Did you ever doubt him? Nobody knows the precise timeline, but it's never looked worse for AI shills
>>109941842I have no idea who that is but just from that image alone I can tell he sucks more dick than Facebooks AI division.
>>109941842>but it's never looked worse for AI shillsLMAO!!! AI is stronger than ever, are you retarded?
Literal who.
>>109941846some grifter who makes a living shouting about how ai is useless, will never get better, and the bubble will burst next monthhe's proven wrong all the time, but he acts like he isn't so he's turning into lolcow I guess
>>109941867He's right but where he's wrong is that they won't let the AI bubble pop because there are too many strategic interests at stake.The US govt wants to turn this into a wrecking ball against China so they'll do whatever possible to keep the hype going indefinitely like that Snowpiercer train
>>109941903he's a brainlet grifter
>>109941842most retards that think ed zitron is retarded are correct that he doesn't understand capabilities, but that isn't really his main argument. And if you can't parse his argument, then you are losing chess to a dog bro
>>109941842OpenAI and Anthropic in 2023:>so yeah we’re gonna spend a ton on compute and probably won’t earn that much until 2028 ish>hyperscalers start building tons of data centers and buying chips>2026 arrivesthe goyim:>OH MY GOD THEY’RE LITERALLY TRILLIONS OF DOLLARS IN DEBT!Hey retards. They’re literally earning more revenue than they predicted. Did you think the investors were completely baffled by this year’s events or something? What is like being this retarded? If you think you are so clever, then you should be able to easily calculate when they’ll run out of money, right? And bet all your money that the bubble would pop sometime around then? And become a millionaire? You can do that, right?
>>109941903>He's rightNo he objectively just isn't. If your timing is off that's still you being wrong. That's like Euro sceptics screeching about imminent collapse for 30 years, yeah at some point the Euro will collapse, every single one of those niggers was still wrong and they know which is why they never bet any money on their great predictions. """"Predictions"""" aren't impressive when you just shout the same shit twice every hour for three decades.
>>109941932I mean he's right about the fundamentals not being right, but he's assuming at this point the market is efficient. When the whole thing is being run with media stunts and exaggerations and controlled leaks.As long as they can fool companies to pay for textual autocomplete, it works.
>>109941969>brainlet feeding off of brainlets
>>109941903>why won't the market economy just stay OBJECTIVELY CORRECTPermabears say this completely unironically as they expect stock prices to go down during record inflationtranslator's note: inflation means prices go up
>>109941842nipples protrudingvery very disrespectful
every single tech company in the world didn't just switch to using LLMs for everything just for things to go back to the old wayif a bubble pops, the next gen of LLMs will just pop back up a year later
>>109941842I think you mean SI. Like sigh.
CodeTRANS look like THAT?!
>>109941930Lmao brainlet retardcome back when you pass elementary school math.
>>109941867>>109941908He's literally staked his entire reputation and quit his successful PR business to go after big tech. It's literally the opposite of a grifter
>>109944839you can't be this retarded if you figured out how to use a web browser
>>109941842Who the fuck is this guido?
>>109941867>pretend to be stupid>people think you're serious>make money off it while they laugh at youall too common these days
>>109941903>won't let the AI bubble pop because there are too many strategic interests at stake.That's why it can pop. I'm not saying will just can. Too many chefs.
>>109944580not an argument. I have a math degree
>>109941903lmao then we riot
>>109941842nah bro, ai is fine. they're just gonna dump another trillion in it and stonks gona stonk
>>109941842>never looked worse for AI shillsthey're literally swarming the board with damage control
>>109941930>They’re literally earning more revenue than they predictedyeah, on one day in august, after X subsidized their cost once and multiplying that by 365 lol
>>109946838*SI
>>109946851No, you are literally making shit up. They are earning more annual revenue than they predicted.
>>109946873no they arent also>revenuelol
>>109946851>yeah, on one day in august, after X subsidized their cost once and multiplying that by 365 lolshit like this is why I'll never take there market seriously again
>>109946873>more annual revenue than they predicted~40b loss~1t valuationinference still unprofitable
>>109946897>>109946929Notice how you weren’t doing all this complaining years ago when all these numbers were already predicted, and this was BEFORE the massive gains in efficiency, and BEFORE everyone knew that AI would solve 500+ open math problems, and excel at computer use, 3D design, etc. So where the fuck were all you trannies 3 years ago? Why are you complaining NOW? I know, it’s because you’re a bunch of low-test cucks who go along with whatever you’re conditioned to believe, and right now it’s extremely popular to hate AI, NOT because it’s useless, but BECAUSE it is now obvious that it’s powerful, and this is making furry artists and tranny coders lose their minds. Their only cope now is to attack the financials, which are literally BETTER than the planned investments that the investors willingly signed up for. You are all so desperate and retarded.
>>109946962woah meltie
>>109946814>I have a math degreeIQ status: sub-65I'll take fries with that.
>>109941842I've never doubted Ed Zitron. How dare you make that accusation. The AI industry (and modern American big tech in general) is a scam.
>>109941930>>109946962None of the mental children here can appreciate these fact because feels > reals
>>109947285so when Ed said LLM’s peaked in 2024, you didn’t doubt him? You still don’t doubt him? How about when he said the bubble would pop “no later than Q2 2026”? Or how about just one month ago when he said OpenAI would need to attain $60B annualized revenue in the next 3 months, calling this ridiculous, and yet they’ve already reached $70B? You still don’t doubt him? How about when he went on the DOAC podcast and asked “why aren’t you trying to slow down AI?” to the guy who literally wrote the book “If anyone builds it, everyone dies”? This is the guy who apparently only uses AI to fix his son’s Minecraft game, who is no expert in finance, math, or technology, and you don’t doubt him?
>>109946962delulu
>>109947318So you're saying he's right about 70% of everything he said and he shouldn't be trusted over the people who were right about 0% of the things they said? Or are we supposed to trust you, who suffers from severe AI psychosis, and doesn't understand the meaning of openai's scam-tier reports?
>>109946962> Notice how you weren’t doing all this complaining years ago when all these numbers were already predictedEveryone just assumed they weren’t retarded enough to keep spending like that when no money was being made. Turns out they’re more retarded than we coule have ever anticipated by a whole lot and it’s really wild to watch them double down over and over even when they owe the house generational amounts of debt and keep trying to let it all ride on black unsuccessfully.
>>109947427Im certainly not going to listen to someone who ignorantly declared “LLM’s won’t get any better” a mere 2 years after ChatGPT was released. And it’s one thing to say “I think the bubble might pop around x time” but he said “NO LATER THAN” Q2 2026 with all the confidence in the world. What HAS he been right about? Reporting public numbers that everyone has access to? Woah, that’s some real impressive stuff. This guy has been ignoring the profitability of inference for months now, and rather than addressing it, I think he and his cult are transitioning to the ultimate cope: “ok even if inference is profitable they still need to train models forever”. These people think that if you don’t keep training at the same scale, that the model will suddenly forget how to do math or something. If you ask them to calculate exactly how much money these labs need to spend on training, even assuming they reach the ideal LLM that’s good at everything, they have no fucking clue, they’re literally just operating on vibes, and their feelings tell them to hate AI and tech CEO’s. They will cry for years and years even when they know their opinions are outdated because it’s exactly what trannies like you want to hear.
>>109947318No one knows what the future holds. Everything is still yet to be determined. You're free to bring up whatever metrics you can find, but that doesn't change the unfixed nature of our reality.When there's a growing industry building itself on the idea that their product is the future of technology, and is going to spearhead the biggest cultural shift in history, and that you need to get on board now or get left behind, that should be met with skepticism.
>>109941842When Anthropic IPO bombs, the bubble will literally pop and burn to the ground.
>>109947504Uh oh, melty!Take your anti-psychotics!
When this guy is proven to be dead wrong he’s just gonna start complaining about global warming or AI dangers>i didn’t think AI could be this smart but it’s too much! Ok yeah it’s profitable now but like, where is the humanity? Why are we entrusting everything to this… machine?and his cultists will keep eating it up
>>109947683>unironic schizo actually thinks any of the ai companies are profitable in any way, shape or form todaymy sides
>>109947696>todaythat’s why you are, and will always be, goycattle
>>109942422>nipples protruding>very very disrespectful
>>109941903They haven't let a bubble pop since 2000
>>1099478042008, but they kicked that can down the road.2015-2016 shadow crash is when they mastered the art of not letting a bubble pop, but as we all know it just means the next one is more dangerous, and so forth, until there's no containing it.
>>109942463>every single tech company in the world didn't just switch to using LLMs for everything just for things to go back to the old wayThe LLM will gather dust in the cupboard next to the Google glasses, the Oculus headset, the 3d printer, le Metaverse, le blockchain, etcMeanwhile>every single tech company in the worldwill be too busy playing with the next earth-shattering TRANSFORMATRIX to notice
>>109947729Give me libertyGive me fireGive me $800B for data centers or I retire
>>109947848I’m screenshotting this because of how absolutely retarded it is. I will post this screenshot on this board every single day for the rest of time to remind people just how retarded humans can actually be
>>109947875Be my guest
>>109941842Who?
>>109942463I'm still waiting to see that Photoshop coded from scratch using AI.Nobody's done it?Wouldn't they want to sell it for like one tenth of the price and get megabucks?
>>109946962>AI would solve 500+ open math problemsThey bruteforced them by just combining all possible language terms and then having humans pick whatever looked right.
>>109947784lel at the cameraman fighting the need to zoom in
>>109947318Again his problem is that he thinks about this rationally, as if the market should be efficient.But the market is scammed by techbros with infinite money glitch using every possible trick to lead everyone by the nose. And telling politicians if you don't help us keep this grift alive, we're all going down and China wins.
>may we see these profits?>may we see these active datacenters?>may we see these optimized native software?and openai's already throwing out that 500usd/month sub lmao, the (((subsidy))) is running out
>>109947585I predict this stuff will still be around, but it will all be raw API rates, no subs or public free service, and it'll be this little niche thing at the back of Azure, AWS, etc's offerings for anyone genuinely interested in spending that much money.
>>109941867How is he a grifter? What's he trying to sell outside of his podcast?>>109941903>The US govt wants to turn this into a wrecking ball against ChinaAny talk about an "AI arms race" between countries is a farce, and a laughable one at that. Trump does not give a shit about AI outside of it being the only thing keeping the economy from total collapse, and is just juicing it for as much money as he can squeeze. Ffs he literally decreed it be renamed "super intelligence" and Jensen Huang complied like the pathetic cuck he is. It's just another pump and dump in the endless conga line of pump and dumps that is the current US economy.
>>109941908he's a hero
>>109942463>this copeDangerous for AIggers
>>109941842this guy just writes financial collapse fanfiction for the retards in his podcast subreddittaleb said it best - if you’re unwilling to have skin in the game your opinion doesn’t matter
>>109951920That makes zero sense. If anything your opinion is worthless if you have skin in the game, because you'll say anything to keep this shit from collapsing.
>>109953152The spectator sees more of the game. Le "skin in the game" making one's decisions better is a rationalisation made by those in the game, i.e., the suckers, the mugs
>>109953152>If anything your opinion is worthless if you have skin in the game, because you'll say anything to keep this shit from collapsing.So that other anon is basically saying that Ed is a grifter who will shift the goalposts of his predictions every time that they are wrong and there will be no consequences from being wrong because he already has an audience that will keep eating his slop until the day his sliding predictions aligns with a real reckoning in the industry. Since there's nothing for him to lose, you can't trust that he actually believes what he says
>>109953362He has his credibility to lose, the same as anybody elseHe also loses the profits on the AI investments he doesn't makeIt was a bad argument, btw who is this guy people are talking about? Why does anyone care what he says?
>>109953362He's only been wrong on predictions of when the bubble would pop because he isn't clairvoyant. Nobody could've predicted that the markets would be so irrational that they would spend TRILLIONS of dollars on the world's dumbest boondoggle. If anything most of the stuff he was wrong about was from how he didn't want to believe that this bubble is being driven by essentially a cult up until recently. He still provides a ton of evidence that is sound though. His leaked IPOs were 100 percent legit seeing how OpenAI both never sued him for libel and pushed back their IPO in response.
>>109941903If something isn't capable of functioning as a wrecking ball why would you try and force it to be used as one?Oh, right, people on salaries will never admit to being wrong as it would risk losing their salary. The entire planet has been captured by salaries.
>>109953576>only been wrong about the bubblelol, no. He also said that LLM’s peaked in 2024. He also said, just one month ago, that it would be ridiculous for OpenAI to achieve $60B annualized revenue in the next 3 months and they’ve already achieved $70B. In February 2025 he said that Anthropic making $34B in 2027 is “laughable” (and now they’re on pace to already make at least $40B in 2026). Again in 2024 he said that these models aren’t getting more energy-efficient (GPT-6 Luna can solve difficult math problems for $0.004). When all these things are pointed out, especially about the capabilities of models, his cultists say, “well he was always just focused on the absurd financials”, which isn’t true, as he also tries to predict the capabilities and energy efficiency, saying that LLM’s were a dead end and that weren’t going to do anything more than what they were already doing (he has recently conceded that they’ve gotten better). And even on the financials, he’s also been wrong! This guy is a clown
>>109948371lmao, I hope this is sarcasm. Even Terence Tao, who hates that AI is progressing so fast, will tell you that this is not just brute force. It’s hilarious that you feel the need to resort to this cope, the most extreme cope you could possibly choose. Just accept the fact that AI is now excellent at math, even though people like you made predictions a few years ago that turned out to be horribly wrong. And I think that’s why you have such a hard time accepting reality. Your ego won’t allow it.>>109948387what a convenient excuse lol. >they’re using infinite money glitches! It’s not fair!Is that why Ed was completely wrong about several predictions? >>109956861Did the AI companies forge all their numbers? Have they actually not improved the models at all, but are just running them at 10x the previous expenses, burning trillions of dollars and keeping it quiet? Is that actually what you believe..? It’s easy to say “he’s being rational” but he’s simply wrong about so many of the basic facts that he isn’t being rational at all. He’s a retarded goy who doesn’t understand the importance of capex and long-term profitability.
>>109951801> What's he trying to sell outside of his podcast?the $70/year newsletter, of course. You know, the one he advertises on his subreddit constantly.> Any talk about an "AI arms race" between countries is a farce, and a laughable one at thatWoah, great argument there. Yeah, we would all be fine just allowing China to have the best AI in the world, you’re right dude! Let’s just send all our money to China.
>>109941903>they won't let the AI bubble pop because there are too many strategic interests at stakeEven if they don't want it to pop the current US administration demonstrated it's retarded, they'll somehow find a way to fuck it up harder.
>>109957076>>109957472the other thing that I've noticed about Ed's arguments is that he makes one enormous category error and never gets called out on it by anyone in his audience - he constantly makes the assertion that demand for AI is flat, completely illusory or declining with absolutely zero supporting details or evidencethis invented fact supports most of his other arguments - if the data centers are being built to support a zero growth environment, then they won't be utilized at all - this has been completely disproven with regard to companies like Lambda signing long-term agreements for older GPUS (forget if it was A100s or H100)
>>109947480> when no money was being madethey’re literally making more money than was predicted. Ed called it “laughable in many ways” that Anthropic would make $34B in 2027. They’ve already surpassed a $65B annualized revenue in 2026!
>>109941842never ever EVER trust a man who suffers from gyno.
>>109958590you know he just buys budget whores in vegas and they have to deal with his disgusting nipples
>>109958579He does the exact same thing for the cost of inference. Watch his DOAC podcast around the 39:40 mark when he was pressed on the true cost of inference after suggesting it costs $3 to do anything with ChatGPT. He immediately changed the conversation because he had nothing to support his case. Literally, he said “gosh it depends on the model, now I want to talk about something else you said”. Even now he’s still saying he doesn’t believe inference is profitable, but according to what..? He simply doesn’t want to believe it (or he knows the truth, but is grifting). So he and his followers have this narrative that inference is and will always be too expensive to serve, and that’s why they think this will mostly disappear in a few years.
>>109958661Some sources about inference margins below. Note that they will only improve with better chips designed specifically for LLM’s and the many other optimization methods. https://perspectives.nvidia.com/ai-infrastructure/total-cost-of-ownership/task/faq/revenue-per-rack-ai-inference-datacenter-economics/https://bunbunlabs.com/notes/when-inference-has-to-pay-for-itself/https://www.tensoreconomics.com/p/llm-inference-economics-from-firsthttps://mlkan.substack.com/p/how-to-estimate-gross-margins-forhttps://clusterbid.com/blog/ai-inference-margin-analysis-gpu-provider-economicshttps://multigrid.ai/learn/inference-marginshttps://www.aeaweb.org/articles?id=10.1257/jep.20261506https://www.vistaequitypartners.com/insights/inference-economics-enterprise-ai/https://arxiv.org/abs/2506.04645https://www.theinformation.com/articles/16444But even you use Ed’s own numbers, in 2025 OpenAI paid Microsoft $8 billion for compute. They received 2.5 billion messages a day, so if you do the math, each message only costs about $0.01 on average, and that was with the efficiency of last year, before caching improvements, looped transformers, and so on. So we can get a rough estimate of inference without even testing the specific hardware. But now the cultists will say it’s all a conspiracy theory or something
>>109958661most boomers that I talk to also have this weird abiding belief that the cost of inference is either heavily subsidized and going to be subject to an imminent rugpull - I think it's largely driven by horror stories they hear on their boomer podcasts about how Leo Laporte got hit with $1000 in Fable costs because he wasn't keeping track of his spendI don't understand how the coping got so intense that they deny the fact that the actual cost of almost all modern technology has consistently declined since the beginnings of the industrial revolution but "it's different this time" because cory doctorow told me that everything is getting enshittified or some other absolutely moronic drivel
>>109958742Doctorow, Cal Newport, Ed Zitron, these guys are just trying to stay on brand with their anti big tech ideology. Ed even admits that a few years ago he was asking himself “what do I even write about anymore” and now he’s taking advantage of the anti-AI stance (which he continues to fuel)
>>109958817the biggest problem for Ed is that AI and agents pretty much make PR obsolete so he's a "primary researcher" now kek
>>109958851>>109958817This is what profound AI psychosis do to a pajeet
AI Bored Ape NFT is the final destination.
>>109958858I'm white
>>109959012And I'm the Queen of England (alive and well).
>>109959016cool how are your corgis doing
>>109959045I was starving, all alone in my cave, so I ate them and had a coat made of their skin.
>>109958858kek you lost, Chang
>>109941846The T-1000 from terminator 2
>>109941930I don't understand how no one gets how the modern economy works yet. Debt literally doesn't mean anything these days if your company is still immensely valuable.
