AI Is Getting Way Too Expensive
wheresyoured.at
AI Is Getting Way Too Expensive
1–10 of 17 posts
Re: AI Is Getting Way Too Expensive
#2Re: AI Is Getting Way Too Expensive
#3The reality is that there is more and more subscriptions available everywhere, and for $400 you can easily get $10K of tokens from Openrouter/official API pricing, you can pile up as many subs as you want and there is virtually no constrain once you make CC an API, you don't need Claude Code, you don't need Codex, you don't need Antigravity, you don't need Kimi Code.
We use now about 20 subscriptions (mixed, around $5K) and we average $60K-100K in token a month, THAT is the price, that's what's debited from our account. There is also many providers that resell tokens for cheaper, why wouldn't that be the price for us consumers?
If that changes in 6 months, so be it, but the current price of AI is that one.
PS: Many enterprises (maybe not large scale) have at least 1-2 subscriptions for each of their employees, so it's not really only startups & consumers.
Re: AI Is Getting Way Too Expensive
#4I disagree on the consumer side and I honestly can't really comprehend why people aren't talking about subscription prices like they ARE the prices, as consumers (and startups), the price we pay for IS that price, that it's sustainable or not on providers side, that's another matter and not our concern. The reality is that there is more and more subscriptions available everywhere, and for $400 you can easily get $10K…
So in a way your argument that users are able to spend 10k in tokens while paying a sub of 200 is actually agreeing with his point.
Re: AI Is Getting Way Too Expensive
#5I disagree on the consumer side and I honestly can't really comprehend why people aren't talking about subscription prices like they ARE the prices, as consumers (and startups), the price we pay for IS that price, that it's sustainable or not on providers side, that's another matter and not our concern. The reality is that there is more and more subscriptions available everywhere, and for $400 you can easily get $10K…
I think the article is more about the cost for AI companies like OAI and Anthropic to operate relative to what they’re making (not much). So in a way your argument that users are able to spend 10k in tokens while paying a sub of 200 is actually agreeing with his point.
They can also leverage their own tech to crush most businesses in an automated way if they wanted/needed to in the future. Let's be real, a good 90% of all existing techs company could be replaced and automated entirely with the right approach and unli-tokens.
Re: AI Is Getting Way Too Expensive
#6Re: AI Is Getting Way Too Expensive
#7Earlier quoted context omitted.
I think the article is more about the cost for AI companies like OAI and Anthropic to operate relative to what they’re making (not much). So in a way your argument that users are able to spend 10k in tokens while paying a sub of 200 is actually agreeing with his point.
Agreed, I went a bit away from the article perspective which is why I talk mostly about consumers, however on the provider side, I don't think they are in a great danger, this is at a level of economics that most people (myself included) can't really comprehend, I highly doubt leaders in Finance would invest such large sums if they didn't have a comprehensive forecast of all eventualities, let's also not forget that…
I would recommend to watch the documentary „Enron: The Smartest Guys in the Room“.
Re: AI Is Getting Way Too Expensive
#8I disagree on the consumer side and I honestly can't really comprehend why people aren't talking about subscription prices like they ARE the prices, as consumers (and startups), the price we pay for IS that price, that it's sustainable or not on providers side, that's another matter and not our concern. The reality is that there is more and more subscriptions available everywhere, and for $400 you can easily get $10K…
The amount of money these AI companies are burning is unprecedented and there is simply not enough money in the economy to keep it going at a loss for ten years. Google (Google!) went cash flow negative and is issuing bonds, basically asking for a loan. That's the reason markets are getting so nervous.
And yes, choice is good but how many of these new models are trained from scratch vs distilled from frontier models? If OpenAI and Anthropic go down in flames how much of the cheaper choices we see today are going to keep evolving? I am not trying to make a prediction either way, I am just pointing out the uncertainty. Which, again, is fine if AI is a commodity you can easily cut, but not great if your company is betting big on AI.