Top
Best
New

Posted by simonw 5/27/2026

I think Anthropic and OpenAI have found product-market fit(simonwillison.net)
1094 points | 1245 commentspage 8
try-working 5/28/2026|
Product market fit for the models, sure. But their cost structure is going to kill them as token costs drop this year, following DeepSeek and Xiaomi and soon more providers.
trash_cat 5/28/2026||
Discovering that Enterprise customers tolerate higher prices compared to retails consumers is discovering demand elasticities not PMF. Am I missig something?
CuriouslyC 5/27/2026||
Companies are kool-aid drinking now due to hype, but given how much they're spending, if they don't see REAL, BIG wins from it soon, they're going to scale it back quickly and switch to Chinese models. Claude isn't worth the API cost for a lot of development work, and once companies have had time to collect and crunch data they'll see this.
grttq 5/27/2026|
Swear people like you were hyping the frontier labs so hard not long ago.

Funny to see the change of tone - a lesson for people not to get too ahead of themselves.

CuriouslyC 5/27/2026||
There's no change in tone, I'm still very bullish on the tech, Claude in particular just isn't worth the API price, which I've always felt was too damn high. I have paid for Gemini 2.5 Pro, Deepseek 3.2/4 and GLM 5 tokens happily though.
grttq 5/27/2026||
Lmao you won’t admit it will you?

You financially benefit from stuff like agents. Of course you will be the last to admit publicly when things aren’t quite heading in the right direction. The gap between hype and reality is ever increasing.

CuriouslyC 5/28/2026||
You misread my comment because of your personal bias, and now you're acting like it's some sort of own? Come on bro.

The models are putting in work, that's indisputable, the thing I'm calling out is that Anthropic's price and hype are running ahead of the value it's delivering. Agents are a great boost at <=5% of SWE salary, a mediocre tradeoff/risk at 20%+, and insane at SV tokenmaxxing levels.

chipsrafferty 5/28/2026||
I think my company is going to cancel our subscription once it realizes we're spending more money in a week that a developer gets paid in a month.
Legend2440 5/27/2026||
>Somehow this fragment turned into headlines like Uber’s COO says it’s getting harder to justify the money spent on AI tokenmaxxing, because the market for stories about AI failures remains enormous.

I notice this all over the place. Many people hate AI and want it to fail, and they're willing to invent misinformation if it supports that idea.

hansmayer 5/27/2026|
Well, it is a big news when the COO of Uber says it no? Not quite some small consultancy shop here.
Legend2440 5/27/2026||
But the COO did not say that. The headline was deliberately misrepresenting what he said.
uncivilized 5/27/2026|||
The article was posted on HN and discussed a day or two ago.

https://news.ycombinator.com/item?id=48268871

hansmayer 5/27/2026|||
No, he said exactly that, if you remove the corporate sanitised language designed to not offend the Uber CTO.
simonw 5/27/2026||
I think you're putting way too much weight into what one person said in unprepared remarks at the 27 minute mark in a 32 minute podcast conversation.
philipwhiuk 5/28/2026|||
Unprepared remarks are the best source of info frankly.
hansmayer 5/27/2026|||
That "one" person is the COO of Uber. And the other one - the one based on whose statement about burning through yearly AI budget in the first few months - the whole discussion sprung up internally at Uber in the first place is the bloody CTO of that huge company. So yes, their words do have A TON OF WEIGHT. Thats why they are in such important positions, arent they? They're not quite the Derek from the pub, casually commenting on how Liverpool will fare this season.
simonw 5/27/2026||
I think the way people reacted to those statements was entirely out of proportion to what was said.

I repeat: a CTO saying that they spent their entire AI budget for 2026 when that budget was clearly set in 2025 before anyone knew what those November models + harnesses were capable of is entirely unsurprising. Any analysis that doesn't also point out the difference between 2025 and 2026 era coding agents is either ignorant or deliberately misleading.

hansmayer 5/27/2026||
Yes, but that's irrelevant, because the COO uses that to base his core argument - that all that jackshit 1800 code changes per week that the CTO boasts about, mean absolutely nothing in terms of value. It means they are spending a lot on it, to gain as he diplomatically said "perhaps 20% more" - and I wonder 20% of fucking what - it's a ride-sharing app, what could they be possibly building on top of it with all that token crap?
simonw 5/27/2026||
You have to try pretty hard to get to "all that jackshit 1800 code changes per week that the CTO boasts about, mean absolutely nothing in terms of value" from what he said on that podcast.

(We still don't even know what Uber's planned AI budget for 2026 was. They didn't reveal that when asked - in https://www.theinformation.com/newsletters/applied-ai/uber-c... it says "He wouldn’t disclose exact figures of the company’s software budget or what it spends on AI coding tools").

hansmayer 5/27/2026||
I don't have to try at all - I think anyone who spent as much as an internship, let alone years at a modern tech corp would have no trouble distilling the absolutely clear message - we are spending too much for too little value. And actually wtf am I explaining myself? Every major tech outlet interpreted it like that too. It's not that hard Simon.
simonw 5/27/2026||
I think that both you and the other major tech outlets interpreted that poorly.
hansmayer 5/27/2026||
Oh really? How about the king of non-tech-outlets and digestible, shallow bites for the not-reading-books-middle-class, the Business Insider?

https://www.businessinsider.com/uber-coo-andrew-macdonald-ai...

You know their business is literally correct interpretation of the C-Suite statements.

simonw 5/27/2026||
I think Business Insider are not a particularly high quality publication - they're prone to clickbait - and that headline is an example of why I think that.

Hah, I just checked their homepage and here they go again amplifying that COO fragment from that podcast:

https://www.businessinsider.com/tokenmaxxing-debate-uber-exe...

> "That link is not there yet, right?" Macdonald said in comments that went viral, racking up over 2 million views on X. "I think maybe implicitly there is more that is getting shipped, but it's very hard to draw a line between one of those stats and, 'OK, now we're actually producing 25% more useful consumer features.'"

Yeah, something going "viral on X" is clearly a sign that it's quality information!

hansmayer 5/27/2026||
So no one got it right? Myself, tech outlets, non-tech outlets, everyone on twitter, etc?
simonw 5/27/2026||
That's why I wrote a whole section about it. I was deeply disappointed at how little thought people had appeared to put into this before amplifying the misleading headlines.

For someone who cares about media hype - https://hn.algolia.com/?query=author%3Ahansmayer%20hype&type... - you don't seem to be very discerning with regards to this particular story.

sandeepkd 5/27/2026||
If we take out the circular interests and investments here then there is no way that this is a feasible business in current state.
airstrike 5/27/2026||
Who's to say those enterprises won't churn after XYZ comes out with a decent enough model that costs 10x less to use?

There's a whole bag of clever tricks you can play to juice short term results leading to an IPO that may not work longer term.

I'll believe they've found product-market fit when they have a product. Right now they're selling the infrastructure, in a highly subsidized and undifferentiated way (at least over a sufficient long period of time of, say, a couple of years).

rhubarbtree 5/28/2026||
PMF is such a bullshit term. There’s no precise definition. It’s bs.

The fact is that investment is at a scale so large that current trajectories are nowhere near going to provide ROI.

The model companies are trying to milk every drop of ARR they can before IPOing. That is entirely the current narrative. You can hurt ARR before it churns if your goals are short term.

Anthropic being profitable is laughable. Sure, by some accounting measure that no one seriously uses. But looking at revenue and what they’re raising you can see the true story.

The truth is we need a revolutionary step up in capability to justify capex spend. It’s possible that might come - Opus 4.5 brought one - but failing that we’ll see the bubble pop once the IPO pumping is done.

cedws 5/28/2026||
My employer went from the Max plans to Enterprise this month which was utterly stupid. We went from paying $200 per head to like $500 for some people, even more for others. For the same product. Oh well, guess we’re doing our part to prop up this bubble.
MaxPock 5/28/2026|
The writer is on to something. Musk saw the writing on the wall and bought cursor. AI money is in coding plans
More comments...