# Ed Zitron Unfiltered on OpenAI, Anthropic & Why the Whole Thing Is a Con

Markdown copy of https://realalpha.app/v/wOBTi8dAb1Y/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

Ed Zitron argues that AI's apparent value rests on subsidized pricing, while returns and the cost of scaling remain unproven.

- Show: Newcomer Pod
- Hosts: Eric Newcomer
- Guest: [Ed Zitron](https://realalpha.app/people/ed-zitron/) is a writer and podcast host.
- Published: Jun 8, 2026
- Length: 58 min

## The best minutes

- [12:01](https://realalpha.app/v/wOBTi8dAb1Y/?t=721) [AI subscriptions hide the real cost of a unit of work](https://realalpha.app/s/ai-subscriptions-hide-the-real-cost-of-a-unit-of-work/) (7 min)
- [33:06](https://realalpha.app/v/wOBTi8dAb1Y/?t=1986) [Data center expansion depends on an enormous debt market](https://realalpha.app/s/data-center-expansion-depends-on-an-enormous-debt-market/) (2 min)
- [41:37](https://realalpha.app/v/wOBTi8dAb1Y/?t=2497) [Executive demos can make AI look more capable than it is](https://realalpha.app/s/executive-demos-can-make-ai-look-more-capable-than-it-is/) (7 min)

## Also worth watching

- [8:56](https://realalpha.app/v/wOBTi8dAb1Y/?t=536) [Anthropic’s reported profit coincided with a compute discount](https://realalpha.app/s/anthropics-reported-profit-coincided-with-a-compute-discount/) (2 min)
- [27:17](https://realalpha.app/v/wOBTi8dAb1Y/?t=1637) [Anthropic has not defined its reported run rate](https://realalpha.app/s/anthropic-has-not-defined-its-reported-run-rate/) (5 min)
- [37:59](https://realalpha.app/v/wOBTi8dAb1Y/?t=2279) [Tech firms pushed AI into products before users asked for it](https://realalpha.app/s/tech-firms-pushed-ai-into-products-before-users-asked-for-it/) (4 min)
- [50:49](https://realalpha.app/v/wOBTi8dAb1Y/?t=3049) [Some AI safety stories omitted how tests were staged](https://realalpha.app/s/some-ai-safety-stories-omitted-how-tests-were-staged/) (3 min)

Skip the extended model-cost debate after the Uber segment and the closing exchange about AI’s trillion-dollar cost.

## Summary

Newcomer introduces Zitron as a podcast and newsletter writer whose critiques include 18,000-word takedowns. This is Zitron's second appearance on the Newcomer Podcast.

Ed Zitron argues that foundation-model companies have not shown a clear return on investment and that subscriptions conceal the high token costs of AI use. He says companies and users are being trained on subsidized pricing, while Eric Newcomer argues that AI already provides practical value and that lower model prices or business adaptation could matter. They dispute the comparison with Uber, the reliability and usefulness of coding tools, whether broad AI adoption was pushed on the public, and what level of accuracy is reasonable. They close by questioning whether present-day large language models justify the trillion dollars invested.

### Minute by minute

- [1:38](https://realalpha.app/v/wOBTi8dAb1Y/?t=98) **AI writing feels lifeless.** Zitron says AI-generated prose feels lifeless and relies on recognizable linguistic patterns. He contrasts it with even boring human writing, which can still convey something human.
- [5:05](https://realalpha.app/v/wOBTi8dAb1Y/?t=305) **AI's return remains unproven.** Zitron argues that AI companies talk about theoretical potential rather than demonstrated returns. He says flat-rate subscriptions hide token costs and calls that pricing a deliberate con. He also says skeptics are treated as outsiders in Silicon Valley.
- [9:18](https://realalpha.app/v/wOBTi8dAb1Y/?t=558) **Zitron opposes OpenAI and Anthropic IPOs.** Zitron says regulators should not let OpenAI or Anthropic go public, arguing that their leaders sell theoretical capabilities. He claims a SpaceX filing projected Anthropic profitability in May and June, when Elon Musk discounted its compute. Newcomer presses him on what a public listing might reveal.
- [12:01](https://realalpha.app/v/wOBTi8dAb1Y/?t=721) **AI subscriptions distort unit economics.** Zitron contrasts Anthropic's spending with Uber's history, saying Uber burned about $32 billion while Anthropic had raised over $75 billion in the previous six months. He argues that flat subscriptions have trained users to consume AI without seeing its real cost; Newcomer says Uber also faced criticism before becoming profitable.
- [18:59](https://realalpha.app/v/wOBTi8dAb1Y/?t=1139) **More tokens can offset lower prices.** Zitron says newer reasoning models can consume more tokens, so lower prices per token do not necessarily reduce total costs. Newcomer points to cheaper older models and open-source alternatives, while Zitron questions how widely companies can replace the models they use.
- [24:19](https://realalpha.app/v/wOBTi8dAb1Y/?t=1459) **Subscriptions conceal AI's full cost.** Newcomer says his company used AI to build ticketing infrastructure, but Zitron argues that a subsidized subscription does not show its true cost. He cites Anthropic documentation estimating average Claude Code use at $13 a day in tokens and says Uber capped engineers at $1,500 a month.
- [29:11](https://realalpha.app/v/wOBTi8dAb1Y/?t=1751) **Run-rate revenue can obscure volatility.** Zitron says annualized run rates can mislead when API revenue fluctuates and argues that Anthropic has not clearly defined its run rate. Newcomer says investors understand forward-looking revenue measures; Zitron counters that companies should report revenue and disclose their non-GAAP calculations.
- [32:00](https://realalpha.app/v/wOBTi8dAb1Y/?t=1920) **Zitron says data center financing is strained.** Zitron acknowledges that he misjudged how quickly the AI boom would end. He points to Google's $84 billion equity raise as a possible bubble signal and says credit markets are drying up. He estimates the projected GPU buildout would require about $800 billion in debt.
- [35:11](https://realalpha.app/v/wOBTi8dAb1Y/?t=2111) **Zitron questions AI's value beyond code.** Zitron asks why coding dominates claims about AI's usefulness. Newcomer says users need to explore the tools to find value, while Zitron compares their reliability unfavorably with the immediate impact he felt from his first iPhone. They also dispute how much current AI use is driven by user interest.
- [40:21](https://realalpha.app/v/wOBTi8dAb1Y/?t=2421) **Zitron calls AI adoption manufactured.** Zitron argues that companies and media have pushed AI into everyday products and framed it as a threat to workers, rather than letting people choose whether to use it. Newcomer says interest in AI is also organic. They then debate whether executives use AI demonstrations as a reason to cut staff.
- [43:43](https://realalpha.app/v/wOBTi8dAb1Y/?t=2623) **Zitron separates generative AI from other AI.** Zitron says executives can overestimate what AI demos prove and argues that leaders often make poor decisions about workers and products. When Newcomer points to uses such as drug discovery, Zitron says those techniques should not be conflated with generative AI.
- [48:03](https://realalpha.app/v/wOBTi8dAb1Y/?t=2883) **They dispute the standard for hallucinations.** Zitron says models cannot learn from mistakes as people can and argues that hallucinations make them unreliable. Newcomer says requiring no errors sets an unrealistic bar, while Zitron responds that companies have marketed the systems as intelligent and capable of replacing workers. They also dispute how AI safety tests and their results are presented.
- [52:05](https://realalpha.app/v/wOBTi8dAb1Y/?t=3125) **Zitron links AI skepticism to its costs.** Zitron says executives and investors have not done enough due diligence and points to data centers' effects on local environments. He argues that media coverage dismisses people who question AI. Newcomer presses him on what positive future he sees; Zitron says he loves technology but thinks the industry is damaging it.
- [56:12](https://realalpha.app/v/wOBTi8dAb1Y/?t=3372) **They question whether AI was worth the cost.** Zitron asks whether today's large language models justify the trillion dollars spent if development stopped now. Newcomer points to expected company valuations, but Zitron clarifies that he is asking whether the money spent so far was worthwhile.

## Transcript

Newcomer Pod, hosted by Eric Newcomer, with Ed Zitron.

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

[0:00](https://realalpha.app/v/wOBTi8dAb1Y/?t=0) Today on the podcast, I finally bring Ed Zitron around on AI. Lifeless, it's empty, gonna look like a dog's asshole on Thanksgiving. You stupid, you moron, why don't you like data centers? Okay, maybe not quite, but I do have a lively debate with tech's loudest AI skeptic. So how are you going to get around to the return on investment thing? He's like, yeah, somebody will work it out. Sammy, clammy Sammy, this is your job. From his podcast, his newsletter, and his 18,000 word takedowns, He has made one thing very clear. AI is a bubble that's about to burst. I'm Eric Newcomer, author of the Newcomer Substack.

[0:36](https://realalpha.app/v/wOBTi8dAb1Y/?t=36) Let's get into it. Ed Zitron, welcome back to the Newcomer Podcast, round two. Thanks for having me. We're live, in person, just you and me, mano y mano. I asked Claude for, I don't know, what did I say? give some monikers for you all right we've got the hater laureate cassandra of compute the doom sommelier tech's designated mourner the bard of the burn rate chief bubble officer uh the last honest man it's that that's a charitable one i guess i just like i asked the dog would it give me just i'd be just as excited it wouldn't be as good though the dog i would care i would have more respect for it it's just like those are things that you get people saying in your mentions every so often it's like like a regular person like good on you like a normie but like it's like claude slop okay the hater laurier okay just like unex such it's so unexceptional that's the thing it's not just mediocre and all these different ways it's just like you read the text it just feels lifeless and if you go on substack it's i think it's mostly clawed written all subsack yes so many we saw yeah there were i forget the company's name but there's some pancrum right yeah yeah yeah and it's just it's very depressing and so it's lifeless

[2:04](https://realalpha.app/v/wOBTi8dAb1Y/?t=124) it's empty even human written linkedin slop evokes some kind of reaction even even if it's just reminded of the feeling of being at a conference you still there's still something human about it even if it's quite boring but you read the clawed slop as it's not just this it's it's this and that was the twist that was the turning point there's these little linguistic tropes it's just sad it's not even good it's not like we're still in this years in we're still like look yay we can kind of write like a person and it sucks and you can tell every time yeah it's just so boring so that that gives us a taste of where where you're coming from i to give context you know uh to people company this episode you know i host an ai conference i write in sub stack newsletter that is not ai slop ai copy written you know i pay you i pay thank thank you it's great um big fan of your newsletter as well and um though it obviously has a point of view we will get into here um yeah um we did we did one episode before nine months ago sort of disagreed for for context For the viewer on me, you know, I feel like I'm somewhere, you know, I host an AI conference.

[3:17](https://realalpha.app/v/wOBTi8dAb1Y/?t=197) I wrote a case against OpenAI's valuation at, I think it was $157 billion, which is March past. I sometimes report data that you're happy to have. So I feel like I sit in some middle ground where sometimes we're allied, sometimes we're clashing heads. And here's the thing with you. I know we don't agree on everything, but you published Tom Dutan, who's a fantastic journalist, who is also quite critical. You yourself are one of the few people who's been even positive you are still willing to critique Anthropic, the Kutu, what's the Kutu? Kutu D, yeah. Where they said, oh, we think it'll be worth 1.99 trillion.

[3:56](https://realalpha.app/v/wOBTi8dAb1Y/?t=236) I don't want to say two. That would be ludicrous. Yeah, that was just that we couldn't possibly value it. a company that burns billions of dollars that much. And then Elon Musk comes along and is like, watch this, it's so based. And it's, but you at least try. You at least try and find some middle ground of, okay, you think that this will work, I don't, whatever. But you at least try and found that in something. And even then you bring in some critical. And I think that's healthy. I think it's the healthier way of approaching this for someone who is going to be pro AI or at least willing to accept a pro AI opinion.

[4:26](https://realalpha.app/v/wOBTi8dAb1Y/?t=266) I never want to have a shtick. I feel like my disposition is sort of I always like to poke and prod at things. So I think this, you know, we'll get into it and I'm sure we'll push back. But at first, I just sort of want to give you a chance, like lay out your case for a couple of minutes. Like, where are you today? Like what is there? There's sort of this like. It's a fraud. It doesn't work. Nobody likes it. You know, there's sort of the self-dealing aspect. There are lots of pieces to it. Like, where are you right now on the, like, it sucks, nobody wants it versus these numbers are made up? Or like, what is your core case against the foundation model companies today?

[5:04](https://realalpha.app/v/wOBTi8dAb1Y/?t=304) So we're really talking about like OpenAI Anthropic, like the case against them. So fundamentally, there is no ROI in AI. You can dress it up however you want. Look at how Anthropic and OpenAI discuss their products in the sense that they don't. They're always talking in theoreticals, in, well, the potential's there. Now, everyone's always talking about the fact that we are still having a conversation about AI's value and whether AI is real or not is kind of the foundation of my argument in that if this was real, we would definitively know. We wouldn't have people saying emphatically, like, AI is here and it's real.

[5:39](https://realalpha.app/v/wOBTi8dAb1Y/?t=339) You don't need to say that. You don't need to say that if something is real. so i have never disputed that llms can do some code but i think people have over become a little overwrought about what kind of coding they can do and how reliable or stable or functional that coding is but the biggest thing by far is that the con is the financial side the con is the fact that most people's experience of ai is unrealistic when it comes to the cost so you using chat gpt i believe the subscriptions will go away i think we are just at the dawn of token-based billing meaning that most people's use of clawed code for example is based on a relatively unlimited you can burn between eight and thirteen and a half dollars for every dollar of your subscription in tokens so most people's experience of ai is separate from the cost divorced from the cost and this was a deliberate con on the part of sam altman and dario amade because they both knew that if they actually made people pay on a per million token basis to use these things they go wait how much what whoa this this thing messed up and i still had to pay and you're kind of seeing that already with the enterprise customers the ubers and walmarts of the world who are starting to

[6:51](https://realalpha.app/v/wOBTi8dAb1Y/?t=411) have to pull back this is the token maxing soul searching we're seeing going well the thing is uber isn't wasn't token maxing they i don't think they had a leaderboard i think i think they might have at some point but they burned through their entire annual token budget in four months in the space of the first quarter and then their COO Andrew McDonald said it's getting hard to justify because you can't connect it to outcomes this is we are years in we can't we why are we all acting like this is going well it isn't and so the experience of AI the value judgments of AI are all based on phony metrics an outright con about the costs and quite frankly a very weird psyop the whole thing is a psyop the cult of personality around ai the very much if you are in the out group you don't like it you're you're othered you're different you're a you're a hater you're a skeptic and it's frustrating because silicon valley is meant to be a place of meritocracy it's meant to be a place where you're evaluated based on your value theoretically speaking remember the i don't know if you ever saw the github thing the meritocracy doormat they had no i look like they had like a white house yeah it looked like a cia thing which

[8:06](https://realalpha.app/v/wOBTi8dAb1Y/?t=486) is also good we love that but the thing is silicon valley is meant to be a meritocracy yeah everyone coddles these things everyone talks about these things like a gifted child like they're afraid that they'll get turned into an ear of corn like that twilight zone episode for saying the wrong thing and that's oh i watched your bloomberg interview recently and you know i used to work at Bloomberg. I'm very sympathetic to them. But their response to some of your criticism was very, but they're smart people who believe this. Like surely it can. I felt a little bit like, is that the best AI defense they're going to muster? So I just want to make sure. I mean, I think, you know, argument number one is that the AI companies talk about what they can do in the future, not what they can do today. Your assessment of what they can do today is not that exciting. To the extent companies are using it, it's often subsidized and or there's sort of this delusional sense of like, clearly we're supposed to get some value out of it. We're going to spend that that's going to come to a stop. What's your view? Anthropic has now confidentially filed to go public. I think you said in our last conversation, there wouldn't be

[9:12](https://realalpha.app/v/wOBTi8dAb1Y/?t=552) exits is now the view that they're going to try to exit, but they're going to pull one over on us. Like, what's your view on, I guess, when does this all come to an end? Like, does it survive an IPO? So I said this on Bloomberg. I don't think regulatory bodies should let Anthropica or OpenAI go public. I think that they are both irresponsibly run companies run by craven liars. Both Dario Amodei and Sam Orman sell their products based on what they might do, theoretically, rather than what they can do. Every time anyone asks Sam Orman how this will work out on CNBC, David Faber was like, So how are you going to get around to the return on investment thing?

[9:47](https://realalpha.app/v/wOBTi8dAb1Y/?t=587) He's like, yeah, somebody will work it out. Sammy, clammy Sammy, this is your job. You are the AI industry. But to your greater point, I think one of them is going to try and go public. I think they're going to bull rush this thing. I think it's inevitable opening if I was to go public. And based on some stuff I might have read recently, they sure shouldn't. They're losing too much money. Anthropic, on the other hand, is also losing just as much money.

[10:11](https://realalpha.app/v/wOBTi8dAb1Y/?t=611) **Ed Zitron:** They had this, one of the most disgusting things I've seen done by these companies outside of the environmental damage, the stealing, the CSAM that Grog generated, putting all that aside, anthropic leaking that they are profitable in the two months that Elon Musk discounted their compute.

[10:28](https://realalpha.app/v/wOBTi8dAb1Y/?t=628) Do we not have an SEC anymore? Do we not have any kind of regulatory body? We certainly don't have a conscience in the valley because everyone was like. Your view is that they made a profit because Elon gave them a deal on their compute. Yes, and that's in the SpaceX S1. Specifically, the May and June in Q2, Anthropic will become profitable. Will, because it's June now. It was not when that story came out. And mysteriously, those are the two months, May and June, when Elon Musk discounted Colossus 1 and 2 for Anthropic. it's just like and what sucks is that's publicly available information this is in the s1 for spacex and people are still like they're profitable now it's like yeah you're profitable as long as you don't don't count the costs and so it's in my view on this is we're hope we will get the answers when the yes i'm looking forward to it and you know we'll talk about the uber analogy because i covered uber yes you were one of the and you were actually one of the few people who were really critical of them. I was very critical, but in some ways I learned the opposite lesson that you did, which is they ended up turning a profit. People over extrapolated from some of the private financials and drew these complicated models that sure, like Uber was spending more

[11:43](https://realalpha.app/v/wOBTi8dAb1Y/?t=703) aggressively and they were, they were like Uber pool never worked. Right. So there were things about Uber that didn't work that were worthy of criticism, but I just, there are just limits to how much you can sort of do financial calculuses off of these like limited private numbers that we see. Do you want to wait to discuss that? No, I don't. Let's talk about the Uber case, because it forms both of our views so deeply. So with Uber, for example, a lot of their money was marketing and R&D. I like to equate Uber to Groupon more than anything. Heavy, heavy marketing, and that was where they did the subsidies for the rights and all that.

[12:18](https://realalpha.app/v/wOBTi8dAb1Y/?t=738) Also, R&D efforts within autonomous driving that I think went nowhere. Forgive me if you're not having looked recently. Also, the burn rate on Uber was just a different planet. I think they burned. Way lower. It was scary then, but now it seems like a joke. It's about $32 billion in history. And Anthropic has raised over $75 billion in the last six months. So I think that that's altogether different. And also the level of burn, the scale of burn, and the linear reality of it. It's a word, I think. Basically, as Anthropic's revenue increases, so do its costs. Except the costs are always up here.

[12:52](https://realalpha.app/v/wOBTi8dAb1Y/?t=772) Same with OpenAI. Uber had a messy cost structure, but they didn't change the unit of work. their unit of um revenue so they didn't have a thing where it wasn't they tried the ride pass where you could just pay one fee and drive as much as you want but you still paid for a ride a customer was still paying to go from point a to point b with the subsidized ai labs subscriptions they are training people and they've trained millions of people to use these things in a way that does not make sense and they pay the costs the reason that you're seeing companies burn so much money is imagine if you were driving your car and you just never really had to think about fuel fuel was just included with it perfectly and then one day and so you actually no let's reframe this use a cab service and it will drive you as much as you want you get 300 rides a month you could drive from la to san francisco you can drive from just the upper west side to the upper east side no matter what it's always the same costs then suddenly one day they say actually gonna have to pay on a mile per mile basis you would have no idea how to plan your life you planned your life around this subsidized unlimited thing and perhaps you could say the the oh the

[14:00](https://realalpha.app/v/wOBTi8dAb1Y/?t=840) ride point to point from the original uber was like that no this is like if every uber ride used to cost three dollars and the new ones cost three hundred five hundred a thousand dollars people do not know how much their ai costs they have no idea how to measure a unit of work because they've never had to think about it like if you were used to just loading up claude going fix these bugs and it kind of does it and then it messes some stuff up and you poke it bonk it on the head a few times maybe it gets it right maybe you mess around with it but you don't know what it costs and you weren't paying the cost suddenly doing that would give you cte look at the people so github copilot which is microsoft's coding tool they just moved from a premium request model to a token-based billing go on the subreddit people are screaming like they're being stung to death by bees because they are finding they're burning through their entire monthly allotment which is just the price of tokens in like a day there's someone who did it in three prompts there's so much i want to respond to this so i'd like to take a move sorry but i do the specific one major contradiction that seems to appear to me in your argument is that people see the cost being high and they freak

[15:09](https://realalpha.app/v/wOBTi8dAb1Y/?t=909) out because they want the product desperately like how do you reconcile with the idea that But it's like people who are bullish on AI see the fact that people want to spend money on it, want access to it as evidence that it is doing well, that they see some value in it. Like, do you not see that? I do see value in it. I'm not disputing that. I'm saying that the value they get out of it is based on the idea of subsidies because they have evaluated it based on it not costing what it really does. So if you use the product and you could do all this stuff with it, however accurate or inaccurate it is, you can say, well, that's a great deal at $200 a month.

[15:42](https://realalpha.app/v/wOBTi8dAb1Y/?t=942) is it a good deal at 300 bucks a week probably not you wouldn't be able to do as much and i also think psychologically you look at it differently if something costs 200 bucks for a month and you run up against rate limits fine you can accept that i want to talk through the uber analogy as i see it i mean I don't know. Yes, I think Uber was a similar case where people felt like, oh, it might work at one price, but it won't work at another. And I think what happened was people got really upset in this amazing period where Uber was overpaying drivers relative to what they needed to pay them to get them to do the work, discounting rides to keep the riders from going to Lyft.

[16:25](https://realalpha.app/v/wOBTi8dAb1Y/?t=985) And basically we had this like free giveaway and there was like a crowd of people online who were like very angry about this even though they were getting stuff at a big discount. My treats. Right. They were getting it. And so then that period ends. Basically Uber and Lyft sort of come to a sort of equilibrium where it's, I don't know, it's probably 85, 15 or even more aggressive in Uber's side. And yeah, I mean basically they have to squeeze drivers more especially on airport trips. but there is a price where people take Uber. It works. It's not like the insane Google level business that people thought. And things like Uber pool, which we're supposed to allow it to sort of like get to scale don't work. So there are pieces. I just don't think there's this like something's valuable at one, you know, it's sort of a sliding scale. And I feel like we didn't learn the lesson from Uber that, you know, that this sort of like big hustle was had on us right but here's the thing when you take from brooklyn to grand central 50 60 bucks it's still within the realm of what you would pay for a cap i am saying that what's happening with ai now

[17:32](https://realalpha.app/v/wOBTi8dAb1Y/?t=1052) **Ed Zitron:** is people went from just for the github copilot example this is not hyperbole there were people who were paying 39 bucks a month and being able to burn 1500 to 3000 dollars worth of tokens

[17:42](https://realalpha.app/v/wOBTi8dAb1Y/?t=1062) that is not the same that's not the same at all that would be if like uber rights were five dollars and then became fifteen hundred dollars or three hundred dollars it's just a different economic scale and also you can plan around that still even though it's more expensive you can say okay this ride was 10 bucks damn i wish it was still 10 bucks but it's 50 bucks now with ai it's you don't know how to measure a unit of work that's the biggest thing like it's tough across models across harnesses use open code cloud code using a deep seat model using jpt 5.5x high plus alpha you're using whatever you're using each prompt oh what are you having it do it might act differently because hallucinations are a mathematical guarantee per open ai so you don't know how to measure the unit of work you also pay regardless of whether there is a mistake or not if it gets it right if it doesn't get it right you are still paying on a per million token basis so it's not like you can say well okay was spending 200 bucks a month now i'm spending 500 bucks a month you're unaware you have no idea the more complex the work you do and because ai code is very verbose the more ai you use the more convoluted things are going to become and the more reliant you are

[18:53](https://realalpha.app/v/wOBTi8dAb1Y/?t=1133) on the machine to kind of work out what it's doing and those costs are dramatic they are unrealistic for most people they're unrealistic for most organizations the models that we have today will be cheaper in the future right that we see that over and over again what do you mean Like the models of 12 months ago fall in price. The new models obviously are new, have new prices, but you're able – the old models become either offerings from Anthropic and OpenAI themselves and or open source competitors basically copy versions of that model and offer them at a discount. So I can see the argument that – and you've talked about this sort of like the CapEx expenditure.

[19:38](https://realalpha.app/v/wOBTi8dAb1Y/?t=1178) The situation with Uber and OpenAI and Anthropoc are very different in the sense that OpenAI's costs are because they're spending a lot training a new model. Well, but before we get on from the token thing, there is an important distinction I need to make, which is models are burning more tokens. So even if the cost of intelligence comes down, the biggest myth of it all, even if a model might be cheaper. In general, they are burning more tokens, Opus 4.8, but more tokens than ever. And every time Anthropic has released a new model, it seems to get worse. I don't know if you noticed that. They seem to be more convoluted.

[20:12](https://realalpha.app/v/wOBTi8dAb1Y/?t=1212) They do things differently. Again, burning tokens so you can... You're saying every Anthropic model is getting worse. If you go and look at Twitter right now and see how people are reacting to 4.8, see how they react. But they're comparing it to Codex. No, they're comparing it to Opus 4.6, man. Like, they're literally comparing it. they've and but this is the thing good example actually even if it's better it is still different which means you have an adjustment to it which means you are going to burn tokens but anyway they are burning more tokens chain of thought reasoning models burn more tokens so even if the model is cheaper on a per million token basis you're still burning way more tokens so you're burning more tokens and spending more money same thing with open source models this has been a consistent problem it's a problem with grok it's a problem with all of them and the thing is but If you're happy as a company with the model today, you can basically find a version of that model.

[21:04](https://realalpha.app/v/wOBTi8dAb1Y/?t=1264) You can make a version of that model yourself. What do you mean to make a version? People can build their own models. There are open source models that people use that they sort of fine tune themselves. There are companies that build their in-house models. It's not like you're forced by Anthropic and OpenAI to shift off the model architecture that you like. The thing is, you're speaking in generalities here. Who has successfully done that? Microsoft just farted out six new models from Gooner himself, Mustafa Suleyman, who is apparently quite a horrible boss. I hear he goes nuts on Slack. But put it all like gossip aside.

[21:39](https://realalpha.app/v/wOBTi8dAb1Y/?t=1299) Gossip is like gossip aside. Obviously, there are a lot of Chinese companies that have produced models. DeepSeek is running the Chinese wear them down process. They discounted permanently just to drag down American models. That's tactical. I love that. It's just funny to watch them do that because it's the kind of thing American AI labs have no idea how to do, which is political rat effing. I'm not going to swear today. But back to the core thing, which is it's kind of like the mileage thing. Okay, gas is cheaper, but you're driving an extra 50 miles, so it's not. It's still the cost of the journey is more expensive.

[22:10](https://realalpha.app/v/wOBTi8dAb1Y/?t=1330) So every model is burning more because the more complex they get, the more things that the whole mixture of experts thing that was meant to bring cost down. It is not more expensive to do the same task. it is more it gets more expensive because people get more ambitious about what task they're going to do no it's actually that because everything i've read is the go on the clawed code sub i love going on these subreddits because you can get a real taste test for it they will find these things new models going in crazy loops the things that could they could do before magically don't work this is a common experience with large language models speak to these users you're looking yeah certainly it is true that when new models come out they have like all these different edge cases and their personalities change their sycophancy level changes so there is clearly a reaction where people are like i used it for this one thing it doesn't deliver on this anymore but like the companies are clearly have every incentive and are measuring their models to make them better over time like they score them i mean just pushing back there are they because are they incentivized or are they delivering i don't either because i don't think they're incentivized to i

[23:18](https://realalpha.app/v/wOBTi8dAb1Y/?t=1398) actually don't think anthropic has any incentive to make the models more efficient same with open ai the markets are rewarding them for being big lossy messy companies and on top of that they have the various business idiots and various companies who will zillow i've been talking to people at zillow for a few weeks crazy ai psychosis over there burning over a million dollars a month in tokens but the thing is these companies are not incentivized to be more efficient and they're not incentivized to bring costs down because people anthropic believes that they have the mandate of heaven i believe that they're a deeply cynical organization and also they're burning more tokens like they obviously are doing it you can people will say oh we're just picking out a few examples are are any of the people Who are excited about this, actually talking about the real cost? Probably not.

[24:06](https://realalpha.app/v/wOBTi8dAb1Y/?t=1446) I agree with you that there's, you know, public companies, all sorts of companies have an incentive to show that they're getting the most out of AI. And therefore, they have a reason to spend aggressively independent of what they're delivering. I think you underestimate that companies do find value in AI. Like, I feel like you... When? Well, literally my own company, you know, our ticketing infrastructure was vibe-coded. Like, we could have paid, you know, for HubSpot or whatever, some SaaS product, and you build your own version so that you can quickly change it. Quick question. Did you pay on a per-token basis?

[24:45](https://realalpha.app/v/wOBTi8dAb1Y/?t=1485) No, I doubt it. You're like, it's just not. That's the thing. You are living in a dream, but who is the dreamer here? Like, this is the thing. you are experiencing it through the larger con which is yeah it burped out some open source software you were like it burped out a copy of an open source crm you have it you have it privately so there's not really i mean i would make sure it's checked by a real coder but you also did not pay for it you paid for the subsidized rate i'm guessing the 20 a month plan maybe the 100 a month i'm sure it's 100 but that's the thing though this is my point you didn't experience the reality of ai you were you're the victim of a con you have been convinced that something is a price that it is not and anthropic only moved companies to token-based billing along with open ai who do it much quieter about three months ago it was announced by the information in april but it's been happening longer and already they're screaming they're like i don't know about this because it's one thing to say wow i'm spending 200 bucks a head and they get rate limits and they can go nuts nuts bananas on this it's another when you have people spending fifteen hundred two thousand dollars a month uber just had to limit their engineers to fifteen hundred dollars a

[25:56](https://realalpha.app/v/wOBTi8dAb1Y/?t=1556) month which makes me wonder how much were they burning before because that's the thing at two hundred bucks a month very different product to one that is a variable cost of hundred bucks a day two hundred bucks a day anthropic's own documentation says that they think that people spend an average of 13 bucks a day in tokens on Claude Code. That's not sustainable for the average person. And I think, ethically speaking, any journalist writing about AI should not use the subscriptions. I think they should only use tokens. No, no, why are you laughing? I'm not even being, I'm not being aggressive here. I'm saying, why aren't they doing that? Because to have a realistic perspective on the actual efficacy of these models, you have to experience the horrors of this thing just went on a little loop and burned five bucks yeah most consumers are not going to want consumers don't want an experience where the product could accidentally you know way overspend and they'll and enterprises are sophisticated i mean enterprises are not sophisticated not always but at least i mean in general i mean but the point i'm making is i think token-based billing is coming for everyone i think it is an inevitability the economics do not make

[27:06](https://realalpha.app/v/wOBTi8dAb1Y/?t=1626) sense for subscriptions everyone does this funny fusion dance of oh inference is profitable no one has proof of that maybe maybe if they didn't pay any training there would be economics that may be lined up but in truth inference and training are one and the same at this point you have to keep post training you cannot stop i mean we're gonna see the number do you think these ipos are not gonna happen they're gonna fail i think make a specific prediction oh man i guess the thing is things are so wacky because anthropic was like we've confidentially as they announced it filed a night for right for rs1 but we're not going to say when we're doing it open ai the journal reported a few weeks because it's burp gen uh was like oh they're going to file soon but they haven't yet i think that there is going to be a reaction based on the spacex ipo if it rips open ai lists if it doesn't open ai slow rolls but also i could see sam altman getting real mad I and saying, I've got to get this, screaming at Sarah Fryer.

[28:04](https://realalpha.app/v/wOBTi8dAb1Y/?t=1684) Right. I mean, to be clear, I think the SpaceX IPO to me is uniquely worrying. It's a dog. To me, it's like the problem with SpaceX is it has failed to be the AI company it wants to be. And the space company has a limited ups. You know, there's just like you can value the space business and it doesn't reach, you know, $2 trillion. dollars um but so where we differ is well i don't know what the correct price of open ai and anthropic is um i mean but that's it's not my job you know it is not to pull the bloomberg but it is sort of the business of investors to speculate about the price and make bets and have sort of this competitive dynamic and that is ultimately what sets prices so i think whoever goes public out of anthropic and open ai stops the other one i think you think it's so bad i think oh i'm confident open ai's is really bad very confident and i think that like the numbers will appear and it's like this is so abysmal i think that the losses for open ai are gonna look like a dog's asshole on thanksgiving i think that i don't know about anthropic but the fact that dario amaday i have to give him credit he is one of the best media manipulators in history he's better than jobs he just he knows how to do the financial stuff with the annualized revenue he's so good

[29:16](https://realalpha.app/v/wOBTi8dAb1Y/?t=1756) it it but you complain about this a lot annualized revenue isn't is useful to convey if your revenue is growing it's not useful for something like api calls though those are not monthly recurring expenses they are there was some numb nuts who spent 500 million dollars on clawed in uh april I think it was reported by Matty Mills over Axios. Your point of view, right? So if I report my ARR this month and then I have a problem next month, I'm going to be screwed if that number comes out because I'm going to go from a great number to a bad number. So you've been complaining about the ARR over and over and over again, but the revenue keeps going up.

[30:02](https://realalpha.app/v/wOBTi8dAb1Y/?t=1802) The problem with using ARR is that the chickens can come home to ROOST. The problem with ARR is not that they can come home. Moments that you can manipulate it in any which way you like.

[30:11](https://realalpha.app/v/wOBTi8dAb1Y/?t=1811) **Ed Zitron:** Anthropic has never defined what they mean by run rate.

[30:13](https://realalpha.app/v/wOBTi8dAb1Y/?t=1813) Not once. And I've read every single story. The information, I think, Sree, Sree, over there, reported that they do it by taking the day's subscriptions, like the monthly subscriptions they have times 12, then their API revenue for the last four weeks times 13. API revenue is a really weird one to do on a recurring basis because that is going to flow. And as you see these cost cuts happening, yeah, that's coming down. so it's just like revenue is how you should report revenue and you should all of also all of these things are non-gap it's the non-gapiest non-gap that's ever non-gapped and it's frustrating because they share the run rate so that it's a really simple manipulative trick you meant to go oh 47 billion dollars of run rate that's how much they made this if they made 47 billion dollars that it's meant it is but these are i mean they're sophisticated investors doing it right i mean what i'm sorry mate no i'm just saying like to a regular person the idea that it's like forward looking revenue might seem unusual but investors are very aware of what like 2025 revenue is versus like taking are they yes i mean are they getting audited financials or are they getting a powerpoint deck that they do in the elizabeth home ceo voice shows good look at this do you think the books are

[31:32](https://realalpha.app/v/wOBTi8dAb1Y/?t=1892) cooked i don't think the books are cooked i think that at the end of every presentation is like the like a font size three series of disclosures that are like on a non-gap basis yeah if they even show i mean they're gonna go public so they will have to reveal gap financials i cannot wait because my god would it be nice to to stop having no i i don't mean this one but you know what i mean it's constantly defending this but here's the thing what you have I've been saying that we are in a bubble since at least 2024, right? I mean, what have you gotten wrong? Like, you've been negative in a period where everybody else has gotten more positive.

[32:11](https://realalpha.app/v/wOBTi8dAb1Y/?t=1931) Oh, I definitely flubbed how quickly it would end. I will fully admit. Like, 2024, I was like, it landed three quarters. I was just wrong. So what? Like, over and over again, you're like, this is the peak. This is the peak, right? And then early in 2025, I very specifically was like, I'm not going to call the peak anymore. you just did call the peak again right didn't you do like it's a bubble the top is the ipos 100 i mean paul kudruski agrees with me he's excellent like you just had him on the show like the google equity raise google raising 84 billion dollars on the public markets like a goddamn middle low tier cap company come on man but the thing that is i agree with you that is a potential bubble sign i think though but not in the cynical way you're putting it sort of like google's like wow people really think we're worth a lot of money like let's get some cash no i mean it's their share they're allowed to get money you know i mean they're allowed to do it that's perfectly legal i'm just saying that they're doing it because the credit markets are drying up that's 100% what's happening well their equity is super inflated i guess we'd both say so they can get a good deal raising money at a high equity price sure but also the credit markets are running

[33:12](https://realalpha.app/v/wOBTi8dAb1Y/?t=1992) like that's what's happening here and that's one of the big things that people really don't want to talk about which is let's say NVIDIA does the thing they're meant to which is they're meant to sell a trillion dollars of gpus by the end of 2027 yeah sell by which i mean put in a warehouse that would i think that's like 15 or something gigawatts of capacity data centers aren't being built at that scale we are watching one of the biggest kickstarters a pre-order campaign

[33:37](https://realalpha.app/v/wOBTi8dAb1Y/?t=2017) **Ed Zitron:** of all time for there to be a trillion dollars of sales we're going to need like 800 billion dollars worth of debt we're going to putting aside cynicism or anything just dollars to donuts we're

[33:49](https://realalpha.app/v/wOBTi8dAb1Y/?t=2029) going to need 800 billion dollars worth of debt and for NVIDIA to keep growing at their current rate in a couple years they're going to have to be selling quarter of a trillion dollars worth of gpus in a quarter in a 250 billion dollars in a quarter and again the cash flow required to do this is so severe that microsoft google amazon and meta all having to race there all having to do dodgy little um what's called like off balance sheet deals putting aside however we feel about ai that is more money than i think is out there people will say all the money's on the sideline there was a story in the ft saying banks are worried about choking on data center debt for this to continue requires an astronomical amount of money and then there's the messy little question of why what does building more capacity do does open if open ai had 10 gigawatts more capacity today what would they do what did anthropic do when they got access to colossus one from elon musk oh they are up the rate limits on claude yay like i mean the model the consensus is certainly that the models have improved yeah most people think the scaling laws are holding true that the more the larger training runs you run the better output you get i mean

[34:58](https://realalpha.app/v/wOBTi8dAb1Y/?t=2098) severely pulled back on pre-training isn't it more the train i mean there are scaling still the training runs obviously they're scaling on inference with chain of thought reasoning and also pre-training is because it's still happening right it's all of the above they are coming up with new techniques to improve the models in addition to trying to run right larger models Then why, here's the thing though, then why are we only really talking about code? Like that's, that's a very, like we really only talk about code and there's, it's because there's a ton of training data for code. There's a lot of people that can post train the models.

[35:28](https://realalpha.app/v/wOBTi8dAb1Y/?t=2128) I mean, I do think like a key, there's clearly a receptivity among like consumers to the argument you're making. And my view on it is that AI requires like proactive work. You know, you have to go out, you have to use the models, you have to find things you want from them. Still, I mean, when we talked last time, you were really emphasizing that OpenAI was like the only one. And now the top three apps, I believe, are like ChachiBT, Claude and Gemini in some order. So it's not just one, like you've argued. in the past. So that is another thing. But the core issue is that there's a large audience of people that are like, I'm not, I don't, it's stupid. I used it a couple months ago. It like hallucinated. And they're not like investing like the energy and time to like get value out of the models. To the extent they experience it, they get it through Google and they underestimate the extent that they're getting LLM answers in the first place. So they're sort of using Google results, forgetting how the product used to work, how it is now. And sort of, it's pretty easy if you don't pick up the app use it and spend time on it to say nothing's happening and so people like this story of they don't derive any value but then i feel like i'm using a fairly like it's

[36:39](https://realalpha.app/v/wOBTi8dAb1Y/?t=2199) not like i'm building you know like i have like my own server at home you know i'm using cloud in a pretty straightforward way and i find that i derive a lot of value from the subsidized thing but also were you what just this is not an insult were you describing embracing no no no i really want to be clear this is not sure doesn't sound very intelligent you are you're front loading this by saying well they haven't made the effort to make these things useful that doesn't sound like ai to me that sounds like a kind of scam it sounds it sounds like something that some that someone on these model companies would say because they know that they can't just give you something use it when i used the first iphone it was one of the first people to buy it state college pennsylvania Yeah, singular, wireless.

[37:21](https://realalpha.app/v/wOBTi8dAb1Y/?t=2241) That thing immediately, we're like, holy crap. Like, this is the future. Using AI to this day, and I've played around with it. I've used it. I've really tried. I've thrown a few models in it. Messes things up all the time. Still does it. This is the same fight we had last time where, I mean, you sort of presented AI as if it had been sold to the world and we've been, like, conned into it. Yes. But the iPhone was significantly marketed. Like, it took the world a long time. Like, most people did not use the first iPhone. There were a million of them in the first year, man. Most of the... Yeah, because it was limited to a single carrier.

[37:54](https://realalpha.app/v/wOBTi8dAb1Y/?t=2274) I mean, I'm sure ChatGPT has more users than the first iPhone did. Yes, but it was limited to a single carrier, singular wireless, and didn't come to Verizon, I think, for at least, I think it was six months to a year. But Apple is the great marketing company. Like, the iPhone was sold to America in a way that, like, the AI providers are fairly clumsy marketers. What are you talking about? Do you think the AI companies are good marketers? My man, what are you talking about? everyone has been talking about ai non-stop for three because as a culture we're interested in no because it's the stuff of science fiction like come on it is not the companies forcing us to do this it is something that humanity eric is in every country in the world is interrogating eric since 2023 we have stories in mainstream media saying we should be scared of these things we've had stories in the new york times saying turn it off because it conned a task rabbit when it didn't

[38:44](https://realalpha.app/v/wOBTi8dAb1Y/?t=2324) **Ed Zitron:** actually do it we have had it shoved in google we've had it shoved in every app we use every single goddamn app summaries generations where we don't need them shoved it's taken over google

[38:55](https://realalpha.app/v/wOBTi8dAb1Y/?t=2335) search it's been crammed in everything every company has been saying well from ai is coming along it's going to take your job every news outlet has been talking about ai ai ai and it's been done on the terms of the company every product is forcing you to apple intelligence that was not a consensual right apple intelligence i'm negative on certainly they forced chat gpt on people they forced ai on people google microsoft puts copilot everywhere how how can you say with a straight face that this was a consensual exploration of this tech people were not told take your nokia 3210 throw it in the trash you'll lose your job take the iphone you pig that never happened there was no pressure campaign like there has been there's been this thing of but my point is that it is a sort of certainly there's pressure like bosses want employees to use AIBs they see value but it is to me when you you say it's sort of like a marketing effort on the part of the companies like I think this is a organic idea that people are excited about something. I think you are ignoring things I think you were ignoring the fact that both Altman and Amaday. The Superbowl ads that they ran, like, An anthropic Super Bowl ad was like not effective.

[40:02](https://realalpha.app/v/wOBTi8dAb1Y/?t=2402) It was like one of the worst ads that's ever been run. Like, they're bad marketers. Except they managed to pretend they weren't involved in the war in Iran. And then I managed to get Katy Perry and a bunch of other people to say, je suis Claude, because nobody paid attention to the fact that they've been working with the government since 2024. But the point I'm making is Altman. That was a pretty big veer. I mean, that's what happened. But we can get back to that in a second. The point I'm making is Altman has been doing the big, scary technology thing. Mira Murati was on The Daily Show in November 2022.

[40:33](https://realalpha.app/v/wOBTi8dAb1Y/?t=2433) Like, this is not something where this is manufactured consent at scale. iPhone didn't happen. I promise you, because I remember being at PC Zone magazine and turning to my editor and going, we should write about apps on smartphones. He went, ah, mate, it's not going to happen. I'm serious. I don't push back on this because I'm just being a hater. Like, I push back on it because it is. it is manufactured consent for a technology and an anti-labor technology at that and more importantly the media fell for it for it to be anti-labor it needs to work no it doesn't why it just needs to be an excuse but if the jobs weren't needed in the first place then what do you mean the jobs weren't needed to fire people that the companies were not finding value from how do you know they weren't no i'm asking you it's an excuse to do what it is an excuse to start chopping off some people arbitrarily in a way that's damaging the companies because they are companies to decide like that companies hire and fire based on the value like i mean i advise you speak to people from meta and oracle right now oracle right now i think like 40 of their renewals aren't happening because they fired the people and forgot to reassign them you have jobs that aren't getting

[41:42](https://realalpha.app/v/wOBTi8dAb1Y/?t=2502) done across data centers they're just hiring and fire executive ai psychosis basically it is so think of it like this you are a guy who makes 50 million dollars of rsus and cash a year probably more on the stock side and you have this chat bot every idea you've ever had it's like you're a genius you're the smartest man you're so that that's exactly right you should fire all of them you don't need hr and these people they get this tool and they're able ai so mobitar fantastic you're very pro like he likes ai fantastic youtube guy has a good point that ai is really good at a demo it's good at making like a very rough thingy and that's all it takes to brainwash

[42:19](https://realalpha.app/v/wOBTi8dAb1Y/?t=2539) **Ed Zitron:** an executive Nick Suresh quote all you need for an executive is to cobble something together looks close enough and these people don't do work most CEOs go to lunch leave lunch and

[42:28](https://realalpha.app/v/wOBTi8dAb1Y/?t=2548) barely read their emails you know one thing I self-driving cars are obviously a very relevant they're interesting yeah and so you know self-driving cars the industry tried to get us very excited you know like a decade ago and they weren't ready because there are a lot of edge cases i think there's similarity to what you're saying with you know the demo idea which is basically a demo it's like it's mostly there and then people way underestimate how much it takes but imagine it all the way imagine if they used waymo and just they just fired all the cap drives every single one it's like that's good enough and then you just every time it rained you would just have the waymos jumping in but then it'll be a great opportunity for every business that doesn't like that that's just like competitive dynamics in business what business what do you mean what do you mean like businesses that hold on to their if your cases that all these companies are going to find that they needed those employees the companies that don't react this way will outmaneuver them like that yeah probably that's competition yeah that's in fact in the erp space of the oracle they're finding a bunch of churn to other erp enterprise resource planning companies

[43:25](https://realalpha.app/v/wOBTi8dAb1Y/?t=2605) that's happening i think meta could die in 10 or 15 years because of the things that zuckerberg has done to it i think microsoft people at microsoft are miserable because copilot copilot co-pilot. Satya Nadella, MBA. Sundar Pichai at Google, MBA. McKinsey. Mark Zuckerberg ran on MBA. Tim Cook, MBA. I think John Ternis is MBA. Point is, these people don't experience problems, so they have no idea how to solve them. Sorry, you're saying Zuckerberg has an MBA? He doesn't have an MBA. He hasn't written a line of code since 2006, apparently. Okay. Which is really funny. But getting back to the main point, which is, I think people credit CEOs with Thank you.

[44:00](https://realalpha.app/v/wOBTi8dAb1Y/?t=2640) Way more intelligence than they have. I also think they think that these guys at the top, like Google, for example, they'd never make a big mistake. Go and use Google right now and tell me if that looks like a good product or it looks like if a startup released a search engine that randomly responded like a weirdo to you in 2014, Michael Arrington would have them shot. I just, I don't get the, like, moral lens. Like, if these people are wrong, they're going to lose a bunch of money. And, like, if you don't like them, you can be happy about that. Like, what's, what's the issue? Like, they're trying.

[44:28](https://realalpha.app/v/wOBTi8dAb1Y/?t=2668) tens of thousands of people losing their jobs the people across the world who have spent three years being scared being told this thing that cannot do it but regardless still as if it could being fired being threatened they open the newspaper every day there's a completely fictitious story like strategy for them like are you long like just like human heavy businesses or like long humans i don't invest in the market i people have this weird thing like oh what are you showing i'm not i like writing i actually care about this stuff and i think people love love that i'm just saying what's like the positive vision for for the world what what do you mean well i'm just saying it's very focused on on sort of like the negative like my view of the ai industry is that people are building interesting stuff that they think i mean i do think it will help with drug discovery research all sorts of not really generative though you're you're mushing there are there are different some of the things that's the thing though but there are different techniques like that's the marketing technique though they want you to think all ai is one blob when ai is many different things and the thing everyone's so your your criticism is specifically text generative

[45:33](https://realalpha.app/v/wOBTi8dAb1Y/?t=2733) generative ai so video video but like the techniques that like allow for alpha go and alpha i do not believe is generative right but there are related techniques involved and like certainly demis like who is creating them is across all this stuff are you skeptical of him Yes. So Demis, Nobel, Nobel scientist. Yeah, but chemistry. And also AI techniques within chemistry, not generative. Every time I see him on stage, I think that in 10 years, the computer is going to wake up. And irresponsible, disgusting. I think it's disgraceful to constantly do this thing of AGI is just 10 years away. You want to know why they're talking about AGI? Because the current stuff isn't good enough.

[46:20](https://realalpha.app/v/wOBTi8dAb1Y/?t=2780) I don't find a lot of coders who don't, like coders seem to believe that this is impacting their career. Like, do you disagree with that? Some are. Carl Brown, the Internet of Bugs, puts it where it's like, it makes the easy things easier, the hard things harder. There are people that use it who just ignore the outputs. And the thing is, you can make really clumsy fudge software that sort of kind of works until it doesn't before LLMs. There was a bunch of software within major hyperscalers that just didn't work. This has compounded that problem. the problem with llms one of the most dangerous things about them is they can make an incompetent imbecile seem semi-competent they can hang around and do more damage it's far more damaging to have someone with an llm who can't code or can barely code who is able to pretend they can and get responsibility than it is for that person to just not be able to do that not to sound like a tech person but what you're describing is you know democratizing intelligence you're you're saying that there's someone who doesn't know how to code that you're giving them more capability than they had before and that there's something wrong about that because you want to keep it as the reserve of

[47:26](https://realalpha.app/v/wOBTi8dAb1Y/?t=2846) the super intelligent coder you are crediting the models with intelligence they don't have you're not giving these people intelligence you're giving them a very complex autocomplete which is still what it is and people love that when i say that you are giving them something that can do things that it's seen before it can do shit off of stack overflow or what have you they can do things from its training data. And it can, when you hit it enough, produce something that will work. Work does not mean stable. Work does not mean functional even. Work means functional enough that no one gets on their back. It is not democratizing coding.

[48:00](https://realalpha.app/v/wOBTi8dAb1Y/?t=2880) To give an imbecile something that is as dumb as it, but able to do a better impression of someone who can do their job. And I think the problem is, is that people conflate those things. If these models were literally perfect, if they never made mistakes, if they were always good, we'd be talking about something else. What would it take to change your point of view? Like, what could come true? They would have to solve all hallucinations forever, which they are completely incapable of. But humans make tons of mistakes. Like, it is very inherent to reasoning, like, one of us is engaged in logical fallacies here.

[48:31](https://realalpha.app/v/wOBTi8dAb1Y/?t=2911) Like, it is normal for human beings to think incorrectly. The mistakes a human being makes, they learn from. When a human being makes a mistake, even if they don't know they've made it, when they find out the error of their ways through trial and error, they can fix it. Some imbeciles don't. Some people don't, but they can learn. Models have no capacity for learning. They have none. The mistakes they make are mathematical certainties. they are not the mistakes a human being can make can be mitigated they can be learned from they can be trained into in a way that models cannot we do not learn just from having knowledge your bar is no hallucination like none no mistakes of any kind why should that this just seems like humans make mistakes like even the best mathematician this is meant to be artificial intelligence they've sold this is the most powerful thing ever that's going to replace 50 of white collar workers why do i have to lower my standards when they keep raising them This just seems like you don't want to have machines involved.

[49:25](https://realalpha.app/v/wOBTi8dAb1Y/?t=2965) Like, it seems like a truly impossible bar. I use the computer all the time. Then what's the problem with AI? Why do I have to accept mediocrity? Like, that's my question. This just is, I'm so dug in that nothing could change my point of view. We're talking about artificial intelligence, man. And you're saying, can't it be a little stupid? You're saying, can it be perfect? You're saying the date is perfect, I will accept it. They're talking about replacing 50% of workers. But most of those workers, myself included, make many mistakes. Like the bar for perfect just seems like totally outlandish to me.

[49:56](https://realalpha.app/v/wOBTi8dAb1Y/?t=2996) It should be best. Based on their promises, no, it doesn't. Based on what they've been talking about. If they had solved this from the beginning, it's like, hey. Who is they? OpenAI. These companies disagree. They dislike each other. They're competing. Yeah, and they're all lying about what they're all talking about some theoretical future. Demise, Sam Altman, Dario, they're all doing the same crock of crap where they're just saying this will do, this will do. when it does this if it like they're all talking in theoreticals if they have from the beginning solved this as a kind of experimental project that we need to be really careful with because it can make so many mistakes and they were like that's basically in 2023 that's how it was that's not there were people who didn't want to release man go back and read the stories i'm so sorry i mean there were people who are optimistic the employees themselves are true believers like there's a whole you know but it was it was called gpt3 like it wasn't branded in the most like Like in this amazing, I feel like you tell this story of like great marketing, they're bad marketers.

[50:54](https://realalpha.app/v/wOBTi8dAb1Y/?t=3054) In the system card of GPT 3.5, I believe, there was a story that a task rabbit had been manipulated by the model autonomously.

[51:03](https://realalpha.app/v/wOBTi8dAb1Y/?t=3063) **Ed Zitron:** But when you click down into the meter study, it turns out that they had actually prompted it, asked it what it would say, preferentially chosen the responses.

[51:10](https://realalpha.app/v/wOBTi8dAb1Y/?t=3070) And then they intimated that they never did the experiment at all. They used that and allowed multiple different stories, I think 14 different stories, to talk about how ChatGPT 3.5 had convinced, maybe four actually, it was the 2023 era, had convinced the tasker and blackmailed them. Anthropic has multiple times done these stories saying Claude blackmailed someone. And what actually happened was they trained it to do something and it did it. It escaped the sandbox. Actually, it wasn't in the sandbox and it didn't escape it at all. So, yeah, actually I do. Don't you want them running safety tests?

[51:43](https://realalpha.app/v/wOBTi8dAb1Y/?t=3103) Like, the tests are meant to see if the models can be manipulated to do damage. Then if they're running safety tests, why do they keep doing weird manipulative things about them blackmailing when they're not actually... I mean, white paper, like, just most Americans, nobody is reading these white papers or people like us. Yeah, journalists are reading them. And then reporting them in the most preferential way possible, which then informs regular people. So yeah, actually, I do think my standards of- Most American- like Americans are largely skeptical of AI. Like they've failed, right? Don't you agree?

[52:13](https://realalpha.app/v/wOBTi8dAb1Y/?t=3133) Like the polling is negative. Americans- Here's the thing. Don't believe in AI. Most Americans don't run companies and most CEOs believe this because most CEOs are disconnected from labor. That's the problem. Venture capital has predominantly funded AI because of things like this. venture capital has failed to do the due diligence necessary because venture capital has uber brain venture capital thinks that burning money is good and i think it's perfectly reasonable to ask for perfection because that's how they're selling it they are selling it as autonomous intelligence this is the in your view a bunch of rich people who are delusion deluding themselves are lighting their money on fire like yes what's the problem like you don't like the rich people who are lighting their money on fire because they're covering our country in data centers ruining local environments gas belching turbines that are ruining black neighborhoods they are stealing from everyone they are threatening everyone constantly in the media and when people say they don't like this stuff someone in the media will condescendingly say oh you just don't get it you haven't used it you need to do this you're so stupid you moron why don't you like data centers

[53:16](https://realalpha.app/v/wOBTi8dAb1Y/?t=3196) yeah i i think that that's pretty bad you're you're the one you call people imbeciles morons Like, I'm not the one doing that. Like, I'm not saying, I wasn't talking about you, man. I was talking in general. But journalists are much more measured, I'd say. Go read the New York Times, mate. Go read how the Times talks to people when they don't like AI. Come on. Where? When? Casey Newton and Kevin Roose, mate. Come on. Look at them on Twitter. Oh, yeah. You're engaged. I actually don't thought. I'm not doing the horse. That is a personality. I'm genuinely not doing horse trading here. I'm just saying it happens.

[53:48](https://realalpha.app/v/wOBTi8dAb1Y/?t=3228) And that has a real effect on people. And I think the tech media needs more empathy for regular folks, but also a genuine connection with how this has been sold. This has been sold as autonomous. It is not. It's been sold as intelligent. It is not. Like, it's sold on a lie. And it's now inflated markets. When we talked nine months ago, you were very negative. My view is that the models have only continued to improve. I think the funny dynamic is that your crowd continues to not see that anything is happening. Who is my crowd? I don't know. I don't know. Anonymous YouTube fans. Like, you're a very popular guy.

[54:29](https://realalpha.app/v/wOBTi8dAb1Y/?t=3269) Right. Right. Yeah. I think you are selling something to a media consumer that they want to hear that nothing is happening. But you've been doing it year after year after year. I think what is strange is that optimism is framed as never, not saying yours here, The optimism, which I actually don't think most of the AI industry is doing, blind hope in technology and appreciation of capitalism is framed as a moral good, that all investment is progress, and skepticism and scrutiny is framed as other and different. Yeah, the fact that you say you're people, but you can't even designate the group makes me wonder who that actually refers to.

[55:10](https://realalpha.app/v/wOBTi8dAb1Y/?t=3310) Regular folks? Who am I? I don't know. You know your audience better than me. use the term no no i'm like who are your i just see them online they're a bunch of anonymous accounts i talk to a ton of people in the tech industry i have a ton of tech works yeah janitors line line cooks yeah i have teachers ton of teachers ton of people in academia i talk to people in the industry all the time so yeah man i think that the thing i represent is an a person who loves technology watching the tech industry ruin it i just think our you know we hunger for what are we creating as a society what is our society moving towards in a world that is post religion for the most part is unsure of what armenia is supposed to be is looking for some sort of forward direction i don't think ai is perfect post religion what are you talking about i i think church church going is down but There are many spiritual people, tons of them.

[56:02](https://realalpha.app/v/wOBTi8dAb1Y/?t=3362) What is your- Anthropic just meant the Pope, man. What are you talking about? Like, here's the thing. I'm just saying the critique of negativity is we need some sort of like, if not this, what are you proposing? Do you think that the AI industry is a hopeful thing? We're going to- Yes, yes. Dotting the world with data centers for a product, ruining local environments, spending more money than there's ever been spent on anything for a still theoretical outcome, for a still wholly theoretical outcome. Do you think that what we have today with large language models is worth a trillion dollars? God's honest truth.

[56:34](https://realalpha.app/v/wOBTi8dAb1Y/?t=3394) Worth a trillion dollars. Because that's what it's cost. All of AI? Yes, with large language models specifically, because that's where the trillion dollars went. Yeah, I mean, we have multiple companies that are going to get valued, certainly. So strictly on an equity basis, you believe. Right, right, right. Yeah, that's how things are valued. What are you asking me? Like, are they going to be valued over the next five years for more than a trillion dollars? It seems very likely. OK, but that's not where the trillion dollars and the trillion dollars of CapEx, like the trillion dollar. It's more than a trillion when you include investments by venture capital.

[57:11](https://realalpha.app/v/wOBTi8dAb1Y/?t=3431) You include the debt as well. Right, right, right. Obviously, the market is saying it's worth way more than a trillion. OK, I don't care about the market. I'm asking you. Yeah. You think that this is worth it. Say it stopped today. And this was where we did. This is theoretical. Say nothing else happened. would you say that this was worth it the trillion dollars that's being spent what we have today forget the air that it was good that it was money well spent is that what you're asking yeah yes thanks for sticking around with our second face off with ed zitron i i think he held too high high of a bar for perfection there at the end uh this is the newcomer podcast if you want more uh please check out the substack newcomer.co if you really want more i have another show cerebral Valley show where I chop it up on AI with my Cerebral Valley co-hosts. It's on YouTube and everywhere else you get your podcasts. Like, comment, subscribe. We're building this channel out. Tell us who you want to hear from next. Tell them to come on the pod. Thanks so much.

[58:04](https://realalpha.app/v/wOBTi8dAb1Y/?t=3484) We'll see you next week.