Comment by cmiles8
20 hours ago
There are a lot of parallels to the WeWork S1. They’ve made absolutely massive forward cash commitments, on a business that’s increasingly commoditized, and with little hard evidence that they’ll generate the revenue (let alone the cashflow) to pay for all these future liabilities. All while everyone is discovering there isn’t really a secret sauce moat to demand much of a a premium over cheaper options that trade a few points above the direct COGS (cost of compute in this case).
If everything goes just perfect they might eek by. If it doesn’t this goes from high flying startup to insolvency and bankruptcy real fast.
> increasingly commoditized
Is it increasingly though? As of 5.5 I stopped using other models all together (and I used to regularly swap, indifferently). My $20 sub no longer runs out, even though I do _much_ more than ever. I use it for every single task at work, even trivial ones, and my spend has gone down. I was given _free_ Codex and realized... it still wasn't worth it. My rational brain says Anthropic has no moat and this is temporary. The progress of all the other models is amazing. I certainly wouldn't bet on them.
But if I look at the actions of myself and the most competent coders around me, they are all doing this same thing - using 5.5 and not really looking around any longer. And we know its not even their best model. OpenAI may respond, Chinese labs may catch up. But for the moment, Anthropic seem to have added separation, not lost it, based on current vs prior _behavior_ of coders I know at home and work.
A couple of points.
First, it does not really reflect my experience. I find modern models to be largely interchangeable when it comes to common software engineering tasks, including some of the more recent open weight ones. The strongest models become interesting for Ph.D. level of work for me, which is not something I deal with regularly anymore.
None of them are perfect; they all have their strengths and shortcomings, but I can reasonably work with all of them and it honestly doesn't matter much anymore. I do sometimes wonder what exactly people are doing to see such a difference; at a minimum, their workflows must be very different from mine.
Second, you're arguing based on subscription pricing. But Anthropic makes the bulk of its revenue (and has its biggest margins) from enterprise/API pricing, where they are the most expensive provider by far. Anthropic isn't going to make a lot of money off your $20 subscription. With a subscription, you currently get more usage per dollar out of Anthropic's than OpenAI's plans, which approximately offsets the difference in API pricing.
> I find modern models to be largely interchangeable when it comes to common software engineering tasks
If you think that, it means you're not demanding enough from your models.
Opus 5.5 still makes a lot of mistakes. Even if it can technically do all tasks you throw at it, there's no way it can do it at 100% reliability.
To be clear, AI writes 100% of my code since last year. But it still produces bugs, needs multiple prompts, and has less intuition than a human.
15 replies →
To clarify, i use both subscription pricing (personal use) and api pricing (at work). My spend going down was referring to api pricing, while using it more (for "everything").
Personal efficiency is noted as never hitting my limit despite it doing huge amounts of work at a time (ex: rewrite my notes app in swift for ios and install on my phone, 2 hours later its opened on my phone, looks nice, is working (base set of features). No follow up prompts, around 2 hours of work. 45% of my 5-hour limit.
Work efficiency is spend, task complexity, code quality (how often i have to steer - rare now), and wiring it up to more stuff (ex having it setup metrics and dashboards etc etc).
I used to sing the same tune as you are now- "models are interchangeable". For today, all other models seems significantly dumber and waste more of my time than Opus 5.5.
tl;dr: You should try Opus 5.5 and report back.
2 replies →
The history of tech is one where “good enough” beats “best in class” nearly every time. Anthropic’s whole business plan is that for reasons they can’t explain this time “best” is what matters.
All indications are the models are increasingly worth less and less and rapidly approaching zero. Even worse, everyone is rapidly shifting to “router” architecture where they switch models constantly. LLMs are perhaps the least sticky product in the history of tech.
The only money to be made is selling compute to run models on. The only real out I see for Anthropic is if somehow compute becomes so constrained that they can flip to sell all their committed compute at a markup to run commodity models. Absent something like that it would appear they don’t have a viable business plan. As many have been pointing out, they casually left Q3 off their charts. Put that back in and suddenly the charts look very worrisome.
You're admitting the problem here, anthropic needs revenue to keep growing and growing, yet you're already settled into a willingness to pay of only $20/month as probably a fairly ideal customer, bigger better models just aren't valuable to most people at the price they need to be to make the money back.
It is estimated that only 2% of the global workforce is paying for AI right now. If that grows to double digits - especially in the 30%+ range - and Anthropic captures a significant proportion of it, their valuation is justified. This is assuming they don’t offer virtual employees and these are 100% in the pipeline.
1 reply →
Sort of. If they can rely on differentiation and loyalty now, pre-IPO, from their large-but-cheap customer base, they can focus their growth on serving enterprise/other markets, and lean on their existing base as a low-churn-risk (compared to the rest of the AI ecosystem) customer and marketing surplus. Focus is good.
Anthropic does not make money if everyone is like you, spending $20/m on their SOTA model and are happy. That sounds very commoditized to me.
An example is I like Claude a lot, but for the same $20 spent with Google to get Argon 4, I also get ad-free on YouTube, a ton more storage, and it can access all my Google Drive files natively. I’m not going to pay for both. And I already stopped paying ChatGPT last Nov when Claude got good.
1 reply →
Ngl, I've been on Opus 5.5 but had to switch to open for a while. DSv4/Minimax 3.1 were pretty damn good/comparable. Definitely not the same but completely servicable.
If some later version of Deepseek had the capability as Opus 5.5, would you use Opus?
What many of us experienced this year is that Fable 5 and Opus 5.5 are good enough and future model improvements may have diminishing returns even if benchmarks improve.
I’m very excited for smarter models but what I really want now is not capability but speed, cost and open weights.
There will be a version of Deepseek with the capabilities of opus 5.5, but by that time there will be Opus 6 and it will likely cost the same or cheaper than Opus 5.5 and have nearly unlimited usage for subscribers.
Every model release people say the model is good enough, yet Opus and Sol are still more popular than Deepseek. The frontier is shortening, but there are still considerable room for models to improve. SlopCodeBench is new and not close to being saturated.
this conversation happens over and over every 6 weeks. It is absolutely a commoditized product. The models are getting better but they're _all_ getting better. GPT beats Claude beats Gemini. Then open models distill them at cheaper prices, etc.
This is great for us, but that behavior is the behavior of a commoditized product.
Right, the thing is, for most tasks, there’s a finish line. Once a model can do a given task right, future improvements may be of little to no benefit beyond cost reduction.
Right now OpenAI and Claude (and maybe Gemini) are over the finish line with reasonable prices for many tasks whereas the rest are still in the race.
But what happens when the open weight models cross the line and it just comes down to price?
I do a similar thing with 6.1 Sol. I find it pretty comparable to Opus 5.5 in many ways.
> My $20 sub no longer runs out, even though I do _much_ more than ever.
Somewhere in there is a definition of commoditization.
Eh, we have cursor at work with most of the top models and I switch regularly. I went from Claude to sol 5.6 and preferred its prose. Then my work took away the frontier models because they want to see if we can get by on cheaper models. I using Gemini Astra now because it’s discounted and it’s 90% of sol level and completely functional for planning and coding. Some coworkers are using muse coding, some Kimi and glm. Everyone is still producing code. I don’t think it’s a real moat as in you don’t have other options or switching models is actually a difficult process
> I using Gemini Astra now
Probably meant Gemini Argon
At $20/month you're the reason they're 80B in the red lol
"Yahoo! is the best search engine, and I cant imagine anyone making anything better."
Wouldn't a better analogy be Google, which would actually contradict the point you are trying to make?
3 replies →
that's a real stretch. WeWork was in a business that had already been commoditized (e.g. Regus) for decades.
Anthropic is in a business that didn't exist 3 years ago.
Yes it is a newer business but doesn’t change the fact that models are trending towards a commodity. And that’s a huge problem for the business model on offer.
I don't think the models being a commodity matters much. Consider: Email providers are a commodity. Cloud linux server hosting is a commodity. Checking accounts are a commodity. There's a lot of great businesses, trillion dollar businesses, built on commodities. DeepSeek v7 Pro might open source and be better than any Claude model (unlikely), but JP Morgan and Eli Lilly are not going to go spend $500M on chinese AI tokens routed through OpenCode. They just aren't. The models really matter far, far less than the experience these companies build around the model, and both Anthropic and OpenAI are doing a decent job of making that experience reasonably sticky and compelling.
3 replies →
Right – they can't argue for acceleration factored in their own value, if it's a property of the whole surrounding market.
IMO, US AI providers are not leaders in efficiency anyway, they're still running the original ~2010 playbook that led to LLMs ("What if we took ML researchers' work, but funded it at 50M/yr instead of 250k/yr?").
They're currently arguing for the continuation of that approach: Scale funding first, and the rest will follow. It is high-risk and WeWork-like, by nature.
1 reply →
Selling subscription productivity tools is a business that has existed for decades if not centuries. Just because their product is new does not mean their business is.
Oh come on. That's like saying "This isn't novel. People have exchanged money for goods and services for thousands of years"
Its already commoditized. Thats why things like OpenRouter are useful and successful because it lets you route inference at the click of a button between different providers after balancing cost and availability. Exactly how commodities behave.
> Anthropic is in a business that didn't exist 3 years ago.
but it's already commoditized.
And is already commoditizing. If not already commoditized
when they start removing biscuits from the Anthropic offices, that’ll confirm their WeWork-like death spiral
Someone should do the same thing Karen Hao did for OpenAI with the furniture costs of their office, if Anthropic's HQ decoration and office equipment value stops growing then we start worrying
This is hilarious lol
> There are a lot of parallels to the WeWork S1
Time has tempered it, then–re-read it [1]. It's a kaleidoscope of stupidity with no coherent economic case. The Anthropic S-1 is ridiculously ambitious. But it puts forward an argument for its valuation that is at least falsifiable.
[1] https://www.sec.gov/Archives/edgar/data/1533523/000119312519...
This is an excellent way of looking at it: buying something for the long term (computing capacity) and selling it in short-term, smaller pieces (tokens). The risk is that the obligation is long term while the income is short term.
It's like signing a 10-year lease on an apartment in NYC and renting it out by the day on Airbnb. It may or may not work.
On top of that, there's the question of how you treat the remodeling of that apartment (model training). If it benefits you for years to come, it's an investment that you depreciate. If your guests frequently trash the apartment, or tastes change (the model becomes obsolete quickly), that remodeling (training) is just a cost.
Yes, and 5 years and 1 Chapter 11 bankruptcy later, WeWork is now (EBITDA) profitable.
And they've gone private now.
Nobody was saying subleasing offices wasn’t a business. It was just WeWork’s setup that was untenable and ultimately insolvent.
Similarly nobody is saying AI can’t be a successful business, but here too the big labs have dug themselves into a giant hole such that a bankruptcy reset may ultimately be the only way out of it. Time will tell.
was it though? sounds like wework had the last laugh. is bankruptcy part of the strategy? people like you call insolvency the end of the journey. the journey is whatever you want it to be.
5 replies →
EBITDA profitable is not a substantial vote of confidence.
and are much much smaller than they were at IPO
My last shred of skepticism has been killed watching people try to imply this doesn't prove Anthropic (and likely OpenAI) were actually much better off than most people assumed and will certainly stay multi-trillion dollar companies
> little hard evidence that they’ll generate the revenue (let alone the cashflow) to pay for all these future liabilities
According to the S1:
- FY24 revenue $386 million -> FY25 revenue $4.59 billion (+1,008%)
- Q126 revenue $4.7 billion -> Q26 revenue $11.5 billion (+145%)
That seems like pretty good evidence of current and ongoing explosive growth. Much of this revenue is pre-paid (re: cashflow). Meanwhile, operating margins went from -184% to -$4%.
Seems like pretty good evidence. Will growth continue apace? Unknown. But the arguments for Jevon's paradox applying to intelligence seem pretty convincing, so the TAM here seems likely to be enormous.
> All while everyone is discovering there isn’t really a secret sauce moat to demand
Also not convinced by this. The Chinese models seem to have distilled their way to current levels, with Anthropic and OpenAI now getting serious about defending against this.
I don't think you're definitely wrong in all this. Rather, I think your claims are too confident.
> Q126 revenue $4.7 billion -> Q26 revenue $11.5 billion (+145%)
I suspect this started stagnating in Q3/Q4, which is why they're looking to IPO now.
I think this is unfairly underplaying WeWorks contribution to financial innovation with their creation of “community adjusted EBITDA”
Well every startup plays the game of “we’re profitable if you ignore all the reasons why we’re not profitable.”
GAAP isn’t perfect, but it exists for a reason.
except, I didn't need WeWork to get work done, I use claude like crack to get work done.
So you'll be shorting the stock, I take it?
Shorting a company is not a trade against the future of a company, it is a trade against the market. You can know a company is going to zero and the company can go to zero and you can still lose your shirt from shorting it.
"The market can stay irrational longer than you can stay solvent."
This quote stops me from shorting pretty much anything.
A lot of people will, if this IPO actually happens. But as is frequently pointed out a short is a market bet not a bet against the company and the market remains wild and irrational. You can be right about the company ultimately failing and still lose your shirt on a short.
However the price of CDS on bonds is already a strong market indication of where the market thinks all of this is heading. And none of that looks good if you’re betting on the AI exuberance continuing.
The bond market and the equity market care about different things. E.g. if the company has a 10% chance of going bankrupt and a 90% chance of doubling its market value, then paradoxically its bonds will look bad and its equity will look good.
When you've got companies like Tesla valued at what they're valued at. Shorting is pretty dumb.
Shorting is not the only way to protect yourself from a bad investment.
You can simply not buy shares in companies you think are on a doomed trajectory.
That said, depending on the IPO target price, I might short it.
You say this as if retirement accounts, and S&P500 ETF’s don’t exist and aren’t very common. Many people were forced into some amount of investment in spacexsi
1 reply →
"token factory" is indeed a terrible business model - no one is arguing that.
IMO the whole thing is based on the fact of "we will reach AGI/ASI eventually, and when we do, we will reach escape velocity".
No one knows - you, me, or Dario if this will happen and when.