Garbage. You can't include training by the companies that develop an llm in the comparison against companies that merely use the same llm. Apples and potatoes.
It'll get paid from revenue, not by redirecting employee salaries. All that AI+compute is literally what customers pay Anthropic for.
Big AI labs are not software companies where payroll dominates expenses. They're capex-heavy industrial entities; it just so happens that the "machines" (whose output they sell) are nominally the same category as the devices that their knowledge worker employees use on their desks.
I wonder if they ever will be. If the chinese open source models are only 3-6 months behind every major frontier model release, I can't see the business model. GLM-5.2 is supposedly on par to Opus depending on the case. And everybody and their mother can run that model in their datacenter and charge Dollars for tokens.
There is distillation going on where chinese providers give the model lots of outputs. We don't live in a world where chinese providers are not doing this so we can't compare the advantage of this distillation, but there is some advantage to it otherwise they wouldn't do it.
If Anthropic can block distillations somehow (which are fair game imo given that Anthropic et al did the same with the written works of mankind), then they might stop or slow down the chinese from catching up.
Chinese also have like 40% of the AI researchers of the world, plus they have access to a lot of cheap labour for writing training data. I'm sure an hour of training data creation from one of China's 162 million university educated people is much cheaper than an hour of work from one of US's 97 million. Probably still cheaper than someone from the grand area.
China is behind in AI chips/GPUs but they are catching up. One thing where they have a hard dependence on outside is their energy imports: they have to import a lot of stuff from third party countries. The US on the other hand is energy self sufficient.
Stopping distilation is a condradiction to growth: The more active users Anthropic gets, the easier it will be for chinese companies to distile the model. Heck, I can see a paid browser extension being issued that does nothing but send copies of your AI chatbox prompts+results to China - with a "hidden" feature that creates distillation prompts every now and then. Give each 5$ a month for installing the browser extension, and you got an unstoppable distillation botnet.
I think the panic around distillation misses the fact that US labs also benefit heavily from Chinese breakthroughs like Deepseek's work on sparsity, MoE and training architecture
It may be that US labs use Chinese models for distillation but we'd ofc never know because they can host the models themselves
xAI is discounting the first few months of their rental to Anthropic, which will help them seem profitable for a bit longer. We'll need to see if that lasts.
Compute (and IP theft) is the raw materials being used by the AI companies to build their products (the models).
It doesn't make sense to compare OpenAIs or Anthropics compute spend to that of our average software company, because different products require different raw materials. Dropbox also use way more storage than Snapchat, that's an equally silly comparison.
Luckily we're not, because then your tip to a waiter/waitress would be dependent on their student loans remaining, especially considering how many expensive liberal arts majors struggle to find a sufficient career.
Apples and potatoes are both something people will need to eat if we want to see it from the human utility perspective, and they both require some land space to be allocated for their culture (though one can of course conjugate both culture).
If you want to take the DDG LLM summary at fate value, apples are lower in calories and sugar but higher in fiber compared to potatoes, which are richer in vitamins and minerals like potassium and vitamin B6. Overall, apples provide more dietary fiber, while potatoes offer more protein and essential nutrients.
Comparison rarely lead to one obvious all superior option that discard every other considerations.
the saying "comparing x and y" implies that you compare something that one of them can't compete ; if people praise the softness of the skin first and foremost, comparing apples and potatoes won't lead interesting results
Yes, that’s certainly what people mean generally. Now if we consider perspectives like the one elaborated by Marcel Detienne in Comparer l'incomparable[1], we can go a bit further.
The comparison no longer starts with the goal to assess distinct objects in the frame of a given more or less established framework, and instead our attention is framed toward challenging ourself. That is, anchored toward finding what frameworks would allow to assess anything meaningful. And latter on, what does frameworks and framework creation reveals about ourself.
Excatly, further more it really depends on who is using it and how value is compared not $. Senior engineers growing without ai will probably deliver 10x in value then ppl who just found out about ai yesterday. Doesn’t matter how much you spent dolar wise
Maybe an ESL thing? "Potatoes" are literally called "earth apples" in some languages (e.g. pommes de terre in French; Erdäpfel in some German dialects.)
I came to appreciate typos, slightly ungrammatical sentences, and creative plays on common phrases. They are completely absent in LLM output without a specialized prompt (and they seem to struggle to make believable language mistakes even when prompted), so they can serve as a signal when judging whether the text I'm reading was written by a human.
As an aside, I observed my stance on proper English use changing in real time over the past 2-3 years. I used to be proud of writing clinically correct prose, and found mistakes in grammar and vocabulary grating. Now, I kind of welcome them and have stopped caring at all about committing such language crimes. I used to cringe when someone didn't capitalize the first words in their sentences - not anymore. I think we're years away from LLMs convincingly faking human-like mistakes (since all the work currently goes into avoiding them), so it's going to remain a useful signal for a while.
OpenAI and Anthropic aren't charities, so whatever cost they inccur for training will be passed down to the companies using the models. So you absolute should include it.
OpenAI and Anthropic aren't charities, so whatever cost they inccur for training will be passed down to the companies using the models
You should, but with two important caveats. First, you don't know what their amortization schedule is like so you don't know what the impact on the pricing will be (are they going to pass the cost on over 5 years or over 20 years?), and second they may go bust before paying the cost down so they may not get a chance to pass it all on. If someone buys the company then they'll get a discount on the value, which means the training costs are just eaten by the investors.
Anthropic spends [...] about $2m of compute per employee per year against a likely all-in comp of $500k+.
The rest of the software market trails. The top 1% of companies spend $89k per engineer per year on AI
This framing makes no sense. The reason Anthropic spends so much on compute per employee is that they are building models. Anthropic employees aren't opening Claude Code and spending $2m in inference every year, so comparing it to other software companies, where AI expense is mostly inference, is completely incoherent.
Yes, the cost has to be passed down eventually, but it's not passed down to one company; it's passed down to all of Anthropic's customers, so the actual share of that money will be distributed among Anthropic's clients.
Look, I 100% agree with the idea that OpenAI and Anthropic are both unsustainable companies that have dug themselves so far into a debt hole that, most likely, the only way they'll be rescued is with government intervention, but this is still a terrible article.
Because the money that OpenAI and Anthropic have to pay those costs with comes from the users who pay for the service. There is no "passing the cost" of anything. The consumer always pays all costs.
I don't know, compute is compute. Arguably making complex software with LLMs isn't all that different from training a model to do a thing. You're throwing a lot of compute at the problem and hoping for a stochastic solution. The distinction will become even blurrier with time.
Though I agree it might be informative to split it by industry sector.
Which is probably a backwards anti-pattern companies have built.
Your most expensive engineer's time is most valuable, so if you give them standard issue which is half the speed, you are throttling the value you can get from your engineer. Not to mention the mental drain of your cursor barely being able to move due to all the bloated virtual networking systemization.
It would seem to make sense to give more valuable employees faster equipment, so that their time isn't spent toiling with the slow machine, but rather actually producing value.
Eh, computers are so (comparatively) cheap, that you can give all your employees top-of-the-line equipment by that logic.
(Well, for this to make sense, you need good computers to be cheap, but the returns to diminish quickly after that. Which used to be the case until fairly recently: single core performance was pretty much capped, and after your whole compile is parallelised adding more cores to a machine didn't really help your best engineer either. Similarly for RAM.)
Meh - our dev machines are over-provisioned cloud pc's from MS, that run as fast a laptop from 2005. And I'm sure they are paying a premium for the privledge.
My 10 year old desktop sitting next to me is a lot faster, but alas, I can't BYOD.
> more expensive engineers don't use more expensive computers
They don't? If you give your best engineers substandard hardware to work on, you're going to get worse output from them compared to if you give them more expensive computers to work with.
Not completely true. Giving developers hardware that is too beefy is the main reason why so much software breaks down when run on users' machines, which are generally old, on spotty Internet connections, and RAM-starved. Devs just don't need to think about performance unless it's really asymptotically bad, while the users bear the full brunt of inefficiencies.
I agree, but good computers are so comparatively cheap and engineers so expensive that this logic applies to (nearly?) every one of your engineers, not just the most expensive ones.
... and that actually shows - senior engineers have spent actual paid time to train juniors. Plus they used to spent time contributing to open source projects or Stack Overflow, all the stuff which every company benefits from.
why stop there? Count how long and how much energy it took for evolution to produce that 3 chimp brain that is then educated, and add how long it took culture to produce the knowledge in text books for said education to be possible.