Join the discussion
Write your take first — we'll ask for email only when you're ready to publish.
- Hacker News
- I more or less agree. When Apple starts using Chinese and expanded South Korean production memory chips, their hardware pricing problem will go away.
I am running new betas for macOS/iOS/iPadOS and Siri is actually useful for a much wider set of use cases. I asked Siri last week what models it was using and one of those listed was Gemini which is confusing because I enabled free use of OpenAI in the settings. Regardless, Siri is much more useful than it used to be.
- Needs an ( “says Ed Zitron”) title change haha.
This is an interview with Ed Zitron.
by sailfast - > While people get some sort of benefit out of AI-generated code, these tools actually end up making them slower
Curious that he references a METR study from July 2025, before the leap in model and harness performance towards the end of 2025/early 2026.
by scrlk - I don't know if his final analysis is right or wrong, but if he believes this, he's completely clueless (about this aspect at least).
Edit: to be clear, I am talking about Ed’s contention that AI coding isn’t net productive.
by loumf - Welcome to Ed Zitron. There is a reason this man doesn't heavily short the same companies he criticizes. Be wary of anyone that won't put their money where their mouth is.by solenoid0937
- I stopped reading at maybe 25% into the article. It's obvious garbage.
He selectively quotes the max theoretical enterprise pricing equivalent of fully using the private subscription. Did SemiAnalysis not also claim a very high margin?
He throws in doubt about the providers having decent margins, which he claims is made up by "AI boosters" rather than leaked financials and open-weight pricing.
Then next he talks about "the real cost" of inference as if it was in any way realistic that labs price the enterprise plans near cost, like he seems to imply.
Then next he claims AI is actually slowing developers down and there isn't much difference between the models.
It just seems delusional.
by user43928 - Interestingly, that METR study has updated data for 2026 that shows a speed up, although they admit that the data may not be reliable because of changed pay rate for participation, but this quote is telling:
> The primary reason is that we have observed a significant increase in developers choosing not to participate in the study because they do not wish to work without AI, which likely biases downwards our estimate of AI-assisted speedup.
So compared to just last year, they had a hard time finding participants because too many didn't want to work without AI.
by thewebguyd - There is a difference between "AI is overvalued" and "AI isn't valuable". We've seen entire industries deliver real tech progress while still going through harsh valuation resets.by kiaansaraiya
- This is quite a short sighted analysis. I do think the valuations are quite and they would need to meet the reality, but don’t think there’s gonna be a crash or we’d ever go back to pre-AI era. It would more or less would be a correction to valuations.
The future of AI would be on-device models which are as powerful as current frontier models and also I can imagine companies have their own deployments of inference of open weighted models for most of the use cases and use the frontier models for extremely niche or higher intelligence tasks.
As an example I use Claude code heavily for every day development and Opus 4.8 was already good enough for my use cases and never used Fable. Also note that I use AI as a tool to help with my work and I do not offload everything I have to do to AI in a single prompt
- How's it "quite short sighted"? Are you saying the math _does_ make sense? if so, how?by s0ss
- Why is on-device AI the future? What is your reasoning behind this? Look at the proportion of things we compute on someone else’s computer relative to what we compute on our own device. Why would this change for LLMs?by d4ng
- The money being poured into AI infrastructure means there is a market for new ways of doing things that take 1/1000 of the power or are 1000x faster, or both.
And there are many such moonshot startups.
AI on GPUs is an efficient as gaming on CPUs.
All that math where perfect precision is not required means that you can’t tell do things in different ways.
by sroussey - > As an example I use Claude code heavily for every day development and Opus 4.8 was already good enough for my use cases and never used Fable.
I think the flaw in this logic is thinking about how AI is currently used only. Yes, Opus is good enough for the task you are asking it to do, but that doesn't mean that is all you will ever need.
As AI gets better and better, it will open up new use cases that require the better performance.
by cortesoft - > [local, but for] extremely niche or higher intelligence tasks
Access to a centralized copy of the web/literature/media, scraped and indexed/integrated, seems another non-local center of gravity. Versus a local model's last minute reaching out to "manually" search and browse.
Also mass parallelism for large ensembles. Perhaps unless/until we get those local models as 10k+ tok/s chips. Local can follow frontier because frontier is still sort of "expensive rack local". If STOA becomes massive burst-parallel ensembles, that following may get harder.
by mncharity - The validity and permanence of a technology has literally nothing to do with how irresponsible people have been while placing speculative bets on it.
In this case, it’s really irresponsible.
by DANmode - > but don’t think there’s gonna be a crash […]. It would more or less would be a correction to valuations. [...] The future of AI would be on-device models which are as powerful as current frontier models […]
That’s the crash… that’s pretty much exactly Ed Zitron’s thesis
by dgellow - > I do think the valuations are quite and they would need to meet the reality, but don’t think there’s gonna be a crash or we’d ever go back to pre-AI era. It would more or less would be a correction to valuations.
Something to consider: would your description also apply to the dot-com boom of the late 1990s? The internet was real, the ideas for internet business were real, and we were not going to the previous reality. But the valuations weren't quite right and a "correction" happened at some point.
When people talk about AI crash, that's what they mean. Not that AI is a hoax, but that the correction could be quite violent and have effects on the broader economy.
by skippyfish - > If Anthropic and OpenAI believed customers would actually pay the real cost of AI tokens, they wouldn't have to give away 20 to 40 times the amount of tokens to subscribers.
One logical gap in the SemiAnalysis 40x cost of tokens versus subscription: I don't know anybody actually maxing out their account limits.
Sure, if you're somehow always running stuff, you can max it out, but subscriptions like this allow people to max sometimes (or always), while others never come close to the max.
What is the average usage of subscribers? Only Anthropic and OpenAI know, as far as I can tell.
by epistasis - Just to counterpoint the adjacent replier. I have claude, chatgpt, and gemini subs and rarely even use even close to 50%. I code c++ everyday. But im not full in on agentic workflows which is probably why. I mostly do fairly targeted things to my various code bases.by namrog84
- Also enterprises have to pay per-token.by telotortium
- I max out Claude Code and Codex subs both every week without fail and I’m doing a lot of work but it’s not insane.by wilg
- I'm of the opinion that the market will need to expand, horizontally. That's what's happening, with low-cost LLMs. I already know lots of "regular folks" that are totally hooked on LLMs (mostly ChatGPT, at the "regular mensch" level). They usually use the free tier, but a lot of them are doing the $20/month Pro tier.
If anyone is in my age group, they remember when ATMs were free. It was the "heroin dealer" business model. It worked out well (for the banks). The Chinese did it with manufacturing.
Getting people hooked on subsidized junk, is an age-old (and highly effective) business model.
Think of all the shops that will soon be composed of people that simply can't even get out of bed, if their LLM is not available. If the LLM dealer starts raising the price, there's no choice. I suspect many of the valuations are taking this into account.
> "Junk is the ideal product... the ultimate merchandise. No sales talk necessary. The client will crawl through a sewer and beg to buy."
-William Burroughs
- I will never understand the decision-making process that led to "Let's build an awesome VR headset that can't do gaming".
I would love to replace my monitor with Apple Vision Pro for programming and productivity. I would gladly pay $1000 for that.
But at $4000 it really needs to put me in a Microsoft Flight Simulator cockpit.
by sssilver - Same decision making process behind making a VR headset out of heavy materials like glass and aluminumby whywhywhywhy
- > I will never understand the decision-making process that led to "Let's build an awesome VR headset that can't do gaming".
FOMO, someone else being successful in VR (and now in glasses) is a nightmare scenario for Apple because it's not their platform. They've got a lot to fear from someone else owning the platform, having written the playbook on how that platform would be controlled and exploited at everyone else's expense.
Imagine if Apple doesn't make their glasses: Meta keeps selling truckloads of them, as they get more powerful the smartphone becomes an optional accessory and then an unnecessary one, eventually they become an alternative to a smartphone and an app platform for third party devs.
It's the same story with VR, Apple does nothing and risks Meta or Steam building a viable platform where people want to use software instead of on their iOS devices, until the hardware can replace iOS devices entirely. (although VR is certainly more of a moonshot)
by benoau - Who said it can’t do gaming? All the regular graphics APIs are supported.
- There is very silly but very simple reason for that. If they let you run MSFS on the Vision, there will be a non-insignificant amount of tiktok videos showing uninitiated people puking after a half an hour in it. There is no breakthrough in Vision that will resolve the nausea issue present in every VR headset, so they're keeping it AR first as the only way to protect against that.by maxgashkov
- For me, the economics are the exact opposite. I'd pay $1000 for a good gaming VR headset, but not $4000 -- because gaming is just a hobby that isn't really worth that much to me. Gaming VR headsets have been around for well over a decade, I've "been there, done that", and it's just not something that's worth that much money to me. However, I have no problem paying $3500+ for something that creates a new category of productivity and entertainment experiences for me.
Besides that, I actually disagree that the Vision Pro is not good for gaming. It's not good for traditional VR gaming, but using apps like Portal it's phenomenal for, e.g., playing existing PS5 games on a massive virtual screen, which in many ways I actually enjoy more than VR gaming, which is far too limited in comparison.
by nilkn - It's the same decision-making process that tried to do a electric car for years, and canceled it after spending time and money. In that exact timeline, China made great electic cars and new brands, Tesla advanced and Spacex launched dozens or more rockets. Apple is so far from any big new tech, they and market don't even realize it.
Apple had more money and influence than combination of all of them, yet failed with car and VR. Why would anyone think they can't fail much much bigger with AI and LLMs ?
by kingleopold - AI companies are a lot more like traditional manufacturing companies.
They can only subsidize you so far, because they may not be even be covering their variable cost at this point.
There's no software multiplier (build once — pay the cost once — sell many times).
Traditional SaaS is in a middle ground, there are operational costs associated with providing services, but per request they're usually negligible.
AI? I don't know, but it's not looking great from where I sit, unless there's a significant breakthrough in inference efficiency.
by juancn