4 ms·
$250k a year, for now. What's to stop anthropic for doubling the price if your entire business depends on it? What are you gonna do, close shops?
by lbreakjai 9mo ago
$250k a year, for now. What's to stop anthropic for doubling the price if your entire business depends on it? What are you gonna do, close shops?
- drited 9mo agoBy then perhaps it will be possible to continue with local LLMs
- teaearlgraycold 9mo agoWhat’s to stop them? Competition.
- blibble 9mo agothat worked real well for cloud computing aws and gcp's margins are legendarily poor oh, wait
- riku_iki 9mo agogcp was net negative until last year. Big part of why clouds are expensive is not necessary hardware, but all software infra and complexity of all services.
- oblio 9mo agoAll the big clouds are still in market share acquisition mode. Give it about 5 more years, when they're all in market consolidation and extraction mode.
- direwolf20 9mo agoMaybe not worth using then. Your product costs 5x and delivers 0.2x of competing product in the adjacent product class (traditional server/VPS), why use it?
- riku_iki 9mo agothose who don't need cloud services are feel free to use other options.
- direwolf20 9mo agoWhich cloud services do you need?
- riku_iki 9mo agoDB with cross zone replications and HA for starter.
- direwolf20 9mo ago
- lbreakjai 9mo agoFrom whom? OpenAI and Google? Who else has the sort of resources to train and run SOTA models at scale? You just reduced the supply of engineers from millions to just three. If you think it was expensive before ...
- simonw 9mo ago> Who else has the sort of resources to train and run SOTA models at scale? Google, OpenAI, Anthropic, Meta, Amazon, Reka AI, Alibaba (Qwen), 01 AI, Cohere, DeepSeek, Nvidia, Mistral, NexusFlow, Z.ai (GLM), xAI, Ai2, Princeton, Tencent, MiniMax, Moonshot (Kimi) and I've certainly missed some. All of those organizations have trained what I'd class as a GPT-4+ level model.
- lbreakjai 9mo agoAh but I said "_... and running at scale_"
- simonw 9mo agoOf the list I gave you, at a guess: Google, OpenAI, Anthropic, Meta, Amazon, Alibaba (Qwen), Nvidia, Mistral, xAI - and likely more of the Chinese labs but I don't know much about their size.
- lbreakjai 9mo agoI guess where I was leading to is who owns the compute that runs those models. Mistral, for example, lists Microsoft and Google as subprocessors (1). Anthropic is (was?) running on GCP and AWS. So, we have multiple providers, but for how long? They're all competing for the same hardware and the same energy, and it will naturally converge into an oligopoly. So, if competition doesn't set the floor, what does? Local models? If you're not running the best model as fast as you can, then you'll be outpaced by someone that does. 1. https://trust.mistral.ai/subprocessors https://trust.mistral.ai/subprocessors
- direwolf20 9mo agoHave they stopped making a loss yet? They'll all need to raise prices or they'll all go out of business, and now it's a game of chicken.
- coldtea 9mo agoAnd how has that worked out for us in any other software category?
- teaearlgraycold 9mo agoI mean it's kind of hard to say because almost all software I use is free, a lot of it is FOSS. The software I bought outright in the last couple of years was well priced because of competition (ex: Affinity Designer 2 for $63 - the new version is free although I stick with v2).
- elzbardico 9mo agoCompetition doesn’t magically waive costs, the investors expectations of return, neither debt serving obligations.
- ikr678 9mo agoYeah this is just trading largely known & controllable labour management risks for some fun new unknown software ones. You can negotiate with your human engineers for comp, you may not be able to negotaiate with as much power against Anthropic etc (or stop them if they start to change their services for the worse).
- toddmorey 9mo agoI mean… What does your shop even do? Write software? Why? The whole premise is that it’s now easily cloned.
- andersmurphy 9mo agoIf this is successful supply shock will kick in (because of energy/GPU constraints) and we could easily see a 2-4x price increase maybe more if the market will accept it. That's before taking into account current VC subsidies.