> February 2025: "I will keep writing this stuff until I’m proven wrong."
Wrong (Zitron continues to write despite repeatedly being proven wrong)
This did give me a good chuckle
How accurate have Ed Zitron's AI skeptic predictions been?
851–860 of 1001 posts
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#852Earlier quoted context omitted.
re: substituting with cheaper models. I think some people here are oblivious to how much of Anthropic and OpenAI’s usage is artificial. Take this one example of a user with 15 Codex subscriptions ($3k) generating $60k in API-equivalent usage per month: https://hraness.com/writing/my-girlfriend-asked-me-why-i-hav... “there’s a once in a lifetime discount happening at the OpenAI Intelligence Depot, and I brought 15 sho…
It goes a layer up, too. Harvey (legal AI) recently tweeted that a single user query cost that firm 26k USD. This, while they have thousands of users and reportedly whole law firms paying zero. That difference is being subsidized by billions of VC dollars.
A bit of a collective action problem, but if everyone leverages this services as hard as they can without cost concern, better capital decisions will eventually be made. Like the Federal Reserve draining the M2 money supply, AI users must drain the capital supply of subsidized tokens. This pulls forward the future of "Does this tech have value at actual unsubsidized costs?" If it does, tremendous, if it doesn't, also a reasonable outcome to the grand experiment. The current pain comes from the valley of uncertainty we find ourselves in at the moment.
> That difference is being subsidized by billions of VC dollars.
"History never repeats itself, but it rhymes." -- Twain
Doordash and Pizza Arbitrage - https://news.ycombinator.com/item?id=23216852 - May 2020 (514 comments)
> I cut this deal with my neighborhood Italian restaurant! I texted the owner about being miffed they hadn’t told me they were on DoorDash. He replied. They aren’t. We compared pricing, and found the prices advertised are way off from what the restaurant charges. So I placed a $5,000 order to the neighbourhood homeless shelter. DoorDash paid him over $20,000, and I get free pasta for the rest of the year. (My neighbours have also partaken.) Glad to know it’s scaling. SoftBank has assembled a unique concentration of stupidity for itself.
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#853I think cable news channels like CNN and CNBC like to have a "counterpoint" on to make their coverage seem balanced. There's always a guy saying the stock market is about to crash. There's always a guy saying oil will hit $200 per barrel. There's always a guy saying AI is a bubble.
Zitron is never right but he exists so that media can claim to be balanced.
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#854Earlier quoted context omitted.
Why do everyone assume they are subsidized? When we seemingly have no idea what it costs? Maybe average subscription is breaking even and token spend is pretty much pure profit? Case in point, claude code seems hell bent on increasing usage at all cost. Which makes sense in the growing phase (get people hooked) but it does not make sense given the hardware shortage. So, which is it?
Anthropic admitted last year to losing money on inference, it had negative margins. The margins have improved and are now positive but there’s still a significant cost. If plans aren’t being subsidized it would mean that the margin on inference is ~99%+ which would mean OpenAI and Anthropic should be wildly profitable but both are still losing money. So, it’s mathematically impossible that they’re not subsidizing pla…
The last I saw with Claude was that the plans are subsidized somewhere around 10x. The usage of Claude and similar products (for people not paying the actual API token costs) would be lot less if they were paying 10x more per account/seat.
A lot of people are using Claude and ChatGPT for all kinds of minor things at work, and they probably wouldn't be if they were paying the true cost of the product. And this is all the while their work product is suffering because AI is not a great fit for a lot of use cases.
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#855Here's where we're at:
Tech companies invested in AI: AI is amazing (so good, it might kill us all, help?), and the ROI is going to be amazing.
NVIDIA specifically: We will allocate cash in strange deals that makes it looks like we're making investments but actually we're buying sales from those companies as they'll spend the money back with us, and everyone seems to think that's OK because the share price is going stinkers.
Cynics: AI is a scourge on society, and the fiscals are circular and a scourge on the economic stability of the planet.
The more nuanced argument I take is that Zitron is right about the circular finances and hyperinflation of valuations, and wrong about the utility of AI to help people. The Tech companies are just outright wrong about the finances, but right about everything else. And, NVIDIA's time will come, and it won't be pretty.
None of this requires me to believe that if Ed says something it must be true or false. It's just another data point, and he does uncover information I can't find myself that checks out. The fact he is wrong about the functional utility of AI or that he's getting super focused on hyperscalers when the real issues are in Oracle and NVIDIA, I can live with.
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#856Earlier quoted context omitted.
Irrelevant. If AI turns out to be a short-lived fad, a bubble that pops and we all laugh at in a few year's time, Google's investments in AI were simply wasted, it does not follow that Google itself will die from having made the investment. Same for Meta, whose current name is due to a previous bad investment we laugh at. The popping of the investment bubble of AI, that might kill Tesla and SpaceX, perhaps also Anthr…
> If AI turns out to be a short-lived fad, a bubble that pops and we all laugh at in a few year's time I think you can be even more precise here, because “AI” can be a wild success, and the Anthropic and OpenAI product strategy can blow up at the same time. Transformers, LLMs, and harnesses are all technologies with amazing utility like many machine learning methods before them. But there may be no moat that justify…
Sure, yes.
Given the rumours that Anthropic will copy SpaceX with value-by-TAM, I suspect they'll be overvalued even if they did have a moat.
My standard example for why not to price a stock by the TAM is how Wikipedia's not valued at [number of people online] * [peak price of Encyclopaedia Britannica].
> VR still has a path to utility / viability.
Perhaps, but IMO it's a new form factor of games console, nothing more than that.
And the Meta vision was broader, "the Metaverse", without really exploring what that would look like in practice rather than as piece of SciFi world-building ripped equally from Snow Crash and Ready Player One.
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#857Things in general I think he's right about: 1. Revenue if anthropic and openai is unlikely to grow to the high levels they need to pay for their commitments. Many of their heavy users (coding) will eventually offset a lot of usage to more efficient and cheaper open weight models. I know of people in a company I was at that spend thousands of dollars a month on tokens. I am sure that what they're using it for can be s…
I feel like too many people have a binary vision of the world, i.e. you're either "pro AI" or "anti AI". Ed Zitron doesn't say that AI doesn't work, that its impact on the world will not increase. He's basically saying that (1) the current data center investments are based on unrealistic revenue projections and (2) hyperscalers are using accounting tricks to move around "money" in a circular way to make it look like…
If you ever listen to more than a couple of Zitron interviews, it's clear that he's not the "reasonable centrist" viewpoint on AI. He's the doom-and-gloom guy.
Maybe there is someone out there positive on AI itself and calling out reasonable objections to some of the extreme things. But that's not Zitron.
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#858Earlier quoted context omitted.
It would matter if we had politicians who acted before bubbles collapse instead of bailing out the perpetrators after the fact.
That will successfully prevent 200% of market multipliers: bubbles and just plain sudden rapid value creation. It's like cutting out the prostate from all men at age 30 to prevent nearly 100% of prostate cancer.
If actions are extreme, perhaps. But that doesn't have to be the case.
We have made efficient frictionless markets a holy cow. The appropriate response is not to eliminate them.
The better part of "rapid value creation" is not the rapidity.
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#859Earlier quoted context omitted.
It goes a layer up, too. Harvey (legal AI) recently tweeted that a single user query cost that firm 26k USD. This, while they have thousands of users and reportedly whole law firms paying zero. That difference is being subsidized by billions of VC dollars.
A bit of a collective action problem, but if everyone leverages this services as hard as they can without cost concern, better capital decisions will eventually be made. Like the Federal Reserve draining the M2 money supply, AI users must drain the capital supply of subsidized tokens. This pulls forward the future of "Does this tech have value at actual unsubsidized costs?" If it does, tremendous, if it doesn't, also…
Yes, there are good examples of this working, anti-competitive as it may be.
What if, in the DoorDash example, the pizza shop was itself venture backed and selling pizza at a loss to win customers and using this 20k revenue as a basis to raise money?
Re: How accurate have Ed Zitron's AI skeptic predictions been?
#860Earlier quoted context omitted.
That's not defensible.
It absolutely is. What has improved isn't the models, it's the harnesses. Give GPT-3.5 a 1M context window and a modern harness, and you won't see any meaningful difference with Opus 5. It's a bit hard to try with such old models, but for example I use Opus 5 / Fable at work and Sonnet 4.5 at home (because it's free via Amazon Q), and there's absolutely 0 difference in performance. None. Obviously 4.5 is only a year…
>Give GPT-3.5 a 1M context window and a modern harness, and you won't see any meaningful difference with Opus 5.
>models are currently regressing. Opus 5 is much much worse than Opus 4.6
I'm in sheer awe at these takes. Literally beyond parody.