Trashfuture x Better Offline: Goosed In The Shell

28:08So from Kakashi, great pseudonymous analyst guy I follow. So, OK, his dad was talking about him and 60 year old Masayoshi son. he was super wrestling his older brother and losing badly, but he refused to stop. When his father finally pulled him away, the boy's eyes were, in Mitsunori's words, his dad, like an animal's, a wolf's eyes. I thought to myself, this bastard is not human. Doing hype edits for your son-like play fighting is really funny. I just really like the idea of him just getting his ass kicked when he's six. And he's like, he just turns around, he just has like demon eyes. Like, oh.

28:46world's first person to have laser eyes in real life um look here's the there's the other thing right we're talking about masayoshi son's life the the the presentation concludes with what used to be his 50 year life plan but now is his 60 year life it's so good so his 50 year life plan was get acknowledged in your 20s finance a war chest in your 30s um in your 40s take on a challenge in your 50s. Complete business. I'd love to do that. Just any of them. Just any of them. Like an 18-year-old girl's Instagram post. Take on a challenge. Sure. Remember Ice Bucket Challenge for ALS in 2011? Take on that challenge.

29:27Done. Finish the decade. 50s. Complete business. Okay. So, complete all the business. What could that mean? Yeah. 60s. Hand over business to next generation. But no. fucking crossed out no crossed out this is the 60 year life plan 60s crossed out hand over business to next generation gone it's not replaced it now just has a 70s value which is realizing asi or artificial super intelligence a similar a term that's sort of this is like not a new hubris right you could have done this at any point in human history there were guys back in the day whose life plan was 20s, Unite Southern Song Dynasty.

30:1130s, Conquer Northern Song Dynasty. 40s, Establish Mandator's Son of Heaven. 50s, Alchemy. 60s, Alchemy. 70s, Never Die. I feel like it's missing a little detour. I feel like you need a bit of a side quest to really make this fun. You should 55, start a band. I don't know. There's so many... Yeah, 60, learn to ride a motorcycle. No, in this case, it's really just like the most obvious thing of like 20s, conquer entire known world. 30s, weep because there's no more worlds to conquer. 40s, build giant statue. 50s, start reinforcing statue at legs. Yeah, I also want to be clear. It's really funny to do this insane hour and 50 minute long presentation where you're like, slide four, JPY, one quadrillion.

31:03and they'd just be like, and I'm not retiring for at least another decade. Fuck you. I don't think that I live in a functional enough economy or society to have a life plan. Like, my life plan is like 20s, fuck around with mental health and drop out of uni, 30s podcast, 40s podcast, 50s podcast, 60s climate wars, maybe? Yeah, woke battalion in climate wars. She thinks the climate wars aren't going to start until she's in her 60s fucking idiot so this is but this so basically this is this is the story of someone who has i think we come back around to this i think it's very important someone who has been driven to mania about this thing but who happens to be able to force that mania upon everybody else yeah and you know one of the things i wanted to dig a little deeper into actually is his claim that artificial super intelligence is coming artificial super intelligence artificial general intelligence, I think are used pretty interchangeably.

32:02He uses artificial super intelligence. That's basically the same thing. Yeah. And so basically, Masayoshi-san and Sam Altman are now... So Sam Altman's guys, excuse me, told Masayoshi-san probably in the, we just need another, can you take out a loan to give us a loan? Said that an AI model is designing the next future AI model at OpenAI. So like San said... You can't see the model or anything to do with the model. Read this philosopher we got in to talk about the fucking constitution or whatever. It's the Dunst economy. It's just like, they find a new key to Jingle and it's like, well, why don't we invent, I don't know, make the model do something new?

32:42It's like, what if we just made the model do that? Or if we just fucking, you know, like the model could do the model stuff and we could just sit here and take the cash, Masio. And like, it's the laziest thing in the world. It's like if any other situation, it's like, I'll have someone else do it. You'd probably be like, they'd be like, well, no, you've got to do something. I gave you$40 billion. Like, you probably want to do something with it. But Masayoshi-san's like, hell yeah. But what's funny is Masayoshi-san, Sam Altman came to him in 2017, I think, and asked for$10 billion. Masayoshi-san said yes.

33:18But then Sam Altman just kind of walked off and took the money from Microsoft. off. And that drove Masayoshi son insane. He was like, oh my fuck. I missed out the one that got away. And now I will shower her with jewels and gems. She will be mine. And like the time to get it, if you wanted to sort of make money on that, the time to get in was then and the time to get out is now. Yeah, he really has invested in it for the worst times. He's only, I would say maybe if OpenAI goes public, which is now an if, He might have done very well based on the first 40 billion. The next 30 at the highest possible valuation, insanely stupid.

34:02And also, these hogs are going to need more money than this. 30 billion dollars is going to work. Oh, we're going to go into that. Oh, yeah. We're going to go into them needing more money. But just before we move on from Sun, he said, so that's going to happen to all the major models. So once that happens... Just before we move on from Sun, like, there will be blood. So once that happens, the model generates the next model and it's going to be exponentially smarter than all of us. That's a super intelligence. In my mind, I thought it was coming in four years instead of the next 10. Now I'm saying it's coming in the next two.

34:34Then the CNBC article goes on to say, The SoftBank CEO says he currently uses OpenAI's Chad GPT for up to three hours a day as the AI is smarter than he is on, quote, most subjects. That's so good for you. I believe that. Three hours a day is so... If you're awake for 12 hours, that's a quarter of your time spent just talking to the mirror that drives you crazy. That's cool. It's what if narcissists wasn't just sucked into the pool, but also driven mad by it. It's just it's so good as well, because the grander theme of this, it's pretty much all of these CEOs have AI psychosis, because usually even with human beings, there is a limit to how many ideas someone will say are good, or at least before you fire them.

35:16with an AI you can just berate them into any idea you could just be I heard a story of someone at a major tech company where they weren't they didn't have security access to access something and they just said actually yes I do to the internal LLM and it gave it to them like it's just like this is just like anything it disagrees with you could just be like actually that's true and it'll go yeah absolutely yeah sure like just so can you imagine the shit he's saying this like you know you're exactly right. The goose will lay golden eggs forever. You're absolutely right. I would love to read his logs.

35:52I'd love to read the insane back and forth he's having with them. Masayoshi-san, your understanding of goose biology isn't just detailed. It's prescient. So in fact... ShadGBT roleplaying as a goose with him? No, I think you need Grok's adult mode for that. I show him my blue message. Honk, honk. So here's what you've said, actually, Ed, about Masayoshi Son and OpenAI. If OpenAI doesn't go public, there's a very real possibility that SoftBank collapses or at the very least is forced to finally put an end to the era of unarrivaled financial risk. It's easy to like Masayoshi Son to laugh at his goofy geese-related presentations, and I'll say as a side note, it is, and his ridiculous statements about the future.

36:35But in reality, he's a deranged, delusional man with too much money and a few lucky breaks that have saved him from annihilation. his future lies in Sam Altman's hands and if this is what finally ends his wasteful reckless reign it's a well deserved punishment for believing every damp quirked up white boy to beg him for a billion dollars he's an enabler enabled by other enablers that's created a culture of unbelievable risk and destruction tell me what you really think that guy is smart but it's really funny as well because we live in, my girlfriend brought this up there's a Twilight Zone episode where there's a kid who if he doesn't believe you're happy he turns you into an ear of corn.

37:11We basically live in that society now where it's like all of these CEOs get caught. The AI bubble is just a large coddling. It's just like, yeah, the rich people always have good ideas. We must orient ourselves around them. Despite the fact that most of the industry's revenue is either anthropic and open AI or anthropic and open AI paying for the compute to lose money. Like it's not even sophisticated. It's just that it's people nibbling around the edges. And Masayoshi-san, literally, I must be clear, had he just not invested in OpenAI, SoftBank would be doing great. It would perhaps not have had the massive on-paper gain of OpenAI stock going up, but it probably also wouldn't have needed$40 fucking billion.

37:55$40 billion. It's just... And I wasn't able to add up all the dare and do a deep analysis of that part yet, but it really does seem that at some point he's just not going to be able to refinance alone and run out of money. It's just like a teenager with a credit card, except it's$150 billion of debt and there's Goose involved. And what SoftBank is doing via OpenAI right now, this is my very refreshingly candid telecoms.com article, is that it's going to take on additional debt in order to build data centers, more data centers in the U.S., tens of billions of dollars in AI data centers in the U.S., and giving more money of that to OpenAI.

38:40So the data center project is called, I believe, SB Neo. Yep. And they are going to, again, say, okay, we're going to build 10 gigawatts of data center, right? But quick question. We talk about data centers a lot, you and I. We talk about data centers a lot on this show. You and I talk about data centers a lot. Do you have any of the enormous supply chain bottlenecks that have largely prevented data centers from being built, but yet let alone the sort of, let's say, lack of demand from outside the industry for their services. Are any of those bottlenecks alleviated? Can you get a gas turbine right now?

39:18No. And the thing is, as well, is just data centers are taking like 18 to 36 months to build. So him saying we're going to build five or 10, he did a thing with France as well. It doesn't bear repeating. It's something with Emmanuel Macron. and he's like, they're building 5 gigawatts there, 10 gigawatts here. You may as well be talking about the year 2090s. Because just the speed of construction, put aside any thoughts I have about the AI bubble, the speed of building these things is very slow. And that's before the power comes in. It's just the power doesn't exist either. Again, if you took away the money, this person would sound fucking insane.

39:57Well, they do right now, but you would laugh them out of the building. Like if just a random guy was like, I'm going to raise 10 gigawatts. So a trillion dollars, because a gigawatt is now$100 billion. Trillion dollars. I'm going to raise a trillion dollars for a data center. And then one company is going to rent all of it. And yeah, that company loses money, but like they won't. And anyone else, you'd be like, you sound like a moron, man. But because it's Mashiyoshi's son, because it's O 'Brien Sam Morton, because it's Dario Amadei. You're like, oh, yeah, yeah, absolutely. Yeah, absolutely. that makes perfect sense.

40:30These are the guys. They're the guys. They're the pharaohs. All things have... With inequality accelerating, with the bets, the actual bets on the future getting better, getting smaller and smaller and smaller, and concentrated on a smaller and smaller number of big hands, if you like, you get a kind of... You sort of rebuild like a pharaonic system of kingship from first principles, where, you know what? You can just be like, you know what? We're monotheistic now. Egyptian polytheism is done. We're going to try monotheism for three generations. But the problem is also, being in a pharaonic system like that, with sort of this much concentration of...

41:09With this much allocation of productive capacity to madness is quite... I'm going to surprise you, quite damaging. Because the question of, can you get a gas turbine? That's super important for lots of stuff other than AI data centers. It's just, these data centers are, in some cases, because of political pressure, like in the UK, able to skip the queue for certain kinds of planning or grid hookup capacity. Or, you know, they're just they're clogging up the order book for these things that will just get deployed to these data centers that we're going to explore if people are, let's say, using enough and profitably enough to make them viable.

41:46And all those spiking infrastructure demand has now is now going out to other. it's like spreading through other pieces of like infrastructure, like hard to build infrastructure equipment. Really? Quick question. Yeah. Has anything else happened recently to maybe fuck up like global infrastructure at all? Something in the World Cup maybe? Yeah. The England world people watch the World Cup. Yeah. I kind of feel like the streets of Hormuz fucking up the supply chain for this and everything else is sort of like one point on the the sort of trifecta the palace logo of we are really fucking bad at long latency stuff as people um so like it's i'm sure all of the consequences of this will just be fine for an extremely like fragile sort of delicately balanced goose fucking portfolio but that actually leads me to an important point which is let's take the AI people at face value.

42:53Don't know why we're doing that, but let's do it. They are saying, because based on NVIDIA CEO, Jensen Huang says trillion dollars of GPUs, visibility sales through the end of 2027. It's about 400, 500 billion dollars a year worth of AI compute revenue. Just needed to pay for that. Put aside my thoughts, just like you need that. This is not a hyperbole. This is like you need that money. The demand doesn't exist. But on top of that, they are just saying, yeah, the AI bubble is going, the boom, even in their mind, is going to just keep going. We're just going to spend$3 trillion on this every year.

43:29And when you ask them how that works, they're just like, the money will be there. How are we going to substantiate it? The money will be there. The money will be there. It'll happen. The demand's there. And you're like, well, there's compelling proof it isn't there. And they're like, well, it will. And it's a great way of magicing away these long latency problems you bring up Nova, like climate change, for example. We were talking earlier. It was like, well, the warming that's happening now was from emissions from, I don't know, 40 years ago, whatever, right? Like this is all baked in. Yeah, if you stopped like any CO2 emissions like today, we would still be logged in on like decades of warming, which is really going to fuck up my life plan, I think.

44:07Yeah, that's good. Climate wars in the 60s? Come on. Yeah, you're going to have to add another decade. But their answer to this, the way that the AI comes and forestalls like actual good long-term latency planning is it's like, well, the AI is going to be smarter than us because it's going to be super intelligent. So we won't have to solve long latency problems. We don't have to worry about the fact that - I love that. Well, that's a relief, to be honest, because I hate solving problems. And to be honest, in my 60s, I'm hoping to get into like alchemy. So here's another stat I picked up, right?

44:39Which is it's no longer, it's not just gas turbines anymore. High voltage transformer lead times have gone from 12 months to 48 in some cases. So good luck. So for example, at a time where we need to do, let's say an energy transition or a time when say even just in the UK, right, national level grid upgrades are more or less needing to happen in the next 10 years or it's going to be like rolling blackouts. All the high voltage transformers you can order are going to the God machine, which will then solve the grid problem some other way. I just love, yeah, dude, screen light. We'll work it out when we work it out using the working it out machine.

45:20You just keep giving it money and then at some point, you know, you work it out. What are you being a quack about? Come on, man. Just give me 10 billion. You're being lame. You're being lame, man. Come on. And also, it's like, I don't think it's really funny that this fear is always, oh, AI is going to be a paperclip maximizer, right? The entire, the whole world and universe and all matter, we made into paperclips by an AI. But if you look at what we're doing with data centers, and we're going to go on again to the profitability of them in a second, because some stuff has happened there, is you don't need the AI to be a paperclip maximizer because capitalism as a system is already maximizing the paperclips in the form of the data centers.

46:04We're cannibalizing useful things to build this. You can't upgrade the grid anymore because all of the Transformers are being turned into effectively paperclips. Yeah. And I know that's not even touching Databricks, which is another N-scale type snafu that got uncovered by a similar situation. I don't know that you can be saying that, Riley. Sorry. Sorry. Check out No Gods No Mayors. But we'll go into that on Thursday.

46:40Support for today's episode comes from Square, the business platform that helps sellers become neighborhood favorites. Whether you're gearing up for a busy season or just trying to keep up with everyday demand, Square keeps your business running smoothly, from payments and POS to online orders, inventory, staff and more, all in one place so you can focus on your customers, not your to-do list. I like Square a lot as a customer, and from talking to various business owners who use it, everybody seems really happy with how easy and intuitive everything is. Business owners don't have to juggle a bunch of tools, and their hardware and software are designed to be easy to use, so day-to-day options feel a lot simpler.

47:14With three clear plans depending on the tools you need, sellers who use Square see 9 % higher sales on average. Square helps you run your business more smoothly, bringing payments, operations, and insights together in one place, so you're ready for whatever's next. Right now, listeners can get up to$200 off Square hardware when you sign up at square.com slash go slash better offline. That's S-Q-U-A-R-E dot com slash go slash better offline. Get started with Square and build a setup that works the way you do. Run a business and not thinking about podcasting? Think again. More Americans listen to podcasts than ad-supported streaming music from Spotify and Pandora.

47:55And as the number one podcaster, iHeart's twice as large as the next two combined. So whatever your customers listen to, they'll hear your message. Plus, only iHeart can extend your message to audiences across broadcast radio. Think podcasting can help your business? Think iHeart. Streaming, radio, and podcasting. Let us show you at iHeartAdvertising.com. That's iHeartAdvertising.com. We've all been there. We've had seasons that have shaken everything. The question is, what remains when the shaking stops? Hi, I'm Gary Valenchano. channel. And on my podcast, Unshaken with Gary Vee, I sit down with amazing personalities and we talk about the seasons that nearly broke them.

48:39The lows that carried them to their highs. Listen to Unshaken with Gary Vee on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts. Bill, why is my eye twitching? Does this mean I'm in perimenopause? Maybe I have adult ADHD. I need to look this up. Where's my phone. Juliana, your phone is hot to the touch. I think it's asking you for a break. You know what? Maybe I should just ask ChatGPT. Or maybe we can ask an actual human. Yes, like a couple sex therapist. Whoa, that escalated quickly, but okay. Because I have questions. We have questions. And I bet everyone has questions.

49:17Like, is it normal to sleep in separate bedrooms? We do that. How about this one? Is bribing your kids bad parenting or just negotiating? Oh, and I still do need to know, why was my poop green that one time? Hypothetically, right? Okay, well, instead of letting the internet guess, we've got actual people answering these exact questions. And laughing with us has got to be better than spending three hours down a rabbit hole online. Listen to Bill and Juliana. The podcast. On the iHeartRadio app, Apple Podcasts, or wherever you listen to your podcasts.

49:54Anyway, I wanted to talk about some profitability stuff, right? As we mentioned that OpenAI's going public is now an if in that earlier episode of Is That Good that we played. So I actually want to quote from a friend of the show, Mike Isaac. The Wall Street Journal reported that the company planned to go public by the end of 2026. OpenAI said this month it had filed confidential paperwork with securities regulators to kick off the process. A trillion dollar valuation of the public market, which would exceed the market capitalization of Walmart, would be staggering for OpenAI, a startup that is not believed to have turned to profit and is aggressively spending on new data centers.

50:27OpenAI's advisors presented company executives with the option of waiting till 2027 or lower the target valuation for a quicker IPO. Mr. Altman responded that any change to a trillion dollar valuation was a non-starter. I assume this is me editorializing because he's in a pissing contest with Elon Musk. I mean, I think it's also the, so their last pre-money, so before the cash they actually raised valuation was like two, sorry, 730 something. So a trillion dollars would be a 30 % premium ish. That's not a lot. And the fact that the bankers are saying that they couldn't get that because that's I'm reading between the lines, but I mean, that's what's happening.

51:08They clearly talked to some advisors who were like, you know, no, that's a negative. That's worse than people are giving this credit for. If you can't get a 30 % premium, it basically means all of your private investors at the last round fucked up because you're not even getting like a good venture capital return would be like 3x. You're not even getting just to it. Like it's, you're genuinely making nothing here. And also they're saying they couldn't list at 1 trillion. So they probably wouldn't be able to list at much above 800, which is fatal. I mean, just, and this is fatal for SoftBank as well, because SoftBank needs OpenAI to go public.

51:48But it's just like these things, the reason we keep having bubbles is because when stuff like this happens, the media and a great many different people immediately move not to go, hey, that's a problem. We should fix that. Or like, hey, we should be worried about the company that's become the single focus of everyone isn't going to make it. Instead, it's like, no, no, no, let's rationalize this bad boy real quick. Let's enable the enabler. We need to make sure we keep this. And it's just completely nonsensical. And to be clear, when this explodes, I'm going to be ripping holes in anyone who tries to be like, how did we not see this coming?

52:29You didn't look is the big thing. Well, the main, because the main, the thing that is at the bottom of this is always the unit economics. And what I think is very telling is that open-air executives, in talking about the IPO, said, well, the company is moving in the right direction because more than 5 million people use codecs on a weekly basis, the company said. Now, more than 5 million people using codecs on a weekly basis, pretty good, which means I assume they've solved their unit economics problem and that's not costing them a huge amount of money. yeah they've definitely avoided they've gotten out of the sort of thing that they were doing of just going hey ai is free until everyone uses it and then they can start charging what it actually costs right right yeah and what's great about that is um they did that and they were like okay and now they jack up the price and within one month they're like oh uh everyone's like we got to limit the cost of this just immediately and by the way on a 200 buck a month subscription to chat GPT, you can burn$14 ,000 worth of credits, of tokens.

53:31Jesus. It's so good. It's, I mean, so this is the thing. We've had everybody panic, right, and go, oh, God, it's expensive now. We got to, like, maybe think twice about using it. They're still charging a fraction of what it costs, just a slightly smaller fraction, a slightly larger fraction? Yes, exactly. And amazing. And also, instead of it costs like, I have$14 ,000 for a dollar. You can set that on fire in front of me. What they've gone is it now costs$2 to do that. And everyone has. Well, what's happened? What's happening is it's a little different from that where it's the enterprises are having to pay the actual per million token rates.

54:14So they, and they, this is, it's really, it was actually a very funny punchline set up because they, all of the business idiots who've been using AI who have got AI psychosis told all their people because they were all able to burn as many tokens as they like, use this constantly. Use it for everything. If you don't use this, I'll kill you. I will shoot you with a gun. And then the moment that they, because in, I think it was March or April, Anthropic and OpenAI were like, okay, you're going to have to pay on a per million token basis. They, within one month, were just like, like John Major style, like, I've got austerity measures.

54:47We've got to be very careful with the cost of AI. We couldn't possibly. And it's just so funny because it's like, even if they cut the price in half, it's not clear if that would actually fix the problem because you can't measure the ROI of any of this. so it's just a certain degree of like great now it's half of half of a thing that we still can't measure so we probably have to cut that too it's just it's truly ludicrous it's genuinely i mean it's genuinely hilarious as well i mean the uh i don't have you have you heard of the new plugin for claude uh called caveman oh yeah where it's just like it just gives you very simple single sentences.

55:26Yeah, this is a plugin that is useful to let's say be austere with tokens by saying, okay, well, maybe we can save a little bit of money on our token budget if we make Claude be kind of rude and abrupt. It's like, I don't think that says anything good. If that's the solution you're using. It's kind of funny how much of the cost of this was making it glaze of the user, at least. So in late June, There has been this huge, it's not because this is not just prices rising. There's been a huge sell off in names across the stock market related to AI. NVIDIA, SpaceX, SpaceX already is down like hundreds of billions of dollars from its post IPO pop.

56:10So if you thought you were going to get on the Elon scam, you got hosed. But this is also related to Microsoft, Alphabet, Amazon and Meta all suddenly becoming like high CapEx heavy industry companies more or less overnight. Right. So the customers of these companies, right? They're the ones who are supposed to provide the money to help them lose money more efficiently, still aren't able to capitalize on them. Consulting firms aren't using it to get better. Banking companies aren't using it to get better. You're not using it to really generate a lot of return on investment. And so more and more people are realizing this, subsidized pricing is coming up against reality.

56:43And that exuberant investment pop that's been happening for the last little while, those names are beginning to go down. In fact, this is from Katie Martin in the FT. The once vaunted Magnificent Seven Stocks, an imperfect but nonetheless go-to gauge of hyperscaler performance, performs slightly worse than the UK government debt. Oh, how are you a worse investment than Keir Starmer? Yeah, that's like at the World Cup. It's like getting rolled by Qatar at the World Cup. Like, come on. You shouldn't be beating us. Can't believe they did that to a bike, yeah? Yes. And I mean, this is a, the Bank of International Settlements is also getting it on this as well.

57:31Yeah. Now suggesting that with the five largest hyperscalers setting to spend over a trillion dollars on AI-related CapEx from 2025 through 26, that the commitments are outpacing earnings and free cash flow, leading some to issue debt to raise additional financing. and that, again, the Bank of International Settlements is now saying this is now, let's say, a significant global risk. Yeah. If only someone had fucking said something. It's so funny as well when the BIS put that out. So like the conservative bank of central banks. And immediately I saw an analyst, this guy called Daniel Newman on Twitter, who is a massive word I won't say, was just immediately like, oh, what was this, a press release?

58:16a press release from the haters. It's like, no, it's a statement from the central bank of central banks. You're a financial analyst, but this is the media behind this. The media is just like immediately ready to just attack any bad sign. And it's very strange. I know it's always been like this, but to see it reach its kind of pornographic heights, I think it's just the data centers are too easy to invest in. The AI companies are too easy to invest in. There's so many places to put your money. And they're so brain damaged from years of ignoring the consequences of their actions that they don't think anything can go wrong.

58:53Masayoshi-san, all of them. I mean, all of the data center operators, same deal. Hyperscalers, same deal. They all think that they have found the mandate of heaven printer. Yeah. And the thing that was that the test is it was this the mandate of heaven is, is this going to do to knowledge work, for example? what containerization did to stevedores. Is this going to do, is this going to be like the great enclosure of the digital commons once and for all? Is this going to be the thing that creates the new, I don't know, six person ruling class of the entire world by fundamentally transforming the economy with this like huge technological displacement of labor?

59:37Like that was the actual question. Are you going to do that? And on day of recording... To which the answer is, yeah, I reckon. Yeah, probably. Sounds good. Well, here's the great thing. And look, this is inferential. This is not deductive. This is not provable. But the Wall Street Journal has posted that big tech has suddenly flipped on the AI jobs wipeout scenario where they've now collected some, let's say, barometers of opinion from the big guys, the Sam Altmans and Dari Amadai's and stuff, surveys. And they've said, hey, wait a minute. It looks like all those jobs aren't going away. A lot of people were still fired with AI as an excuse.

1:00:18A lot of them got rehired in some cases. But this is excerpt one that the Wall Street Journal quoted. A year ago, the message from many business leaders was that AI was going to wipe out jobs. The past month or so, tech CEOs have been striking a more optimistic tone. In late May, Sam Altman said, We've been roughly right on the technological predictions and pretty wrong on the social and economic implications. saying to CNBC, our industry underestimated how much we're going to be able to keep people at the center of everything. But your thing was the replace all human labor machine. Yes. And your replace all human labor machine, what you're saying is it's failed to replace all human labor, at least in the timescale that you need it to, in order to go public at write down a value and you can decide how rich you are.

1:01:03Yeah. I do kind of wonder whether, like, the reason why they don't think of the idea of consequences is because the AI is not telling them that this might be the thing that happens. Like, I know there's that sort of thing where like Sam Altman sort of sounds like he is kind of just, he sort of, his brain is entirely like ChatGPT and that sort of, you know, the guy who sort of drinks too much of the Kool-Aid and becomes a Kool-Aid guy. That's how you make new Kool-Aid guys. Kool-Aid guy values zero. That's right. Presumably that's how it works. Um, but yeah, like even when I sort of think about the boosters of the stuff that like Ed, you were just sort of saying now about like, um, sort of boosters and tech types who are like genuinely surprised when people are like, Hey, maybe this thing kind of sucks or is overhyped.

1:01:50Like to me, it does sort of strike as like, not, not just that they believe that this has to happen. And this is like the inevitable next frontier, because if not, then like, what else you've wasted all this time, you've wasted all this money, you've wasted all this energy, basically kind of doing nothing except creating a vanity project. but like one of the things that feels so evident about this entire cohort is that like they just don't think about consequences because i guess they don't believe it exists because to believe that consequence exists you have to believe that other people live among you and that their lives are important and worthy of some kind of dignity and once you do that when do you think sam when last carried a wallet when do you think do you think he he carries money do you think he do you think any of these people do because this is are they just so separated that the only people they talk to are just worm tongue like figures and chat gpt and one another well i mean yeah i mean it's sort of it's sort of just strike me as like the guy who sort of has never really had to experience any type of friction in his life and so doesn't have to like understand that no there are like parts of the world where you just can't i mean it's also like why they always get confused whenever like the european union is like yeah we're gonna maybe regulate you and they're not like oh the regulation is bad and we can so it's like but why why would you want to regulate us what would the purpose of regulation do like why would you why what yeah exactly why would you want to restrain us like the idea of sort of being held back is like completely foreign to them and so imagine like sort of then sort of trying to explain to them but hey look like there are a lot of societies in the world that like you know don't operate using kind of you know the computer is not the be-all and end-all of like communication you know um people actually like talk to each other there there exists like communities and bar systems and all those types of things like imagine just trying to explain to them in general like hey like people just don't live like you what do they do when they want to what do you mean they run out of money what does that mean you know you know it reminds me of like there was a clip that was circulating today and it's a very old clip it's like an Anthony Bourdain clip where like he's explaining to a bunch of like uh very rich Singaporeans I think that he does his own laundry and one of them is just like well what what do you mean you do your own laundry and he's just like yeah I go to the laundromat and I put the coins in and the machine kind of does it and I and I fold it and I feel quite good about doing it afterwards and like the reaction that this guy has is like you are speaking to him in a foreign language like he has never understood the concept of like touching his own closed before.

1:04:25And this is like, this reminds me so much of just like, what, how all these guys view the world. They got too rich. We got time for one more bit before we have to close. And this one is really interesting, right? Because this is one of my favorite kinds of statistics, which is, how is AI actually being used in companies? What are they using it to do? Is it profitable? Are the pilots successful? And so on. The sentiment change isn't limited to tech leaders. This is from the excerpt again. A survey by EY Parthenon found that the percentage of CEOs who believe AI investments will result in significant reductions in headcount fell from 46 % in January 2025 to just 20 % this May.

1:05:02And even the most credulous worm eventually turns. Yeah. And like, so it just goes back to, well, if the firms aren't more productive and they're not able to be the same level of productive with fewer people, like if these things aren't happening and the thing is starting to cost what it costs and what it costs is enormous and it's still being subsidized. And in order to continue building up the demand that they think that they'll need, even if the thing suddenly goes to 100 as everyone needs it for everything, even if all that happens, they still can't build the data centers. Then it's just like in the grand story that you've been telling, Ed, and that we've been sort of checking in on you with every once and again is that like if the coyotes run off the cliff he's held up the sign and is looked down and has held up the sign that says oh no yeah um or he's opened he's opened up the uh the little um parasol and the boulders falling on top of him this is the moment where he has looked down yeah because that was that like the moment where he looks down that was going to be when the thing has to start costing what it costs right not even making a profit, just costing what it costs.

1:06:17And that's now happened. Yeah. And now all the companies are freaking out that it costs a lot of money. And I think it's just that, so, you'll notice we haven't had a revenue update from in a while. And I have to wonder if that story might change too. But honestly, the biggest warning sign, the biggest sign the coyote looked down is when Sam Altman, when asked about these costs, said on stage, yeah, it's become a huge issue. Yeah, someone should look into that. it's like a it's just like oh all right well and he basically get the exact quote but he was like yeah someone's gonna have to industry's gonna have to sort that out it's just like you are hopefully the big problems hopefully the big black problem solving box will do that i shouldn't check what's in it sell the computer who sam fucking aquaman look um i know we're uh we got a bit of a hard stop today so i'm gonna i'm gonna call it there uh ed joint venture 19 staten islands 17 billion dollars raised and 18 and we have lost the money already so we're gonna have to keep doing this thanks a lot everyone for joining us we've been tf and ed's been better offline and we will see you on our various you if you're listening to this, you know the deal.

1:07:36If it's TF, we'll see you on the bonus episode. If it's Better Offline, you'll see Ed on Better Offline. We should. See you on the bonus episode either way. Yeah, exactly. Thanks, everybody. Cheers. Bye. We've all been there. We've had seasons that have shaken everything. The question is, what remains when the shaking stops? Hi, I'm Gary Valenciano. And on my podcast, Unshaken with Gary V, I sit down with amazing personalities and we talk about the seasons that nearly broke them, the lows that carried them to their highs. Listen to Unshaken with Gary Vee on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

1:08:22I've been hearing for decades that the markets can solve climate change. Today, we have more incentives for market solutions than ever, and emissions are rising. on this season of Drilled Carbon Cowboys, the story of three market solutions colliding in one multinational boondoggle. You gotta give Bruce and the guys credit. They're Republicans. They don't give a s*** of money. Listen on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts. Here at the Happiness Lab, we're serving up some hot takes for the summer. Big ideas that just might reshape how you think about your well-being.

1:08:59Like the radical notion that we should get rid of small talk completely. We talk about current events. We talk about what you do for a living. But not, do you love what you do for a living? Is this your dream job? For more surprising ideas backed by psychological science, check out our new series, Happiness Hot Takes. Listen to The Happiness Lab with me, Dr. Laurie Santos, on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts. This is Michael Rappaport, and my podcast, the I Am Rappaport Stereo Podcast, is unlike anyone you've ever heard. If you're looking for strong opinions about sports, entertainment, politics, pop culture, and whatever else catches my attention, then subscribe now.

1:09:39This kid Jafar Jackson should absolutely, positively get nominated for his portrayal as Michael Jackson. Listen to I Am Rap Report on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts. This is an iHeart Podcast. Guaranteed human. Thank you.