55 min read
Good journalism is making sure that history is actively captured and appropriately described and assessed, and it’s accurate to describe things as they currently are as alarming.
And I am alarmed.
Alarm is not a state of weakness, or belligerence, or myopia. My concern does not dull my vision, even though it’s convenient to frame it as somehow alarmist, like I have some hidden agenda or bias toward doom. I profoundly dislike the financial waste, the environmental destruction, and, fundamentally, I dislike the attempt to gaslight people into swearing fealty to a sickly and frail psuedo-industry where everybody but NVIDIA and consultancies lose money.
I also dislike the fact that I, and others like me, are held to a remarkably different standard to those who paint themselves as “optimists,” which typically means “people that agree with what the market wishes were true.” Critics are continually badgered, prodded, poked, mocked, and jeered at for not automatically aligning with the idea that generative AI will be this massive industry, constantly having to prove themselves, as if somehow there’s something malevolent or craven about criticism, that critics “do this for clicks” or “to be a contrarian.”
I don’t do anything for clicks. I don’t have any stocks or short positions. My agenda is simple: I like writing, it comes to me naturally, I have a podcast, and it is, on some level, my job to try and understand what the tech industry is doing on a day-to-day basis. It is easy to try and dismiss what I say as going against the grain because “AI is big,” but I’ve been railing against bullshit bubbles since 2021 — the anti-remote work push (and the people behind it), the Clubhouse and audio social networks bubble, the NFT bubble, the made-up quiet quitting panic, and I even, though not as clearly as I wished, called that something was up with FTX several months before it imploded.
This isn’t “contrarianism.” It’s the kind of skepticism of power and capital that’s necessary to meet these moments, and if it’s necessary to dismiss my work because it makes you feel icky inside, get a therapist or see a priest.
Nevertheless, I am alarmed, and while I have said some of these things separately, based on recent developments, I think it’s necessary to say why.
In short, I believe the AI bubble is deeply unstable, built on vibes and blind faith, and when I say “the AI bubble,” I mean the entirety of the AI trade.
And it’s alarmingly simple, too.
But this isn’t going to be saccharine, or whiny, or simply worrisome. I think at this point it’s become a little ridiculous to not see that we’re in a bubble. We’re in a god damn bubble, it is so obvious we’re in a bubble, it’s been so obvious we’re in a bubble, a bubble that seems strong but is actually very weak, with a central point of failure.
I may not be a contrarian, but I am a hater. I hate the waste, the loss, the destruction, the theft, the damage to our planet and the sheer excitement that some executives and writers have that workers may be replaced by AI — and the bald-faced fucking lie that it’s happening, and that generative AI is capable of doing so.
And so I present to you — the Hater’s Guide to the AI bubble, a comprehensive rundown of arguments I have against the current AI boom’s existence. Send it to your friends, your loved ones, or print it out and eat it.
No, this isn’t gonna be a traditional guide, but something you can look at and say “oh that’s why the AI bubble is so bad.” And at this point, I know I’m tired of being gaslit by guys in gingham shirts who desperately want to curry favour with other guys in gingham shirts but who also have PHDs. I’m tired of reading people talk about how we’re “in the era of agents” that don’t fucking work and will never fucking work. I’m tired of hearing about “powerful AI” that is actually crap, and I’m tired of being told the future is here while having the world’s least-useful most-expensive cloud software shoved down my throat.
Look, the generative AI boom is a mirage, it hasn’t got the revenue or the returns or the product efficacy for it to matter, everything you’re seeing is ridiculous and wasteful, and when it all goes tits up I want you to remember that I wrote this and tried to say something.
Ed embodies how anti-AI sentiments slide from “justifiable venting backlash” to “identarian circlejerk.” And he’s putting it all up-front with a headline that concludes ‘yes I am over-reaching, my motte has no bailey, do not offer benefit of the doubt.’ Every post is the same rant about a few companies spending more than they make. He’s forgotten how to talk about anything else. He refuses to tolerate nuance regarding the underlying technology, as distinct from marketroid figureheads making shit up.
No kidding there’s a bubble. When it pops - the tech’s not going anywhere. Local models work fine. Pick any benchmark and they’re only a year behind the behemoths trying to justify their own scale. ‘What’s the next word?’ is the silliest fucking approach to a neural network, and it’ll still do anything you ask. Not perfectly! Sometimes, not even passably. And yet it moves.
Movie studios will in fact use video generators to describe CGI into existence. It’s already happened and nobody noticed. Audio filters will let two voice actors play a whole cast, none of whom sound like a specific human being. Music generators can spit out a three-minute song in four minutes. Bespoke subject, arbitrary genre, finished recording. Four minutes. Nitpicking the quality is kinda missing the point.
Little of that shows up as economic activity because the most interesting applications are local. For some reason nobody has even tried to sell local models. They just hand ‘em out. What people do with them seems to involve a lot of pornography… allegedly a major driver of tech adoption since shortly after the printing press. When corporate studios seek experts in generating video that would be impractical to film legally, I suspect they won’t look too closely at new hires’ resumes.
Local models work fine.
Even these specific doomed companies could slash their VC-powered spending if they let customers do their own hardware. This whole thing only happened because consumer gaming cards are accidental supercomputers. Any of these companies could sell the open-weight models they are currently giving away. It’s just software. Products don’t have to be services.
These companies pursued scale to avoid competition, and it’s mostly worked. But they still got scooped by DeepSeek, by several orders of magnitude. The expected quality for any amount of training keeps going up. Half this money was spent on engineering greater efficiency. The other half was spent doing more of the more-efficient training. See: Jevon’s paradox.
If the hyperscalers cratered tomorrow, none of the weirdos tweaking published models would disappear. That decentralized infrastructure is noncommercial.
If Nvidia disappeared the day after… CUDA is not required for matrix algebra. Their monopoly on this whole mess was already a criminal scheme, fifteen fucking years ago. And Microsoft has demonstrated that quantization shenanigans make CPUs viable, if not competitive.
But you’d never do anything for clicks. Like… re-write the same frothing opinion piece, three times a month, for two years. This man spent three thousand words, in this fifteen-thousand-word article, railing against tech-bro ‘people doubted the internet!’ crap. The existence of success doesn’t prove some new thing will succeed. There. Done.
No need for a wall of text, over and over and over.
He’s making the point that the entire tech economy is dominated by this bubble, and gargantuan amounts of money is tied up in this with no hope of getting any useful or profitable business.
Yet the mainstream press continues to coddle the egos of Musk and Altman and Zuckerberg and Nadella and Bezos and Pichai, as though their business use of this technology is worth the trillions of investment value they have attracted. It’s a fantasy. While that continues, the message is not getting through and the bubble continues to inflate.
For as long as the bubble goes on inflating, yes there is urgent need to keep repeating that message until the mainstream tech and financial press starts accepting it as reality (because that’s what investors read), so that people stop hooking our economies into that bubble.
Yes, he’s been saying ‘it’s the whole thing!’ from two years ago until this very headline… ignoring whatever doesn’t fit his monomania. DeepSeek gets an offhand reference and is not investigated further. Nvidia’s wild overvaluation doesn’t affect AMD, who was not invited, despite making goddamn near the same product. Every company he’s rambling about could disappear, and the tech would still be a big fucking deal.
Ed seems to think the tech itself is a bubble. It’s hard not to reflect on the aforementioned ‘people doubted the internet!’ crap, when he’s not just saying Pets.com and Webvan are unsustainable… ‘it’s the entirety of the web trade.’ Like it’s as empty as NFTs.
Meanwhile, you can download a thing and it does the thing. A startling variety of perverts have demonstrated that diffusion models work. That shit’s gonna follow the same trajectory as CGI, from ‘ugh, they used computers’ to ‘holy shit how’d they do that.’ Language models are a lot dumber than we’d hoped, but the fact they work at all is borderline miraculous, and they will half-ass anything you ask for.
None of this is going to make anyone a trillionaire. But knowing that is a long way from becoming a professional hater.
I know it’s a lot to cope for you, but you really need to calm down if you want to be taken seriously.
Tone policing is trolling, even when it’s not spat at polite and impersonal criticism. It’s a no-effort, all-purpose dismissal you’re liable to repeat in response to this comment, because the nature of bad faith is that there is no right answer, and engaging with arguments would be hard.
That’s almost entirely because of CUDA, which is Nvidia’s proprietary code for parallel GPU computing. AMD has OpenCL, which you don’t hear about, and only managed to get CUDA running on their stuff last year
CUDA-on-AMD has happened half a dozen times, but it keeps getting shut down over legal fears.
CUDA is Nvidia’s worst anticompetitive abuse.