4 ms·
I am confident that Anthropic make revenue from that $20 than the electricity and server costs needed to serve that customer. Claude Code has rate limits for a
by simonw 9mo ago
I am confident that Anthropic make revenue from that $20 than the electricity and server costs needed to serve that customer.
Claude Code has rate limits for a reason: I expect they are carefully designed to ensure that the average user doesn't end up losing Anthropic money, and that even extreme heavy users don't cause big enough losses for it to be a problem.
Everything I've heard makes me believe the margins on inference are quite high. The AI labs lose money because of the R&D and training costs, not because they're giving electricity and server operational costs away for free.
- Esophagus4 9mo agoI always assumed that with inference being so cheap, my subscription fees were paying for training costs, not inference.
- smashed 9mo agoDoubtful
- simonw 9mo agoAnthropic and OpenAI are both well documented as losing billions of dollars a year because their revenue doesn't cover their R&D and training costs, but that doesn't mean their revenue doesn't cover their inference costs.
- overgard 9mo agoDoes it matter if they can't ever stop training though? Like, this argument usually seems to imply that training is a one-off, not an ongoing process. I could save a lot of money if I stopped eating, but it'd be a short lived experiment. I'll be convinced they're actually making money when they stop asking for $30 billion funding rounds. None of that money is free! Whoever is giving them that money wants a return on their investment, somehow.
- simonw 9mo agoIt matters because as long as they are selling inference for less than it costs to serve they have a potential path to profitability. Training costs are fixed at whatever billions of dollars per year. If inference is profitable they might conceivably make a profit if they can build a model that's good enough to sign up vast numbers of paying customers. If they lose even more money on each new customer they don't have any path to profitability at all.
- citrin_ru 9mo ago> If they lose even more money on each new customer they don't have any path to profitability at all. In theory they can increase prices once the customers will be hocked up. That's how many startups works.
- bildung 9mo agoBut only if you ignore all the other market participants, right? How can we ever reach a point where all the i.e. smaller Chinese competitors perpetually trailing behind SOTA with a ~9 month lag but at a tiny fraction of the cost stop existing? I mean we just have to look at old discussions about Uber for the exact same arguments. Uber, after all these years, still is at a negative 10 % lifetime ROI , and that company doesn't even have to meaningfully invest in hardware. IMO this will probably develop like the railroad boom in the first half of the 19th century: All the AI-only first movers like OpenAI and Anthropic will go bust, just like most railroad companies who laid the tracks, because they can't escape the training treadmill. But the tech itself will stay, and even become a meaningful productivity booster over the next decades.
- viking123 9mo agoI am also thinking long term where is the moat if it will inevitably lead to price competition? Like it's not a Microsoft product suite that your whole company is tied in multiple ways. LLMs can be quite easily swapped to another.
- overgard 9mo ago
- vrighter 9mo agoModels are fixed. They do not learn post training. Which means that training needs to be ongoing. So the revenue covers the inference? So what? All that means is that it doesn't cover your costs and you're operating at a loss. Because it doesn't cover the training that you can't stop doing either.
- simonw 9mo agoTraining costs are fixed. Inference costs are variable. The difference matters.
- vrighter 9mo agoNo they are not. They are exponentially increasing. Due to the exponential scaling needed for linear gain. Otherwise they'd fall behind their competition.
- simonw 9mo agoFixed cost here means that the training costs stay the same no matter how many customers you have - unlike serving costs which have to increase to serve more people.
- 3form 9mo agoYet somewhere above you said: >Training costs are fixed at whatever billions of dollars per year. Which I think is the part people disagree with.
- simonw 9mo agoI used the word "fixed" there to indicate that the cost of training is unaffected by how many users you have, unlike the cost of serving the model which increases as your usage increases.
- beAbU 9mo agoIs inference really that cheap? Why can't I do it at home with a reasonable amount of money?
- tverbeure 9mo agoNobody questions that Anthropic makes revenue from a $20 subscription. The opposite would be very strange.
- brandensilva 9mo agoYeah it's the caching that's doing the work for them though honestly. So many cached queries saving the GPUs from hard hits.
- xienze 9mo agoHow is caching implemented in this scenario? I find it unlikely that two developers are going to ask the same exact question, so at a minimum some work has to be done to figure out “someone’s asked this before, fetch the response out of the cache.” But then the problem is that most questions are peppered with specific context that has to be represented in the response, so there’s really no way to cache that.
- marcyb5st 9mo agoFrom my understanding (which is poor at best), the cache is about the separate parts of the input context. Once the LLM read a file the content of that file is cached (i.e. some representation that the LLM creates for that specific file, but I really have no idea how that works). So the next time you bring either directly or indirectly that file into the context the LLM doesn't have to do a full pass, but pull its understanding/representation from the cache and uses that to answer your question/perform the task.
- simonw 9mo agoA lot of people believe that Anthropic lose money selling tokens to customers because they are subsidizing it for growth.
- Drakim 9mo agoBut that has zero effect on revenue, it only affects profit.
- what 9mo ago>make revenue from that $20 than the electricity and server costs needed to serve that customer Seems like a pretty dumb take. It’s like saying it only takes $X in electricity and raw materials to produce a widget that I sell for $Y. Since $Y is bigger than $X, I’m making money! Just ignore that I have to pay people to work the lines. Ignore that I had to pay huge amounts to build the factory. Ignore every other cost. They can’t just fire everyone and stop training new models.
- simonw 9mo agoSee comment here: https://news.ycombinator.com/item?id=47057874 https://news.ycombinator.com/item?id=47057874
- ben_w 9mo agoAs I understand it: Gross profit = revenues - cost of goods sold Operating profit = Gross profit - operating expenses including depreciation & amortisation Net profit = Operating profit - net interest expense - taxes