a supercharged version of that that we'll see in more and more

places. Just functionally different to what I mean, it's just, just because at this point, that is everything. That is everything you use. They're already, for the last 10 years, Big Tech has educated the economy and consumers against having customer service. If your Facebook account ever gets banned, good luck. My name was banned for three to six months and it took begging someone on Twitch being like someone, because they're going to delete it. And they never told me why, but they were locked out. They never explained why.

You look up, oh, sorry. Now you can pay for Facebook premium verified. It's like 15 bucks a month and you can get customer support with that. It's not like, and that is 100% true. It sucks. Jesus. I know it's fucking bleak, but that is everything. You go on Amazon now, you can get to a customer service rep.

It just takes a few tricks. You have to like do a sleight of hand. You have to be like a, like a trickster to get around. You have to mess with the bot until it goes, ha! And then it just goes, talk to a person. And they're usually very pleasant to give them credit, but they're already doing this. They're already been working on this. So oh no, we're going to have AI chatbots everywhere.

We already fucking do, they suck. It's just we're going to have, at some point they're going to hand it off to something important and something that's gonna break really badly. There's gonna be someone, a banking institution probably won't be it, because they think about the SEC, well we don't have the SEC anymore. I don't know what's gonna happen with some of it. So many variables as well, where it's like, if they just delete part of the government, is that gonna...

Well, let's come back to the government. I mean, you mentioned crypto and how crypto had more juice. And I've talked on the show for years about crypto being a bubble and not really being a product, having no use case and just being a constant, you know, pump and dump scheme. It seems like crypto's plan to save itself has been to embed itself in the U.S. government, to get the U.S. government to like buy a strategic reserve of crypto and hold some.

And then that de facto turns it into a much stronger, you know, commodity or currency than it would have been otherwise. At, you know, basically a taxpayer expense. Like it's, you know, it's being bolstered by, it's being propped up by, you know, the government of the United States. That's actually a pretty good social engineering scheme

to create sort of perpetual value around your useless product. If you can get the federal government to invest a lot of money in it. Is that part of what is going on with, you know, Elon putting, saying we're gonna put AI into the federal government is to try to, as you say, we're going to force this product, we're gonna to put AI into the federal government is to try to, as you say, we're going to force

this product, we're going to make it penetrate everywhere so that people are forced to deal with it as a way to socially engineer the omnipresence of this useless technology. Mason- So crypto is weird because crypto found another use case, which is they found some way to turn the fake money into real money and then shoved it into campaigns. It's all nakedly illicit. It's all nakedly manipulated. But they found a way to financialize it to the point that they can get some money out of it. I do think that there will be more crashes just because this whole thing is unstable. But there is a use case, which is dark money. It's the dark money machine. They found a way to do that.

By comparison, large language models haven't got shit. And also, let's say that they do have AI in the entire government. Let's say government does half a trillion dollars a year of just AI contracts. That would require them to do so much that doesn't exist. There's not enough reasons to use it. There's not enough use cases, even if they put it fucking everywhere, a chatbot everywhere. First of all, this stuff's horribly unprofitable. Second of all, it's not that much money will come out of it.

There really isn't. Like there isn't that much cash. There aren't that many use cases. Even if they put it everywhere, what can it do other than what? Rag search so it can do like generative search. Yeah. Okay. And you've got chat bots. And then you've got war. And that really is it. What is it with crypto? There's money, money and money out. Like it's nakedly illegal, but we don't have laws. Like, I mean, you can fire all the air traffic controllers and

put all the flight paths into a generative AI and have it output landing instructions. I'm going to assume that you could rig up an AI to do something like that. That might save you how many millions of dollars a year by not having any air traffic controllers? Ten million, twenty million dollars a year. Well, they would never do that for a number of reasons. But if they did the first private plane to crash would be the end of that. Yes. The second all of this stuff stops. I mean, we're hoping as somebody who takes 50 flights a year to go to Adam Convert.net for my tickets and tour dates.

You know, I'm worried about flight safety and the only thing keeping me confident about it is, you know what? This is the one thing that might affect the Mark Andreessen's of the world and the Donald Trump's of the world that like they do need their fucking planes to land. And they really do. But even then, assuming that model, how much revenue comes out of that? Because right, that's my point. If we're thinking about who would also be the because this this comes a lot of it's

like, okay, Elon will put AI into everything. He'll put grok into everything. Horrible idea. Terrible. How much money though? Because generative AI is unprofitable. Inference, which is the thing you do when you prompt it. Everyone's saying the price is coming down. They're not saying it's profitable because it isn't. It isn't a very, very inefficient

software. And what's terrible is America's AI companies do not care about efficiency. They are fat and happy and lazy. And so, even if you put this everywhere, there's not that much money coming out and there's definitely not more money coming out. It's not like they're going to make enough money to cover the costs. I guess you can just have this big fucking loss leader, but for whom? Elon Musk? So there's going to be, what, fight?

Let's assume the rosiest thing. And to be clear, OpenAI only makes a billion dollars on API calls. So people plugging into their models. It's the most popular generative AI company in the economy. Barely scratch a bill. How much money is reasonably, even though the whole government had a chat bot, how much money is actually coming out of that for Elon Musk? Probably not that much. And I guess this is putting aside of how horrible this would be. It'd be fucking terrible. It sounds terrible and awful. But as far as the saving the industry thing, which is the classic thing people are saying, OpenAI is just going to stick into the government, then money will come

out. How much money? How much? Well, the government's got unlimited money. Yeah, but are they giving it to him? Are they going to, how is the money? I mean, that makes OpenAI a giant contractor that is, you know, their services being- They're losing money on everything. And even makes OpenAI a giant contractor that is, you know, their services being... They're losing money on everything. And even if they're a giant contractor, they make for... Right. If they lose money on every API call, then if the government is doing billions of API calls, they're just losing more money.

And they're not going to do billions of API calls. These are government services. I guess even if it's crunching inside it, but at that point, even the dumbest imbecile will go, why don't we do an on-device language model? Now again, they're not doing efficiency, they're doing something stupid. But on top of that, open AI is not going in the government. Elon Musk will stop that. He hates Sam Altman.

He truly despises Sam Altman because he's like, I'm the only person who can be a giant con artist in AI. Sam Altman, this he's like, I'm the only person who can be a giant con artist in AI. Sam Altman, this is my job. Immigrant coming to America, taking proud American con artist jobs. Elon Musk is taking everything from this country. Let's talk about the efficiency piece of it because Deep Mind, Deep Seek, excuse me, Deep Seek is the Chinese AI company that, I mean, there was literally,

this had an effect on the overall American stock market when this product came out, because this is a Chinese model that is far, far cheaper to run than the, this by the way was the original reason we booked you on the show. Yes. Because we want to talk about this.

And then in the time between the booking and this interview, a lot more shit went down with AI and the government and Elon. But so how does that reshape AI in America? So, to explain what DeepSeek did in the most simple way, what they did was because they didn't have access to as many GPUs and mostly had access to old GPUs. I must be clear, DeepSeek was trained and run and funded by a hedge fund called High Flyer. Now, the big thing is, is they had limitations. They had limitations on how many, they had like tens of thousands of chips, but some of them are older generation. As a result, when they were

training this model and when they were running it and making it work, they did multiple models. They did V3, which is compared to the main model for chat GPT, GPT-40. They also did a reasoning model called R1, which is compared to O1, which is OpenAI's reasoning thing. Get to why that's big in a second. But they had these kind of barriers to entry to make this thing work. So they actually had to do hardware level. They had to do like a digital processing unit. They had to work out how to do things with less memory bandwidth. So they were forced by constraints to make something more efficient. What's this sound like? Silicon Valley. The problem is that Silicon Valley

is no longer this. So DeepSeek, and yes, they might be funded by China. We don't know. Doesn't matter because they open sourced this model. They put all their research out there. It was allegedly trained for, I think, they said six million, the capital expenditures are probably like one and a half billion dollars, but that's for all the infrastructure, which is still like much, much cheaper than what OpenAI uses with Microsoft, like Microsoft billions and billions of billions into CapEx. But nevertheless, they were able to train this model they claim for like five and a half million dollars, probably more than that. But still, OpenAI's 4.0, that cost a hundred million or more than that. But still open AI's four row that cost a hundred

million or more to train. So this model is cheaper to train and it's cheaper to run much more efficient. All their models are there are one reasoning model is like 30 times cheaper than open AI's. Wow. So the big reason this is so scary isn't just the cheapness. It's the fact that it kind of broke the paradigm of we need the biggest, most hugest chips every single year without fail. Because they didn't. It also broke OpenAI. Because OpenAI, when they released O1, their reasoning model, there was a big buildup to it with all these rumors calling it strawberry.

Everyone really lost it. And all the media, again, shat their pants on this one. They were like, OpenAI's O1 reasoning model's here. It can think. What can it do? I'm afraid I have a phone call. I will be right back. Every fucking time.

But it was like moon magic, everyone's like, open AI, they've made reasoning, everyone started copying open AI. And then this random Chinese hedge fund is like, we did it. We did it, we open sourced it, it is all the research, go fucking nuts. And the funny thing is, the model was trained off of outputs from OpenAI, so OpenAI was like, should we sue them for copyright? And it's like, oh whaaa, my plagiarism machine got plagiarized! Fucking cowards.

Anywho. So you now have the broken myth of OpenAI being magical in this big research hub, because someone else did it real easy, like. And you have this thing of, why do we have to keep buying new Nvidia chips as well? We can do it for cheaper. And also why do we need all this CapEx? Why do we need to spend all this money to build this out? We could do it with all the chips. What do we need? And still, the market's still going, but there is genuine, there has been a vibe shift, a nasty one too, where everyone's like, huh,

did we put all this money and attention into something that's just a kind of a boring piece of cloud software? Yeah. I sure hope we didn't do that. Yeah. So now we've got this weird situation where OpenAI, before this happened, was like the god of reasoning and no one could possibly work out how they did it. Now they're not. Now it's commoditized. Now everything's commoditized. And now, well, they only have to build products. And as we've established, they're not good at that. So we have this weird thing. And then over at the side, we have Mark Zuckerberg. Looks like a clown without its makeup on. Very horrible looking man. And he, by the information report, this, he convened four war rooms to work out what DeepSeek did.

And he has said very clearly that he wants to commoditize large language models. He wants to make it so that it's the products that stand out. Great fucking idea, Mark. That sounds great. Facebook's products sure looking good right now, buddy. But nevertheless, if Zuckerberg decides that, he has unilateral control over better. They have hundreds of thousands of GPUs. If he wants to make his own deep sea reasoning thing, he can, and I think he will.

So now there's this uneasy tension. There's this thing of, well, shit, I hope open AI does something. I hope Microsoft does something. I hope something happens. I hope something good happens. And also a few other narratives are catching up with them. Don't look so good. So open AI is revenue. Fun fact, most of their revenue does not come from people plugging their software into OpenAI's

models. You'd think, right? They're the biggest game in town, so most of their revenue must come from this magical thing they sell. Over 70% of their revenue comes from subscriptions to ChatGPT. That is an insanely bad business. Sure, they currently have 15.5 million subscribers. They make a bunch of revenue. But yeah, the name synonymous with the movement that everyone's talking about that no one can tell you what it does, of course you're going to go to the biggest company and you're going to go, I think this shit, I'm going to see what this fucking does. And yes, entire organizations and management consultants and businesses run by people that don't do real work will pick it up. But man, if that's all this company is a cloud software subscription service, right, they will be commoditized. There is no if

deep research. They're one like other than operator, the agent thing that does not work. Deep research got commoditized immediately. Perplexity had deep research like a week later and Aravind from Perplexity's not the greatest CEO. Like he tried to merge with TikTok just over the Christmas holidays. He was like, yeah, what if Perplexity merged with TikTok? Yeah, my company fucking sucks and loses a bunch of money. Can I merge with your big successful one? We're an AI search company. You're TikTok, you're globally renowned, we are a search engine that like 10 million, 50 million people use. Can we merge with you? This is the brain trust. This is the people. So yeah, it has raised the temperature on this whole thing. And even in the last earnings, where every single

hyperscaler was like, yeah, so this doesn't do anything. So this sucks. You hate it. It doesn't really make us money. Yeah, we're going to double our capex. Fuck you. I don't know if they doubled it, but they've all raised their capex. So we're in this insane kind of group delusion almost where no one really wants to be the

one to say, hey, we all doing this for no reason. Is this really stupid? And now OpenAI is raising money from the god of investment, Masayoshi's son of SoftBank. They're raising allegedly $40 billion. What's weirder is, and this is the thing with this whole narrative, there's always asterisks, it's always weird. So OpenAI is raising up to $40 billion, led by SoftBank, famous for WeWork and Wirecard, the fraudulent payment company. They've lost $30 billion in the last few years. It's fucking crazy. They are leading this round and they're

up there, $15 to $20 billion SoftBank's going can put in, but what's funny is you've heard of Stargate, right? Stargate's this new alleged private partnership. It was announced with Trump there, but up to $500 billion to build miscellaneous data centers for private AI companies led by open AI. Open AI commits $19 billion to this. Wouldn't you know it? Part of the reason they're having to raise $40 billion is because they do not have $19 billion and they only lose money. And SoftBank is also putting

money into this deal. And I don't know for listeners if they know what SoftBank is. SoftBank is a giant Japanese conglomerate that has a venture uncle, The Vision Fund. And what's great is they had a really early win on Alibaba. Really insanely good pull that they're still making money off of because they were so early. Most of their other investments not so good. This Masayoshi-san, the CEO of SoftBank, he was like begging Adam Newman from WeWork, begging to invest. He begged Sam Altman to invest in OpenAI.

SoftBank is one of the only investors that lost money on Uber. They got in so late that they were like, double it down, baby. And they lost. You lose money on the most obvious. Softbank rocks. Anyway, when Softbank invests in the company, it's a great time to start questioning the fundamentals of the company because they do not look at them. So you've got this weird thing where all of this money sloshing around, all these companies are promising all this money. where all of this money sloshing around, all these companies are promising all this money, but in the middle there's nothing. There's nothing going on. There really isn't going on. And critics of my work will say, look, oh, well, something's going to happen. They're going to work it out. When? When? Why do the rest of us have to go through our lives, like repeatedly having to justify our existences in the workplace and our relationships? But these chunder fucks, they get

out there and they're like, so search engine, it kind of sucks, it's really expensive and loses a bunch of money. It could be God. Can I have 40 billion dollars? And also I need that now. I need that now. I know I just raised 6.6 billion dollars late last year. I need more money than that.

And I'm also, by the way, you're going to give me 40 billion dollars. I'm going to need another 50 in like six months. Why? Because my software fucking sucks, dude. It's so bad. But it's also the most powerful thing in the world. My company is sick. I need money, but it's also so good. We're going to birth God. If I get enough money, I do not have enough. I feel crazy when I talk about this stuff because it's, there's so much out there just saying AI is the future. But when you read the article, it's not about AI. I mean, how much is this the sickness of the tech industry overall that I mean, we talked about this last time that they had, you know, a couple of big wins early on that, you know,

the tech industry picked all this low hanging fruit actually changed the world in many, many ways. And now they're addicted to that model and they're pumping up these things in order to create that appearance, but now there's nothing at the core. And does that explain, again, them taking over the federal government? That like, when you have nothing, you need to take over something that's actually something.

So it's what I call the raw economy, which is this growth at all costs mindset that has dominated pretty much the entire economy, but software is so perfect for it, because other things you need to build things, economy, which is this growth at all cost mindset has dominated pretty much the entire economy. But software is so perfect for it because other things you need to build things and things exist in meat space. But with software, you can put it all in places and you can sell to new people and you can upsell to those people. You don't need to do many extra things to upsell to them. This worked really well for a while. And when you look at everything through the lens of growth, you start cutting away

the troublesome people like, Hey, is this useful? Is this good? Do the customers like it? Fuck off. Yeah. Out with you. How dare you? We're trying to grow. All these companies are oriented around growth. They want to grow double digit percentages every quarter. Every single product that they put out is kind of oriented around that. Microsoft's 365 suite, for example. Anyone who works in a business that uses Microsoft will know that Microsoft Teams does not work half the time. That now you load Word and it's like, Hey, do you want to do AI? It's like, no, mate, I want to write it. It's like, you sure? And it pops up again.

It's like, Hey, you seem to have, you seem to have closed the window real quick. Just bring this up again. Oh, so that's never a good look. But they're doing this because they, by putting AI in it, they're thinking, okay, now we cannot sell customers more. The problem is, it's been like 10 years since they've had something big or good, like a real growth market. And by growth market, I mean another way to bolt on $5 to $10 billion of revenue a quarter. None of these companies have had one of those for a while. What has always worked is throwing a bunch of money at stuff, building a nasty monopoly, making it so that you can lower the price of your software so that you can

sell into the enterprise, so you can sell to a bunch of fucking people, or finding ways to trap people in your ecosystem so they can never leave, so they have to keep paying you so you can raise the rent on that. It is rent-seeking, ultimately. But growth, and really the only thing that grows forever, is cancer. So right now, the tech industry oriented around growth while they see AI. I'm like, wow, that sounds futuristic. And I mean, the whole reason that Microsoft invested in all these GPUs is because they saw chat GPT and they went, but I have that shit on Bing. I'm going to get down Bing, get on Bing right now by all the GPUs,

Amy Hood, go, go, go, go. We need to call in video. And they did that because none of these companies really understand customers or problems. I know I'm speaking kind of vaguely, but really, if you look at what they're building these days, they're no longer solving problems. They think AI, magical, right? And these little pig losers will buy anything. It really is contempt. It's like these, look how amazing this is. Why is it amazing?

Shut the fuck up. And every time it's like, why is this amazing? Well, computer AI. This worked before and it did work before. People would just buy whatever the tech industry put in front of them. But even then, even in the past, even with these little things, they found a way to get in. They found a way to sell in. What they've never done is tape themselves to the side of just an insanely unprofitable thing. And I think the reason they've done it is they don't understand human problems. They don't speak to people. They don't understand business problems.

These companies are not oriented around solutions. Their solutions they oriented around are how can we sell more to customers and how can we come up with new customers? And sure, that's capitalism. Absolutely. That is a lot of them, but that is the poison in the veins of the tech industry. When everything's oriented around number go up, the only thing you optimize for is engagement. The only thing you optimize for is keeping the people on the service in the app, which

can either mean you have a monopoly, you kill off your opponents, or you make it so that the app is harder to use. Use any modern app. You also don't maintain anything. So if you're using, I don't know anyone watching this, if you're using an app and it just stops fucking working, you are not alone. Such a big part of my work, I'm saying this. You are not alone. Everyone has, every single person you talk to, their software is breaking down. Every bit of software isn't working. And it's because these companies do not sell software anymore. They sell experiences, which by me, I mean traps. I mean traps. They've convinced your boss that they need a three year subscription to 365.

They've convinced your other boss that you need Slack, which you also have in Microsoft Teams. Do they talk together? No, but you need some clients on Slack, some people on Teams. Both apps crash randomly. Both of them different, different like video experiences. They both suck. They've just got notifications.

I've been using slack for 10 years Hasn't gotten any better in 10 years the same thing. It's got worse. I would say it's just more bleep bloop So there's like lights everywhere harder to use. Yeah, and that's because they're not making slack better They're making slack more sellable to more people. They're trying to find ways to juice more out of it I realize I'm going on a bit But where this comes in with AI is when you don't really have companies that solve problems anymore, when your idea, when you come up with the next big thing is not what could this do, but what could this could do for us, you're

inherently going to choose something like AI. Because when you are a, when you're Sachin Adela, a management consultant, he's an MBA, sorry, wouldn't dare call someone Sundar P Pichai, Forman McKinsey, Google. But Sachin Adela, when he brings, even adding chat GPT to Bing. Okay, so search is better. Cool. Now he's putting AI models all over as your why? Isn't really obvious. You can do copilot for sales. What does copilot for sales do? You'd be shocked to hear it summarizes your emails and presentations. More fucking summaries. And the reason they're making that isn't because it's going to improve sales. It might marginally. They're doing it because they need a new fucking SKU to sell. Everything's about selling something new. Sam Altman, OpenAI, if he'd have been like,

yeah, it's a large language model, it does this. Never got any investment. By saying, this will become literally the smartest person ever, but it will be on the computer ever, please give me money so I can build that because it doesn't exist yet. Yeah, that's a growth story. Everyone wants the growth story. Everything must be the growth story. And also, when you get to the government, yeah, it has the exact same attitude. Except it kind of inverts, which is, what helps with growth? Cutting costs. You don't do anything when you're a lazy fucking rich piece of shit. All you do is you sit around with one of your 800 children that you ignore called like 192.168.11. One of your fucking kids that you're ignoring, you're like, what does business work like?

And also, Tesla is not doing well either. Either way, you look at the government, you want growth, so you cut things. You save money. You make federal employees that have dedicated their lives to the country. Come fuck them. Who gives a shit? Because we need money go up, number go up. It's sick, but it is that exact rot economy mindset.

You have this attitude of, I don't do anything, so these people working for the government must also not do anything. And how did I become rich? Because of my business mind and efficiency. It's all of these terms you memorize while not doing work. You're like, what the business do? Number go up, number go down. Expansion? Growth? People cost money? What a class for something.

You're making a lot of sense of this because, look, the government has been run by big business for many years in one way or another since hundreds of years, frankly, it's like America is a capitalist country. Right. And so we've been governed by capitalists. But, you know, when we were being run by GE and, you know, even fucking Goldman Sachs, those are parts of the regular economy. And they, I think maybe had a more sophisticated understanding that you need the country to be regulated in some way, right? You need people to be able to make livings. You need planes to be able to land. You need, you know, order in the country.

And so there was, you know, at least a maintenance of the system. But the tech industry playbook has become so brain dead and empty and based around puffery and cutting and number go up to such an extreme degree that these people have decided to ally with an extremely brain dead right wing political movement that also just wants to see the world burn. That is what tech and honestly, Donald Trump's political coalition seem to have in common.

All they want is to set fire to things and they think that that's progress. They're both stupid enough to believe that. Jason McAvoy And it really is that distance between production and capital. Because all of these things, sure, if you're a little bit closer to production, even if you're far away from it, you're like, okay, things still need to work. Maybe there's a bunch of people I don't really totally understand, but there is a logic. Big systems are made up of billions of little pieces.

That makes sense to me. And I understand that as a person who's closer to production or whatever. I understand that because I look at it and I go, okay, it's an intricate system. I don't fully understand it, but I'm smart enough to say I don't fully understand it. I'm not just going to chop bits off. Right. When you are so much further from production or real work, when did Elon Musk last work? Like, come on. He works 130 hours. Fuck you. Just not, but he doesn't know how anything works.

He really doesn't touch his companies. The keys are in someone else's hand. You don't want this man. ALICE His whole MO is in his own businesses, at Tesla, at Twitter, to show up and go, what's this? What's this for? We don't need that. And to make everybody scramble and get rid of that thing. SEAN What is this? That's a toilet, Elon. I will get rid of it. My favorite moment with that as well with Twitter was when he found the StayWoke t-shirts. He was like, oh no. He got so angry at the StayWoke, he was like, oh my god, no, I can't have this.

He was fucking... But that's the thing, though. When you don't really experience human problems, you can't solve them, and if you don't... If you think you're so fucking smart because the legend in your head you've built up is all that's left, because you actually look at what you've done in your life, you've not done much more than shake hands with people and go to lunch. You're like, how does business work? They're efficient, right? Less, but you get more. Number go up and number go down. If you're not actually experiencing how things are made, how industry functions,

or worse still, if you're taking that lack of knowledge and going, and the government is run like a business, which it is not. Which it is not. It's run very differently to a business, has longer timescales, and our entire business, all of the businesses in the economy are now running this not short-termist view. It's an insane way to look at a society. There are many things in society that cannot be provided like a business. Yes. Such as education, such as roads, such as, there's many, this is extremely basic as a fact, but literally not everything can turn a profit.

No. And indeed it shouldn't have to, in the case of like roads and school. These are not businesses. These are functions of society. Unless you don't interact with society, unless your entire everything is above it. Because that really is the problem with Elon Musk. It's the problem with one of the problems with Elon Musk. It's one of the major things to the tech industry. It's why everything with tech has got worse because

everyone's outside. They're not looking at what real people do. Otherwise, they'd look at the large language models and go, this doesn't really help. Mark Cuban yesterday on Blue Sky, he was like, well, I don't understand why people hate large language models. Sure, they get things wrong, but to someone without any education, they could learn anything. It's like, Mark, you fucking moron. What the fuck are you talking about? A person who doesn't know anything wouldn't know when it's getting it wrong. They'll learn wrong things. You God damn. But again, if you don't interact with real people and their real problems, the idea of a magical poor person that you would never look at, let alone touch,

using the special computer thing that you yourself would never use, well, of course that's magical. But it's scary. All of this is scary. And the government stuff's scary, but the tech industry stuff scares me too, just because all of this money is going nowhere. It's going into something that has to collapse. Right. That's the fear about DeepSeek. I want to make sure. Which Deep thing is which? Well, it's a grander fear, which is right now, we we are two years in and we've not seen really anything come out.

But what we've also seen is a bunch of people just leave OpenAI and then go to Anthropic and then leave Anthropic and start a new startup that doesn't have a product. We don't see, I don't see a single sign that things are changing. Even LLM experts are kind of like, yeah, it's just this, I guess. And at some point someone will have to blink. The capital expenditures are too high. It's hundreds this, I guess. And at some point, someone will have to blink. The capital expenditures are too high. It's hundreds of billions of dollars. At some point, someone's going to go, hey, quick question.

You seem to be losing a shit ton of money. Why? The markets analysts are already asking these questions. At some point, this falls apart. And this is actually my greater worry. When the AI bubble bursts, there is something worse to come, which is the collapse of the value of tech stocks. Value of tech stocks is driven by growth.

It's driven by the idea that software can endlessly proliferate. And as a result, software can always grow things. Software businesses can grow forever. Problem is, if the AI bubble bursts, which it will, I'm very confident. At that point, everyone's going to go, did you just sell me for two fucking years, this line of shit about your shitty encyclopedia, waking up and becoming God? Did you do that? Microsoft? Why'd you do that? Google? Why'd you do that? Meta is over here. Meta, you can never be fired. It'd be fine. Stock will take a cut, but that's a separate problem. All of these companies and they'll go, you spent all this money on what? And the GPUs are not

going to be, you can adapt them to other things. There's accelerated computing, sure. But there's not really other use cases for the GPUs. So you're in this weird situation now where like, I think that there is just going to be a reevaluation of what a tech stock should be valued at. And it's not going to be nice. It's not going to be nice at all. And I think that like a third of all venture capital investment into startups went to AI last year. Like the venture capital industry is not oriented around creating value anymore. The tech industry isn't. And once the markets turn on the tech industry, I think consumers will continue to. My experience of talking to regular people, part better off lying, I force myself

to talk to as many people as possible who are not tech people. And I have yet to meet one normal person who is even slightly happy with chat GPT. I've met one person who was like a hairdresser. It was like, yeah, ask it some questions sometimes. Like, great. Welcome to the future. Once this bubble bursts, the tech industry is going on notice. And I worry about that because it will lead to tens of thousands of layoffs. It's going to lead to, I don't know if it's an apocalypse, but it's going to lead to a complete reevaluation of how software is valued. Because to your point earlier, venture capitalists have not been oriented around making the next big thing in the sense of something that changes society. They've been trying to come up with abstract financial things that they can build and then flog. That's everything. That's everything in the tech industry right now.

That is the rot economy. And that is what really could lead to tech's gin... I won't say downfall because it's not going anywhere, but I think there's going to be a massive reckoning. Well, and that's going to affect the rest of the economy and, you know, real people's pensions. I mean, look, I'm on the board of my union. I know how we do our finances, you know, it's like the pension plan is like invested in these companies, right? Everybody has some investment in these companies, whether or not you know it, you participate in some larger, something that you depend on

is invested in the stocks of these companies. And the tech industry has become the only growing part of the American economy because it has eaten every other industry. You look at media, you know, the entertainment industry, manufacture so much of the of our entire economy has been, you know, sucked up by some tech company. And, you know, they're they're cutting and doing layoffs

because tech is doing it instead. So once like tech craters, then why have we disemboweled the rest of the economy? What happens to the rest of us? And that is a good bloody question. Yeah. I also think consumers are fucking furious at the tech industry. Our software is not working right. The experience of using the computer is worse than it's ever been. I have a brand new iPad Pro I love. For some reason, when you try and switch apps,

it covers it in blue. When you're using the keyboard, it covers it, it highlights it in blue, which works really well with Apple apps. It does like a nice little thing. If you use non-Apple apps, sometimes it just covers one side of your screen in blue. You may think, Ed, you could just turn that off, right? Go and turn off the option that controls that. It's called full keyboard access. The problem is that that would not let me use the keyboard I bought from Apple. That was quite expensive. This is how computers work. Now we go on Google search doesn't work. We go on Facebook. It's like ad, ad sponsored person, person ad, ad, ad sponsored AI slop, AI slop. We go, we go to use Microsoft Teams, your audio,

your video doesn't work. Every app you use is a different weird incentive poking you because they need to increase engagement, they need to increase time on app. It's a pop up for a subscription that says 2.99, very small word, weekly. And it's trying to get you to everything's track. You log on to Dropbox. It says, do you want an annual subscription? No, mate, I want to look at my files, get it down the fucking way. It's everyone is being fucked with all the time. I actually think that this is a wider social problem than people realize.

Yeah. That all of our stuff just isn't working. And it's because the tech industry has been rewarded for it. They've been rewarded by the markets for growth. Everything must grow. All numbers must go up. One of my favorites to choose is the match group. So online dating right now, which is insane if you're single, I mean, it's like, you go on there and it's like, hey, do you want all of these seven people that you think you'd really like? That'll be $2.99, you little pig bitch. Fuck you. Welcome to Match.com. Hey, you want to go use one of the competitors? We own them too.

What you gonna do? You're gonna go to Match.com? We own them too. What you gonna do? You gonna go to match.com? We own that. Bumble. Oh, different monopoly, mate. They're also doing this though. They're all being sued now. The experience of using the computer is insane now.

You go on Google and it's like sponsored links everywhere. They've kind of embedded. It drives me insane because it's everywhere and no one's talking about it. And we've been educated by the tech industry to gaslight ourselves, to be like, I'm just behind with the times. I'm just, I must be losing it. Trust me, I'm on the computer all the time. It's been just getting worse. It's not like it. And all of the power users watching this will be like, huh, I deal with this every day. First of all, you're not indicative of the majority of people. Second of all, if you are a power user of any computer, think about all the little

interruptions you're getting with all the little things that the fresh install of windows you put on every computer you buy just because there's all this slop they've added. They've added, they've got like search slop in the, they've got ads. Actually, here's a great example. The most popular laptop on Amazon at the time when I wrote this thing last year was an Acer Aspire 1. It took like 45 minutes to set up.

There are ads in the start menu. There's ads in Edge. There are ads when you're waiting for it to install. The keyboard physically bows when you touch it. The menus will not load properly. They're like judder and shutter. This is how millions of people experience the computer. And it's very easy to be like, Oh, I'm a Mac user.

It's not my problem. It's your problem when everyone's being driven insane by it. And this is all growth because if they were like, Oh shit, if I lose a customer or a few million customers, if I piss off my customers, they might go and use someone else. Except when everyone's like, we're all shit. Everything sucks. We don't need to make this better. Apple has made so much money just being like, what if it wasn't shit?

What if this wasn't as shit? But you still go into your settings. The settings in the Apple, in your iPhone, they're like an MC Escher painting. There's like 19 different options. They have a search bar in there because it's so hard to find the thing. Yeah. And even they have run out of, I mean, look, Apple is, you know, I think in many ways, the most functional of these companies because they've had a focus on actually creating a

good experience for a long time. They've been very much the HBO of the tech industry, right? You pay a little bit more, you get the actual premium experience for the rich people. You're stuck within their walled garden. You have to get all the apples. It's nice, right? And you gotta buy a new thing every couple of years,

but hey, it's actually nice. It's actually nice. And then even they, jamming AI into shit. Some of the worst features I've ever seen from Apple. And the mercy of it from Apple, from as an Apple user is that their AI features, they haven't really jammed them in that hard. They're like, hey, we'll summarize your text messages.

You can make a gen AI emoji who gives a shit. It's on device. And you can say, ah, they'll get rid of it in a couple of years, just like they got rid of the touch bar on the laptops. You know, this is their little flirtation of a new thing to add. They're not going into it as hard.

They're like, and shitifying just a little bit less, but even they are still doing it. And I also hate the AI features. The summaries are hilarious. Oh, I mean, I kept them on simply because it was so much fun to take a screenshot of the summary and send it to the person who had sent me the text message to see how incorrect it was. Once it told me that one of my friends said that they were horny. They were not horny.

What were they? I'm not going to tell you what they actually were that's violating their privacy. Okay. But they were not horny. It was like on my way and horny was the text message. Okay. They had been talking about horny was the text was the text. OK, they had been talking about horniness separately. Right, right, right. Because we're comedians. We're talking about shit all the time. But like these it's it's insane.

I think and I will actually say I agree in that it's less offensive, but still it's forced in there. If you click the wrong button, you hit the genmoji playground. There's like a two second lag while it brings up a full screen fucking thing and you can generate a seal giving you the finger. I don't know why, what human being is using this and being like, finally? It gives you these suggested answers that are all terrible. And it forces, it tries to push the message onto your phone.

It's invasive. And now the new Siri just does not work. It's so funny that they were like, what if we had Siri with AI in it? It's like, mate, isn't that what you said when you bought this company, when you bought the company that made Siri, but didn't you say that was AI? Then what is this? It's Siri, but it sucks. And you need a new phone.

And sometimes it will say, I don't know, do you want to use chat GPT? Which will take 10 seconds from you asking to actually getting anything and maybe that answer. What if my phone was slower and worse? Finally, the power of AI. Jason Cosper I think that the voice agents, the voice assistants are a really good analogy for all of AI generally, because we've had them for 10 years now. And we've been seeing ads with Samuel L. Jackson, you know, like using Siri to help him cook 10 years ago. People got Siri, people got Alexa,

and they found it largely useless. Use it for two things. Play some music, set a timer. That's it. Wake me up at 8 a.m. And then when I'm in my car, I'll go play my station or whatever if I want to hear the shitty algorithmic station Yes, I don't have time to pick something manually or you know, play some prints

That's it and they still lose Alexa lost to Amazon billions. Yeah, they've never made any money They thought that it was going to be like Amazon specifically thought it was gonna be like their big kill shot It was gonna be everything that they did was these assistants, and stories started coming out two years ago, oh yeah, they've only lost money on it, there's never been a purpose, it's for children to ask for a song to be played and annoy their parents, and to say, wake me up at 8AM. ALICE What's great as well is that Amazon has been desperately trying to do a generative AI version, but they keep having to stop launching it because it's really slow and does not work.

I love the future, man! I love this. And the thing is, I think regular people understand this better than the business people. Regular people are just like, what does this shit do? Why do I care? And to your point, it's like, I already have Alexa and it only kind of works. Like if you try and make Alexa do anything, Siri same deal, more than like reminders, like they were, oh, you can buy stuff with Siri, with Alexa.

No, no, no, that doesn't sound safe at all. That sounds actually really bad. And it never was not that you would. No one walks around is like, hey, computer, get me milk. No, you want to choose the milk. You know, the best thing I ever did with Siri? You can't do this anymore. But when it started out, they had a hookup for Wolfram Alpha, you remember?

Oh yeah! Which was like... And that rubs. For folks who don't remember, it was sort of like a computational engine, you could sort of put in natural language queries, it could do really complicated unit conversions and stuff like that. Pretty interesting piece of software, It's very pre-AI. It still works as well.

But Siri used to hook into it. And so you could ask it weird questions. I would ask it, what was Adolf Hitler's height in kilometers? And it would say Adolf Hitler was 0.001 kilometers tall. It was like a fun party trick because in the database that Adolf Hitler's height and it could do a unit conversion.

Doesn't do that anymore. But I used to ask it what Adolf Hitler's height was in light years. And it would tell me, oh, what a fun little dumb ass thing that this thing can do. It's smart in a way I didn't expect for something useless. Yeah.

And you're now describing how people use chat GPT as well, which is crazy. Right. Not even being facetious, that's the kind of shit they're like, oh, what if it did that? Oh, okay. I will now move on with, forget that this exists. And also this isn't AI, it's not autonomous at all. Ah. Like, we're describing things that existed 10 years ago that worked better. Wolfram Alpha works better. Still around, and it still works better than chat GPT. Yeah. Wolfram Alpha works better. Still around, and it still works better than chat GPT. It's also frustrating looking at the actual things we use get worse, but the thing that loses a bunch of money that sucks that they must have us use, that sucks as well, but they're claiming it's

the future. They're not building the future anymore as well. And I think maybe that's what this moment says the loudest. We've been sitting here this whole time being like Silicon Valley's there and they have all this money and we justify them having this wealth and this marketing arm and like all of this media and all of these things because they're going to make the future, right? They're going to make cool shit. Where's the cool shit? The cool shit that they're trying to... They treat us like hogs. They treat us with complete contempt. They're trying to warm up stuff they already did, but it works. Similarly or worse, and also it costs more money than anything should and destroys the environment and steals from everyone, but maybe one day

it will be something else that they can't really describe because average general intelligence, artificial general intelligence, whatever you call it, they're claiming it will be like a super intelligence, right? Human beings, we don't understand how intelligence works in the brain yet. How the fuck are we meant to do it in the computer? It's ridiculous. I feel like I'm going insane sometimes. Because outside you read these AI headlines like AI will by 2030 be a trillion dollar industry. It's like, what? Why? What are you fucking?

Yeah. What are you talking about? Okay. We've been talking for 90 minutes. We got to wrap it up. We could go on about this forever. No, no, I get you. You have a very negative view.

Is there anything that you feel positively about right now? No, I actually do. So in the last year, working on this, working on the show, Better Offline, I only came out positively about right now. I actually do. About the tech industry. Yeah. What is it? So, in the last year, working on this, working on the show, Better Offline, I came out in February last year, writing my newsletter. When it started, people were like, you sound like a nutter. Why are you so mad at the computer? Yeah.

Now, especially towards the end of the year. I just say I love the way you say the word computer. I can listen to you say computer in your... What is your specific accent? It is a mixture of West London, I grew up in a roughen out Wyrmwood Scrubs in London, and then I went to a private school, and then moved to America, so my accent's just a fucking mess. RG You're like code switching on me, aren't you?

Sometimes you say computer. KM I can't control it. I cannot control it. RG Because you say computer in this way where I'm just like, I wanna, I'm imagining you holding a baseball bat and you're about to destroy it. KM Well, the thing I say is I will never a baseball bat and you're about to destroy it. Well, the thing I say is I will never forgive them for what they've done to the computer. Because that is the thing.

What gives me hope is that the people I love the computer, I grew up loving the computer. And that's the thing. I love the computer. I love being online. I love all this stuff. I love my gizmos and gadgets. I love my batteries. I love the fact that we actually have cool climate solutions.

I love electric cars, just not the Tesla ones. I have a Volvo. It's a boring car. I love it. Yeah. actually have cool climate solutions. I love electric cars. Just not the Tesla ones. I have a Volvo It's a boring car. I love it. Yeah, I genuinely love being online. I've met people I love I've met people I hate I've met all sorts of people I've grown a business. I have a learning disability called dyspraxia I don't think without the computer I would have been able to live but in the last year And I was not a very social kid either. So this is how I made friends. Yeah, this is how I keep in touch with people And I was not a very social kid either, so this is how I made friends. This is how I keep in touch with people. I'm in the same boat, you know, I got a cable modem connected into my bedroom when I was like

15 years old and it opened the world to me, right? Because I was a shy kid, I didn't have friends, but I like built a life for myself online and that's why I have the career I have today, it's why I have the life I have. I'm 11 years old with a PCMIA, MCIA card. It's like a 33.4. I was a nerd. Anywho. But this allowed me to make connections that are necessary to like live. And it actually told me, wow, the world isn't this shut off horrible place. London. And it allowed me to do all these things. And doing this show for last year and writing this stuff, the people that have reached out to me have really told me a few things. One, they're fucking furious at tech. But two, I think most people love the computer. I think most people

really love tech. And also most people can understand what these companies are doing and why they're doing it. They've just never had it explained to them. And something that I really, like the most important thing to do is I want to give people the language to understand what is being done to them by these companies. Why Meta is storing ads at them every second, why the algorithm doesn't seem to do anything to help you. It's because Facebook is helping Meta make money. And they know that Mark Zuckerberg, the CEO, Andrew Bosworth, the CTO, Naomi Gleit, and Javier Olivan and the rest of them. The growth team is why. These seem like obtuse concepts, but most regular people, teachers,

I hear from a lot of teachers, fucking rucks, they all say the same thing, which is, I finally fucking get it. I actually believe that there is a consumer awakening happening against the tech industry. Not because the tech industry as an idea is bad, but because what the tech industry has become is so vile and so antithetical to what Silicon Valley was built on and the reasons that we use tech to connect and empower human beings, to make ourselves rise above what we are as meat suits. That's the theoretical

thing. There's meant to be something inclusive and diverse behind the whole thing. And I really feel like people are awakening to this. And the reaction to a lot of my work over time has gone from people being like, this guy's just angry to, wait, I see this shit in my life all the time. I hear sports gambling ads everywhere. I have apps that seem to change the buttons in random places. I have all of these problems that are happening to me. And every, I think every week I get like 50 different people who email saying, I thought it was just me and it isn't. I think every week I get like 50 different people who email saying, I thought it was just me and it isn't. I think consumers are waking up to this and I think that pressure against the tech industry,

it's going to take time, will change things. And what gives me hope is that people really have come around to this idea that they do actually love the computer. They love the internet. They love what it could do. They don't love what's being done to them by these companies. And maybe once something is done to us by the federal government, now that it is run by the apotheosis of the tech industry in the form of Elon Musk,

maybe that'll I mean, we're in a place where we're talking about, hey, maybe after the forest fire, green shoots will sprout. Not great to be living through the fire. Certainly being here in Los Angeles, we have some experience with that. But, you know, that that that's not a great world to live in, but it is at least like a, you know, a blue sky on the horizon, perhaps. And I think the people are learning how to express that this pleasure with tech. You're learning about the incentives. They're learning about the incentives.

They're learning about the people behind it. And I really think there's something very powerful about knowing who, like someone like Prabhakar Raghavan ruined Google search, for example, and I ruined his SEO. Because that's the thing. You search for his name, you find you. You cannot avoid me. But that's the thing. People knowing these people's names and understanding the relatively simple way they're fucking us all, I believe that will the thing. People knowing these people's names and understanding the relatively

simple way they're fucking us all, I believe that will change things. I love participating in it. Well, thank you so much for coming here to spread the message to us today. Where can people find your newsletter and your podcast? So you can go to betteroffline.com and where's your ad for the newsletter, but betteroffline.com has everything, all your links. You can find everything there. Thanks so much for being here, man. It's always wonderful to talk to you.

Thanks so much. Well, thank you once again to Ed for coming on the show. If you want to support this show, head to patreon.com slash Adam Conover. Five bucks a month gets you every episode of the show ad free. For 15 bucks a month, I'll read your name in the credits. This week, I want to thank 90 Miles from Needles, Aaron Matthew, Andrew Harding, Alaska, Amy Thorntron,

David Snowpeck, Eric Karlskin, and Scooty Chimkinuggy. Thank you so much. If you'd like me to read your silly username at the end of the show, head to patreon.com slash Adam Conover. If you want to come see me on the road in places like Providence, Rhode Island, London, Amsterdam, Vermont, head to adamconover.net for all those tickets and tour dates. I want to thank my producers, Sam Roudman and Tony Wilson, everybody here at HeadGum

for making the show possible. Thank you so much for listening, and we'll see you next time on Factually. I don't know anything. I don't know anything. That was a Head Gum podcast. Hi, I'm Caleb Herron, host of the So True podcast, now on HeadGum. Every week me and my guests get into it and we get down to what's really going on.

I ask them what's so true to them, how they got to where they are in life, a bunch of other questions, and we also may or may not test their general trivia knowledge. Whether it's one of my sworn enemies like Brittany Broski or Drew Fualow or my actual biological mother, Kelly, my guests and I are just after the truth. And if we find it great, and if not, no worries. So subscribe to So True on Spotify, Apple Podcasts, Pocket Casts, or wherever you get your podcasts, and watch video episodes on the

So True with Caleb Heron YouTube channel. New episodes drop every Thursday. Love ya. Hey, it's Nicole Byer here. Let me ask you something. Are you tired of endless swiping on dating apps? Fed up with awkward first dates and disappointing hookups?

Girl, same. Welcome to Why Won't You Date Me? The podcast where I figure out love and how to suck less at dating. Each week, I get real with comedians, friends, and celebrities about their love lives. We swap dating horror stories, awkward hookups, and dive into the messy and wonderful world of relationships.

I've chatted with amazing guests like Conan O'Brien, Whitney Cummings, Sarah Silverman, Trixie Mattel, Tiffany Haddish, and so many more. So whether you're single, mingling, or boot up, there's something in it for everyone. Tune into Why Won't You Date Me With Me, Nicole Byer, and discover insights that might just save you from your next dating disaster. Listen and subscribe on Apple Podcasts, Spotify, or wherever you get your podcasts, and catch full video episodes on YouTube! New episodes drop every Friday!