Transcription
I think circular financing should be illegal. I don't think it is an honest way of doing business. It helps establish, it's funny, this era is sold on this level of democratization. What it actually is is helping retrench the largest players except the largest players gambled hard on this thinking that it would magically turn into something profitable and it won't. And they we have the two loadbearing fail sons of Anthropic and OpenAI who must grow because if they don't grow Amazon, Google and Microsoft have wasted their capex. The core business of all of these companies is slowing down. The only thing they have to patch it up is AI and the only way to patch that up is to feed OpenAI and Anthropic. The circle is complete.
Joining me on the tech report is writer of Where's Your Edit and the host of the Better Offline podcast, Ed Zitro. Thanks for joining me as always.
>> Thanks for having me.
>> So, there's two big things I want to talk about this week. There's the Nvidia VP of applied deep learning who admitted that AI is more expensive than a human worker, which obviously throws massive questions on the whole the whole idea of the AI buildout. But I want to start with the news that OpenAI failed to meet its endofear 2025 user targets as well as what I think is at least a few monthly revenue targets. Open AAI called this I think it was ridiculous and clickbait even though it's true.
>> Um what does what does this mean for Open AI?
>> Here's the thing. The reporter of the Wall Street Journal who did that story is a guy called Berber Jim. Uh he is not a clickbait guy. Berber is not that kind of guy. He doesn't do stories like that. And his stories are written with a a calm countenance. They are not fluffy at all. And indeed, I would say there are times where he pulls punches. Not as an insult, but he's a fair reporter. So, this is a real story and it's a real problem because the specific thing that Sarah Frier said was that she was afraid that revenue growth was not there and that it would not keep up with their compute commitments and they would not be able to afford them, which is a pretty big deal when OpenAI represents 70% of Microsoft's AI revenue. And so, I think the magic is gone. I think that it's kind of slowing down. To be clear, if OpenAI wasn't losing what I probably five to 10 bucks for every dollar they made, it would be an incredibly high revenue company, two billion a month in revenue, if you believe that, which I have questions on. Point is, it's partly I think anthropic and others, but I think it's also partly not everything grows forever. Not everything can keep growing. And Open AI, I think the number is they are going to make 600. They claim they claim they're going to make $673 billion in revenue and they going to have $852 billion in burn by the end of 2030. They can't afford to have revenue slow. So there's been a charm offensive with Sam Orman going online and saying he was drunk. That's a real thing. Which is one of the saddest things I think I've seen in my life. He was like, "Yeah, I've been drinking." Go weird uncle energy. I've never really credited him with uncle energy, but he's got it's very strange. Divorced uncle energy, but he's going out doing the charm offensive on Twitter, trying to take advantage of the fact that Anthropic keeps crap in their pants. Their servers keep going down. Uh they're having they're spending more money, spending more tokens using Opus 4.7. There's issues with clawed code still. But nevertheless, OpenAI is in a bind because it's not just that growth is slowing. it's that they cannot afford it to. They have what I think I read recently that they've agreed to 10 gawatt of compute capacity. Well, not that really any of it's built, but that's 100 to$140 billion of revenue a year for those theoretical compute providers. OpenAI can't afford that. Most companies can't afford that. I think Microsoft's OPEX is like 150 billion. What are we doing here? So, OpenAI also has another fun story. It's leaked to the information this week. This Jermaine, they believe that 80% of their $20 a month Chat GPD Plus subscribers are going away. So, they're just losing 80% of their largest business line this year. But the good news is is they're going to make it up with 109 million Chat GPT go subscribers at five bucks to eight bucks a month with ad with ad support. I just want to be clear that's 109 million new users by the end of this year. I don't think that happens. And that story got it really I saw every coverage of that focus on one thing which was that oh well OpenAI is going to have 109 million new users. Not OpenAI's largest business line is falling apart. And these are OpenAI's own projections as well. That should be cause for alarm. It isn't. the tech industry, the tech media, the business media, the finance media completely captured, refuses to bother even thinking about things for two seconds, OpenAI may just explode one day or more slow down and just crash into something. It really depends on your metaphor for the day.
>> So, in reaction to the news of OpenAI missing its targets, stocks dropped. I think we saw with Nvidia, Microsoft, Broadcom, Coreweave, Oracle, and of course, SoftBank all down. And I think that kind of goes some way to demonstrating the insane levels of interdependence and circular financing that's all going on all at play here. How big of a fallout is there if OpenAI misses its next target as well?
>> Well, I mean, Broadcom, they said they were going to build 10 gigawatts of data centers with Broadcom by the end of 2029. By the way, they're yet to make an order as of the latest that Broadcom's earnings haven't made an order. They are a large client of Coree. And the thing is, it's not just OpenAI that is as well. 67% of Core's earnings in 2025 came from Microsoft who was selling the compute to OpenAI. Kind of double booking revenue. Not really. SEC, if we had one, should look into that. Uh, I mean, Microsoft, OpenAI makes up, Microsoft claimed to have 37 billion dollars of annualized revenue in AI. I reckon 24 billion plus of that is OpenAI's compute. So, any connected counter counterparty to OpenAI is in trouble if OpenAI can't pay its bills. The other thing is with OpenAI's revenue is a lot of it is in revenue backlogs for data centers that aren't built yet. Unclear when that will happen. Yeah, the knock-on effects of OpenAI slowing growth are that you've got I mean at least half a trillion dollars, maybe more worth of backlog of remaining performance obligations across Microsoft, Amazon, uh maybe a little bit of Google and definitely a ton of Oracle like most of Oracle's backlog. I think they have 500 something at least 350 of that is open AI. These are to be clear remaining performance obligations I think are a fake thing. I don't think that companies should be allowed to present them without them. Actually, the regulation should be they could only do 12-month horizons. They can't do beyond that because otherwise you have RPOS that are just into infinity. We'll have talking dogs and flying cars by the time that OpenAI is paying for any of Stargame if they're not dead. So, yeah, the knock-on effects are that OpenAI is a loadbearing fail son. They are a company that if they do not survive then I mean a bunch of revenue goes away in AI and it's all part of the illusory demand story. There's a capacity crisis because open AI and anthropic are taking up most of the world's compute. And so there's not really it's not so much that oh there's so much demand waiting in the wings. It's that what demand exists has a very very small amount of compute it can actually use because of the two large valley fail sons OpenAI and Anthropic who I believe are just most of compute spend at this point. I don't think the rest exists. I actually before coming on tried to look around and find anyone who was spending more than $50 million a year on AI compute that wasn't anthropic open AAI or one of the hyperscalers. And the answer is there isn't anyone. Even Curser who Elon Musk may or may not be buying with SpaceX. It'll either be $60 billion acquired or $10 billion breakup. They're never paying open. I think XAI just kills cursor eventually. But I even them they will get they're the largest independent AI company I think outside of OpenAI and Anthropic. They were going to rent 10,000 GPUs from Elon Musk. That's still only like $40 million of revenue. Even if the largest AI company that isn't an AI lab only needs about 10,000 GPUs to train something, where is the rest of this revenue coming from? And I truly think it's just hyperscalers feeding themselves cash and going look look we money go up number go up or hyperscalers making their own systems to just pretend they're competing. It's very strange.
>> you mentioned that anthropic and open AI are using a lot of the compute of of different companies or sort of the world's compute that's kind of on sale at the moment. Do do you know do you have some more specifics on what those figures are?
>> So okay I worked it out that right now OpenAI represents 70% of Microsoft's AI revenue if not more about 24 billion of annualized revenue. I think that Anthropic may represent 80% or more of the 15 billion in annualized revenue that Amazon has. So I mean Project Rania, which is meant to be 2.2 gawatt out in Indiana for Amazon. Um that's entirely anthropic. They have about 500 megawatts of capacity right now. Epoch AI estimates that Amazon has about two gigawatts of capacity. I think Anthropic's threequarters of that by now. I don't think that people realize how much, but on a revenue basis, yeah, Anthropic is 80% of Amazon's AI revenue. Andy Jesse just talked about AI revenue for the first time at the beginning of April. He shouldn't have. That was a mistake. But these companies take up so much space. Their services are extremely inefficient and dependent and demanding on GPU infrastructure. So, and the thing is if there were other big clients, we'd know who they were because right now the the story is, oh, number keep going up, everything looked good. When you look in the numbers, they never really talk specifically about AI. And every time they do, gets kind of muddled. It's it's like, oh, yeah, AI helped boost this in a way we're not going to break down. The thing is is that at some point there is going to be enough compute for the current demand. And I'm not sure whether much more than that exists. Like because like I said, if there were other big clients, we'd know about them. But go and look at the big comput deals signed in the last two years. Microsoft Meta, Microsoft Meta, OpenAI, Anthropic, Microsoft Meta, and the occasional Google. Terra Wolf, I believe, was Google, but that one for Anthropic, I believe. It's always it all roads lead back to two fail sons, Anthropic and Open AI. And the thing is, it's really easy to look at this picture and say, "Wow, look at the demand. Look at all the GPU demand there is. No one can find GPUs." Yeah, because someone's taking them all up. If you're on a if you're on a bus and one very large man is taking up four seats, do you think that there's crazy demand for bus services or do you just think one person's taking up half the bloody bus? And that's the thing with these companies and that's actually the great lie about this whole thing is that the crazy demand for AI never seems to manifest in somebody being able to give me a number that proves there's crazy demand. What it is is actually that these companies, Google just funded another $10 billion to Anthropic this week with up to $30 billion more. Amazon just invested another $5 billion. I think another $15 billion on the way. And Amazon invested $15 billion in Open AI. They're going to do another $35 billion if they get AGI or IPO. Microsoft invested 13 plus billion dollars in Open AI. It's just handing money. It's either hyperscalers handing money to open air and anthropic who can hand it back or hyperscalers handing money to Jensen Huang and Nvidia. And at some point this falls apart. But my real question is how many data centers are actually getting built? Because I we've been over the fact that only 5 gawatt of the ones meant to come come online this year are actually under construction. But I actually cannot find much proof of data centers coming online. Lots of stories about them not coming online. The stories about like say Stargate Abalene operational. September of last year, one building out of eight. U project Rania. Wow. It's completely operational. No, it's not. It's meant to be 2.2 gawatt of capacity. It's only 500 of it up. The actual But this is the AI bubble. The AI bubble is big headline that says number so huge big thing happened when you look in the actual body it's thing happened in future number not so big number you think want big actually pretty small or we're not going to say it and the funniest thing is on Google's earnings this week my favorite things happened they said oh yeah we have user it was so many users of our TPUs and they named three companies I forget all three of them Boston Dynamics uh thinking machines I can't remember the third one and then they had to go to the information afterwards and say, "Actually, only Boston Dynamics is using TPUs." There's crazy demand for these things. If they if the demand was so crazy, you'd be able to come up with three people that use them. And that's the thing. That's the thing. No, I think only Anthropic is using them. And we don't Google is smart. They didn't disclose their AI revenues because they know I would go through them or someone else would. But that's the thing. Whenever you ask for proof here, whenever you're like, "Hey, so this is you've been spending what Microsoft's over $300 billion in capex, Amazon about 300 billion, Google in the 200s, Meta in the 200s, maybe more. Forgive me for not having the numbers in front of me." And when you ask them to explain what AI is doing, you're talking to the Riddler. It's just like, riddle me this. What if a number went up, but another went down and the growth was so big? And you're just I don't know what you're talking about. Sachin Nadella of Microsoft got us in the earnings call. So um how are you seeing the return on investment? 200word response. No numbers just like yes usage is up and usage is good and usage is fair and AI is so big and so huge but because we have no SEC because we have analysts with rebar in their skulls and we have a captured business and tech media. We are left in this situation where these companies are going to blow so much money. Amazon's nearly at negative free cash flow. I It's unbelievable. And people will say, "Oh, it's just like Amazon Web Services." Wrong. Oh. Between 2002 and 2017, Amazon spent $52 billion. And that's normalized for inflation on Amazon Web Services. And that's all of their capex. That's all of it, including Amazon.com. The story, there is no comparison here. The only comparison here is between functional businesses and the dysfunctional businesses of AI.
>> So if we go back to OpenAI is missing its targets, it's you mentioned earlier that to to justify their $600 billion commit spending, they need to basically double their revenue every year from 2028 to get to 280 billion by 2030.
>> Yes. But considering they just missed their targets, what kind of growth are we going to have to see for them to even get close to closing that gap?
>> Based on what I've been coming up with last week, it's about they have to 10x their revenue from here. That's that's the long and short of it. They have to 10x their re because this year they're looking to make about $25 billion. They actually have to more than 10x because they need $284 billion in 2030 and to be profitable. That's the other thing they say happens that year. I don't know. Uh yeah, they would have to 10x their revenue. It's that simple. And yeah, they can't. I just that's not happening. I don't think they're alive in two years. I think they're dead as Disco. And I think Disco actually based on YouTube is quite alive compared to OpenAI. The thing is no one It's weird how few people want to talk about this because when you start thinking about to your question the slowdown of open AI the rest of it begins to feel increasingly more irrational because you're like okay the only way that Oracle can justify their 384 billion sorry 348 billion in capex that they'll spend on all of the Stargate data centers is if OpenAI can pay them $75 5 billion a year. Open AAI has to grow. Has to. It's no longer a they should grow. It would be interest. They need to grow. They must grow now because $250 billion of uh incremental compute on Microsoft Azure, $138 billion over eight years on Amazon Web Services, indeterminate amounts of money on Google Cloud, uh $22.4 4 billion on coreweave over 5 years and $20 billion over three years with Sarah Brus. So that's a lot of money and that's more money than they have even their $122 billion probably raised 60 at most. Yeah, they can't afford that. And what gets a bit scary is when you go and look at how the money's being raised as well. Soft Bank Soft Bank is having to do a bizarre IPO of an unknown robotics company. They're trying to do a hundred billion dollar IPO there. They're doing a $40 billion bridge loan to fund this. Even if all of this works out, this is how much SoftBank has had to do. Now, OpenAI needs them to do this three or four more times, maybe more. Amazon had to split up the investment. 15 billion up front, $ 35 billion if you can do other things. This is how difficult it's going to be, if not more, when they keep growing. They need that to happen. They could go to the bond and the debt market, but I can't find a company in history that has actually raised more than $50 billion on an annual basis. And those were real companies. I'm talking like the aged form of G Capital before Sain and stuff like that. I'm talking about Microsoft. Even Microsoft's not done that much. Oracle has. Look what's happening to Oracle stock. I think that people are very flippant about how bad things are economically and I think they are looking for any excuse to ignore the big blinking blaring warning signs that we see every month. And the thing is OpenAI cannot afford to slow down and neither can Anthropic. Anthropic raised what cumulative I think it would be another60 $70 billion between Amazon and Google. They're apparently working on another funding round for another $50 billion at a $900 billion round. Anthropic also has made 10 gawatt of compute commitments across Amazon and Google. It's this doesn't look good. There's not enough money to to substantiate all of these fail sons. How are we possibly going to keep them alive? I also think that we need a regulatory body, if one existed, to ban this kind of thing. I think circular financing should be illegal. I don't think it is an honest way of doing business. It helps establish. It's funny, this era is sold on this level of democratization. What it actually is is helping retrench the largest players. Except the largest players gambled hard on this thinking that it would magically turn into something profitable and it won't. And they we have the two loadbearing fail sons of Anthropic and OpenAI who must grow cuz if they don't grow Amazon, Google and Microsoft have wasted their capex because there isn't enough revenue to substantiate it. You'll mention I haven't you'll notice I haven't mentioned Meta. I got no idea what Meta is doing. I I don't think that they know. Meta wants to spend another $180 billion dollars on capex. Why? So they can lead another old man with dementia to his death per a story from Reuters last year. That happened. Kylie Jenner LLM bot on Meta's uh platform that led a man with dementia to New York City and then he fell down and hit his head and died. Meta like people should be there should be criminal inquiries into LLMs. But putting all that aside, there is no business from Meta. Meta, same Griddler treatment as everyone else. It's like, oh, we saw a 3% lift here and there, Batman. Oh, you'll never guess what we have coming. It's more crap. You're going to hate it. But we've connected a bunch of numbers here and there, but then when you look at the numbers meta, they had their first daily active people. Their measurement of how many users they have, they had their first drop. They had their first drop, I think, in years, if not ever. Yeah. The core business of Meta is dying. The core business of all of these companies is slowing down. The only thing they have to patch it up is AI. And the only way to patch that up is to feed Open AI and anthropic. The circle is complete.
I want to get on to the the comments from Sarah Frier in a second, but just really quickly, last week we were talking about how Open AI kills Oracle and and I do recommend anyone watching to go and read Ed's newsletter on that exact topic, but I I just want to quickly ask, we saw Coreweave nose dive following the the missed targets from OpenAI about the same amount as Oracle, and I was just wondering why.
>> Basically,
>> well, I mentioned it earlier, but 67% of Core Weave's earning of revenue in 2025 came from Microsoft. All of that compute is for Open AI. All um they have tens of billions of dollars of backlog from Microsoft. And wouldn't you know it, Open AI. Most of their revenue is coming from OpenAI. You know who's also on their backlog? Anthropic. Microsoft. Now they they have a huge backlog as well, but they still have other businesses that are doing okay. That's why they didn't get savage so bad. Oracle, most of their remaining performance obligations are Open AI. Every company attached to this has benefited every time Open AI looks good. Now they're paying for it. And they're not paying for it very hard yet. This is just a taster, an amuse bouch for these douches where they're going to see this is just a taste. This is just the beginning. Open AAI hasn't started to stumble yet. We just rumor mills at this point. When the actual problems start happening, when the actual checks start bouncing, that's when you'll see these stocks get crushed. And it's because while Corewave claims to have this diversified revenue stream again, topline stuff diversified all of this backlog. Actual numbers 67% Microsoft. The rest of it, by the way, I think another 10 or 12% is Nvidia. It's Nvidia buying It's Nvidia buying their compute, which is not something you do. You don't need someone to buy your remaining compute when there's actual real demand, unless of course the demand doesn't really exist outside of two companies.
So on to Sarah Frier that you mentioned the that she admitted she was worried about the company's ability to to pay for future compute contracts if revenue growth doesn't accelerate. I mean I think as far as we know those those comments are themselves rumored or or do you know if those
>> they're real?
>> Yeah the the information reported something very similar about a month ago. Burbagen great reporter very well sourced went and followed up and found them. The the thing that changed is the comments have been made multiple times now, which is I I don't know why people are not a little bit more concerned about this. When your chief financial officer, the person with the money is worried about the money, that's when you should be worried about the money. One of the other comments that was made in Burber's article was really scary as well. She said she didn't think that Open AI was ready for the rigorous standards of going public and public scrutiny. To which I ask, what do you mean by that, Sarah? What? This company is ostensibly worth $850 billion. It's the loadbearing fail son of Microsoft and Oracle and Coreweave. What do you What do you mean? What do you mean that they they don't? What? How much? Because when I hear that I start I start hearing Nortell. I start hearing I start hearing Lucent. I start hearing WinStar. I think of all of the dotcom brilliance because OpenAI is not a sophisticated business. It's not a business like say I don't know a mass manufacturing company where you have a ton of procurement. You have a ton of physical labor happening. OpenAI isn't spending any capex. OpenAI isn't building their own data centers. They have they have four 5,000 people. They have compute costs for training. They have compute costs for inference. And they have compute contracts. I'm sure it's a complex more complex than my business perhaps, but I'm not sure what problems they'd have. And if I had to theorize, it would be that they have piss poor accounting when it comes to their compute spend. if they don't know how much they're actually spending on compute then that demand story is twice as bad because I don't know if you don't know how much you're spending how do you possibly know how much you will have at the end of the month or perhaps it put another way do you not know how much money you're making in which case the reporting I did last year which did not match up I got actual real Microsoft Azurebacked information from a source where OpenAI's revenues did not match up with those reported. Perhaps it's the actual accounting is fudged. I don't know. We are guessing. But all I know is Anthropic raising another round instead of going public. Open AAI making noises like they really want to go public except their CFO doesn't want them to and she would have to sign off on it. But the other part of that story that people keep forgetting which came from the one in a in early April at least was that Sarah Frier has also per the information been left out of conversations around data center to spend. Now here's the thing. I ain't no Fortune 30012 CEO. I ain't no business genius. But well here's the thing. I like to have the money person know how much money he spent that good. If she is being left out, if the CFO of OpenAI is being left out of compute decisionm and actual financial decision making, then this is a company in retreat, a company in collapse because my god, that is the classic dysfunctional business thing to do. When a CEO says, I don't like what my these annoying money people are saying. Uh, I'm just gonna I'm just going to freeball this one. I don't need to think too much about it. Which is not what you do when you need to have $852 billion by the end of 2030. Sammy Sammi is not thinking too good. I think we're in the era of desperation. Fascinating to watch. Honestly,
>> like you say, Sarah Sarah Fry is saying we would probably wouldn't even make the standards to be a public company, which raises the question why Alman is so keen to rush to an IPO and and some analysis that I saw basically just said he wants to offload the massive compute debts that they've got onto real retail investors before the bubble pops. And firstly, I would you agree with that? And where does all of this news that we've been talking about leave the hopes of getting a late 2026 IPO?
>> I mean, Sam wants to rush to IPOs so that his investors get paid off. That's why he wants to further entrench his Silicon Valley relationships. That's mostly it. I also don't think people realize how much money you can get as a public company. The largest IPO of all time, I think, was Saudi Aramco, about $30 billion. If OpenAI and Anthropic and SpaceX each want to raise $50 billion, I think all the IPOs combined in 2025 were 42 or 47 billion. Forgive me if you're not knowing off the top of my head. So, long story short, lot of money to just raise. Also, $50 billion. That's not even 10% of how much money you need, Sammy. So, they'd have to raise debt because Open AI is a big pile of crap. They are not. They are a business that loses billions of dollars that will lose billions of dollars in perpetuity. They would not. I don't even think they'd raise investment grade debt, which means that they would have to raise junk bonds. There's only so much you can raise. They would need 50 to 100 billion every single year. Every year. Or they could do equity raises post IPO, but people don't like those. They could do an at the market share sale like Oracle is doing, but they would have to do everything I'm saying, which is usually something that is spread across years. They would have to do it every year. People can delute themselves into believing that that would happen. Hell, they could go public and do it maybe once or twice, but at some point the money will run out, especially when you've got everything happening with the rest of the economy. I don't think they IPO this year, but if they if they do, it will be under Sarah Fry's juress, which means it will be a messy, rushed IPO. And IPOs are the one thing you don't want to rush because when you file your S1, the paperwork to go public, you actually have to have something called an accountant look at it. You have to get an auditor and they'll look at the numbers and you legally have to say what those numbers actually are. You can't do you can sorry you can do non-GAAP so generally accepted accounting principles like uh we works community adjusted EBIT. You can come up with all the numbers you want, but the real numbers are the things that people are going to look at. And while there's a lot of hype around AI, I don't think there's enough hype to deal with the fact that the innards of Open AI smell and look like a torn from the Empire Strikes Back. This is a stinky company. This is a company that is not wellrun that is pretty much only made possible by hyperscalers feeding themselves money.
>> How many more times can Open AAI miss its targets before it's terminal?
>> It might already be like just because if you think of it like this, there are two billion a month they say. I don't know how much I believe that, but nevertheless, two billion a month.
>> Mhm.
>> That's very high already. It's very high for a private company as well. How much more can you grow? Because even $20 billion a month isn't going to be enough. I like how is this how does this work exactly? I think like Microsoft makes $ 34 billion a quarter from its uh intelligent cloud segment which is where OpenAI's compute spend goes. So um OpenAI will just be larger than that in a few years. How? I don't know man. Oh, they'll get 103. To be clear, if they get 103 million of five to eight dollar subscribers, that might not cover the gap of the $20 buck a month ones. But even then, OpenAI would probably to do what they need to say, keep all of their $20 a month subscribers and add $ 103 million. And they've also projected their ads business will make hundred billion in a few years. YouTube, one of the single most popular places on the internet, makes about $60 billion a year in revenue. So yeah, OpenAI's growth slowing already suggests that it's growth. They've not even done the hard part yet. They've got to the point where like, wow, one of the biggest startups of all time. That's huge. Great job. Other than the billions of dollars of losses, not enough. Not even remotely enough. Maybe uh well, tenth of the way there. They need to grow so much more. and any slowing in growth suggests that it's comical to expect them to reach these targets. And as we saw from the little Ripple and the stock market, I don't know how it's how their counterparties are going to react to the fact that they're not going to have the money. They should already be reacting, but the market's stupid and will stay stupid until it's made to be smart.
>> We've spoken a lot recently about AI being subsidized, and so so has the news started to do that as well. that putting it putting sales targets aside for a second and looking just at the user targets they're if they're being missed when AI is as cheap as it's going to get surely that means open AI won't survive moving over to charging user rates or token rates or usage rates or token rates or or whatever.
>> Yeah. So I think the order of events is going to be June 1st 2026 GitHub copilot goes tokenbased. I think anthropic follows. I think OpenAI is the last man standing. Uh because you've wa if you watch what happens whenever Anthropic does a rate limit thing, Open AAI lowers their uh they loosen their rate limits. It's really a stupid strategy to try and express growth. But I think Open AAI really waits. If Open AI goes to tokenbased billing, that is how you know it's over. That's how you know they are they have decided that they can't lose as much money. We also don't know if influence is profitable. I don't think it is. So, we'll we'll find out. I think they will hang out as long as they can though because they have the largest cash pile of all of them. They have the biggest hugest sugar daddies of all. They got Microsoft and Amazon providing their compute now. Both of them have invested in the company. Got to wonder how much real cash is changing hands. So yeah, any sign that OpenAI is moving to tokenbased and they've already done it with their enterprise customers by the way, but I'm talking about the regular the pros, the people on the street, that will be the sign. Now, it may be that they're made to because if Anthropic does, the thing is Microsoft is giving people permission, but until someone else other than Microsoft does, it's not really cracked off yet. But I think what we're gonna see, it needs to happen because Microsoft announced this at the what week ago. I caught the scoop a week before that. You can read about it early with me. Um Microsoft says June 5th. I think on that day you're going to see because users on GitHub Copilot are going to stick around. They're going to stick around not really understanding how expensive it is. I think once they do and those stories wave out, you're going to see the crisis grow because I think people will realize that their way of life is no longer possible. And if Anthropic ever follows suit with Claude code and or Claude subscription Max Max 20x and so on, I think that that is again beginning of the end and we have every sign we need to say that this is happening other than the final decision being made. subsidized AI kind kind of brings me on to the the comments from the NVIDIA VP of of applied deep learning who who said that AI is costing him more than a human worker would even even at this discounted rate that we were just talking about. Where where does that leave the underlying sales pitch of of the whole AI bubble if you're actually paying more for something that's wrong 10% of the time and it is also way worse for the environment.
>> Well, as we live in the era of the business idiot, it will it event it's really the war between the CFO and the CEO because right now CEOs in charge CEO going AI is the most important thing ever. Coders 10x wow big big number go up so much. I think right now you were seeing because bigger organizations are having to pay their token burn now especially with anthropic. I think you are going to slowly see this message grow and grow and grow and I think eventually I don't know someone's going to go hey are we spending more on the machine than the people why are we doing this you know the things I've been saying they should do for years and I think then people are just going to start really frankly looking at it and saying nah I'm not going to do this best case scenario they move on to Kim K2 or Deepseek 4 what have you but I think even there those are still going to be really expensive and so you are right now as well important detail open AAI has higher rate limits sorry looser rate limits I should say on codeex through May so we're going to see soon even that subsidy start to collapse but I think at some point someone's going to realize they're spending 300 bucks a day on this and they're going to say am I getting 300 bucks worth of value out of this and they're going to say no and they're going to start doing token tightening I reckon We're going to see token austerity. So token token limits at Meta, at Microsoft. I already heard people at Microsoft discussing it. I think across the board, people are going to start hitting token limits at companies. Companies are going to have to start them because you have to um if if everyone was running a microwave at their desk for 10 hours a day, I think they probably ask them to stop. And I think that that is going to be part of what breaks this all because I like you said the whole point of this not that I want this to happen is that people are meant to be replaced by these things. People aren't. They're not. Also these things cost as much if not more as a person. Why? Why? Other than the fact we can all say we're doing AI. I'm not really sure what the point is. And before with the subsidized tokens, yeah, it was really easy to wave this off and claim it was great. People right now, the majority of AI users are used to a service that will not exist in a year. The idea of being able to just spooge as much AI as you want, that's going away. But also, most people are not used to using these products in a way that is germaine to paying tokenbased rates. most people are used to and I think it was one of the deceptive parts of the media campaign as well where it's you read all these breathless sweaty pieces about claude code they're all based on people using them on a $20 a month heavily subsidized product if they had they don't know like you if something goes wrong on one of those subsidized products you go oh this thing these things happen because you're not paying for it I don't know how people feel about it if when one of the many loops that Claude code goes on or codeex goes on. If that loop lost you $2 or $3 or $5 or you spent $50 and you realized you're barely any of the way through the project, I don't think people are going to be quite as excited. And that's the marketing campaign. This is a deception. The whole thing has been a deception. It was never economically viable to offer a monthly subscription rate AI product. And when people use the tokenbased systems, they're going to see that really quickly. And also, by the way, people are going to say, well, I heard Dario Ammed say that they're profitable on inference. It's 50% gross margins on inference. That is not true. I've heard lots of people say this. He said on a interview in February with Dwarkesh Patel that his stylized facts his words were that if you that this company could be profitable with 50% gross margins on inference and then explicitly said he was not talking about anthropic. Again, when you want the numbers that prove that your business is good, every AI founder turns into the Riddler. How do you think the 90 odd thousand employee or tech employees who were laid off this year under the guise of AI efficiency gains are feeling seeing the news that the person selling the picks and shovels is saying is saying well actually it's it's kind of easier it's kind of cheaper to dig by hand. I think that the tech industry has massively massively overplayed its hand and also I think they do not even realize how much they have pissed off everyone. This is a vulgar, wasteful, aimless, directionless technology. It is unreliable at its at its core. It is expensive in a way that nothing has ever been expensive before. There is no comparison. Bubble's nothing like it. the some micros systemystems uh servers you needed was 64 grand but you didn't need anywhere near as many of them people are being fired on a completely fictional basis they're being f people at meta I have heard are being fired who are performers people who are very good they're arbitrarily laying people off now I think the tech industry is radicalizing people against the tech industry they're radicalizing customers by making the products worse by putting AI in everything by taking up massive amounts of land with loud ass data centers that sound like a goddamn Dalek. They are firing people by saying, "Yeah, we found a technology that's better than you, except the technology sucks and it doesn't replace anyone. At best, it makes people more more efficient in a way that no one's no one can categorize. Also, these things cost as much if not more as a person. Why? Why? Other than the fact we can all say we're doing AI. I'm not really sure what the point is. And before with the subsidized tokens, yeah, it was really easy to wave this off and claim it was great. People right now, the majority of AI users are used to a service that will not exist in a year. The idea of being able to just spooge as much AI as you want, that's going away. But also, most people are not used to using these products in a way that is germaine to paying tokenbased rates. most people are used to and I think it was one of the deceptive parts of the media campaign as well where it's you read all these breathless sweaty pieces about claude code they're all based on people using them on a $20 a month heavily subsidized product if they had they don't know like you if something goes wrong on one of those subsidized products you go oh this thing these things happen because you're not paying for it I don't know how people feel about it if when one of the many loops that Claude code goes on or codeex goes on. If that loop lost you $2 or $3 or $5 or you spent $50 and you realized you're barely any of the way through the project, I don't think people are going to be quite as excited. And that's the marketing campaign. This is a deception. The whole thing has been a deception. It was never economically viable to offer a monthly subscription rate AI product. And when people use the tokenbased systems, they're going to see that really quickly. And also, by the way, people are going to say, well, I heard Dario Ammed say that they're profitable on inference. It's 50% gross margins on inference. That is not true. I've heard lots of people say this. He said on a interview in February with Dwarkesh Patel that his stylized facts his words were that if you that this company could be profitable with 50% gross margins on inference and then explicitly said he was not talking about anthropic. Again, when you want the numbers that prove that your business is good, every AI founder turns into the Riddler. How do you think the 90 odd thousand employee or tech employees who were laid off this year under the guise of AI efficiency gains are feeling seeing the news that the person selling the picks and shovels is saying is saying well actually it's it's kind of easier it's kind of cheaper to dig by hand. I think that the tech industry has massively massively overplayed its hand and also I think they do not even realize how much they have pissed off everyone. This is a vulgar, wasteful, aimless, directionless technology. It is unreliable at its at its core. It is expensive in a way that nothing has ever been expensive before. There is no comparison. Bubble's nothing like it. the some micros systemystems uh servers you needed was 64 grand but you didn't need anywhere near as many of them people are being fired on a completely fictional basis they're being f people at meta I have heard are being fired who are performers people who are very good they're arbitrarily laying people off now I think the tech industry is radicalizing people against the tech industry they're radicalizing customers by making the products worse by putting AI in everything by taking up massive amounts of land with loud ass data centers that sound like a goddamn Dalek. They are firing people by saying, "Yeah, we found a technology that's better than you, except the technology sucks and it doesn't replace anyone. At best, it makes people more more efficient in a way that no one's no one can categorize. Also, they're spending more money than ever before while they're laying you off. They're laying you off. They're saying we cannot afford you but we can afford another data center that will get built in three, four, five years maybe and uh we'll lose us money the moment we turn it on. That is more important than you a person. Yeah, I think most people are going to end up hating the tech industry by this point. I think that by the end of the AI bubble I think most people are going to be angry and I don't think that the tech industry has any idea about it. I think the business idiots at the top, the Sachin Nadellas of the world, they are so disconnected from real people's problems that they don't even know how to solve them anymore because how would you know what's going on? Do you think S when do you think the last time that Sachin Adella talked to a janitor? When do you think he last talked to like a regular person? You think he walks into a shop? You think Mark Zuckerberg has gone to a shop recently? You think that Sam Alman has, I don't know, gone and bought a sandwich? You think he's gone and picked up a sausage roll? You think that that's happened? You think any of these people have regular exper You think Elon Musk, it goes to the public swimming pool. You think that any of these people have regular experiences anymore? When did any of them do any work last? They go to and from lunch. They wake up at 5 in the morning. All of them say we wake up at 5 in the morning to do what? To spend other people's money and waste it. That's the thing. When you have so many companies run by people who are so rich and so disconnected, of course they're not going to actually realize this stuff sucks. I paraphrase a Twitter post, but it's like being kicked in the head by a horse every day. They don't experience reality. So, they're going to keep doing this. And I mean, this is kind of what you're seeing across the board with these CEOs who are spending as much as an employee costs while firing real employees because AI is so good because they are not the ones using the product. They're not the actual people doing work. And thus, they are going to create things that are what they consider connected to reality, but it's a reality that they themselves have crafted and can only last as long as, well, the markets are stupid. And yeah, the markets love being irrational, but look at how quickly they freaked out with Open AI. Look how quickly they got spooked by that. And that was a minor story. Wait until something big happens. Wait until something really shakes them because they might be looking for a reason to run. They might be looking for a reason to run high. I mean, but they're also looking for a reason to doubt themselves because even the most lead poisoned hedge fund kind of see something's wrong now. And it's weird because right now we're in the hysterical era where number go up so big. Eventually that's not going to be enough. Eventually, we're going to need to see real ROI from AI. And it's completely insane that we are 3 years in and we're still solving the riddle of the Sphinx every time we ask someone, "Hey, does this make more money than you spent?"
>> Well, Ed Zitron, thanks for taking the time.
>> Thanks for having me.
>> If you enjoy today's episode and you want to hear more of the Tech Report, please consider liking and subscribing. Also, you can get episodes of the Tech Report wherever you get your podcasts.