Live data from Hacker News

Are the costs of AI agents also rising exponentially? (2025)

tobyord.com

81–90 of 155 posts

Re: Are the costs of AI agents also rising exponentially? (2025)

#81
post #67
post #64

AI feels more like a gamble. People like gambling. From casinos (win-loose), to lootboxes (uncertainty) or even extramarital sex (whose baby is it?). This way - AI work is like a slot machine - will this work or not? Either way - casino gets paid and casino always wins. Nevertheless - if the idea or product is very good (filling high market pain) and not that difficult to build - it can enable non-coders to "gamble"…

Under this definition, everything is gambling, including commenting on HN (will I get upvoted or downvoted?).

Everything is modeled after gambling nowadays to nudge and gamify the user experience to a desired outcome

Re: Are the costs of AI agents also rising exponentially? (2025)

#82

Once a model is stable and good enough, for example Sonnet 4.6 or GPT 5.4 (or something else in future), it can be burned into hardware like Talaas chip reducing the cost many times and increasing the speed. At some point we can rely on old model while being productive with it.

No, burning models into hardware won't make them faster or reduce the cost. It will cost way more for similar performance as what you would get with a gpu. I am not telling you why, you can go figure that out on your own.

But isn't this happening here https://taalas.com/ already. They have a demo of llama running at 17000 tokens per second https://chatjimmy.ai/

Re: Are the costs of AI agents also rising exponentially? (2025)

#83
post #59

I think an interesting counterpoint, is whether the value obtained is reducing.

Exactly. AI is not really replacing people but it's definitely allowing them to do more and more interesting things. You should offset the cost of having an AI do something against the cost of doing that manually. Your mileage may vary of course. But I am definitely getting things done that I wouldn't even have started without AI assistance. And that stuff is valuable to me. Although you could argue that anything AI can do is actually deflating in value as well. The economics here will get pretty interesting. But all things considered, I'm not spending an unreasonable amount on all this AI stuff. Probably around 60-100$/month currently. It varies a bit.

Re: Are the costs of AI agents also rising exponentially? (2025)

#84

> On many task lengths (including those near their plateau) they cost 10 to 100 times as much per hour. For instance, Grok 4 is at $0.40 per hour at its sweet spot, but $13 per hour at the start of its final plateau. GPT-5 is about $13 per hour for tasks that take about 45 minutes, but $120 per hour for tasks that take 2 hours. And o3 actually costs $350 per hour (more than the human price) to achieve tasks at its fu…

Where are you getting hourly costs for private models ? The rate limits are pretty arbitrary. If you max out by api tokens it would be like $10k / hour

Re: Are the costs of AI agents also rising exponentially? (2025)

#85
post #81
post #67

Earlier quoted context omitted.

Under this definition, everything is gambling, including commenting on HN (will I get upvoted or downvoted?).

Everything is modeled after gambling nowadays to nudge and gamify the user experience to a desired outcome

If your definition of "gambling" is "things with uncertain outcome", then there's nothing in life that's not gambling.

Re: Are the costs of AI agents also rising exponentially? (2025)

#86
post #67
post #64

AI feels more like a gamble. People like gambling. From casinos (win-loose), to lootboxes (uncertainty) or even extramarital sex (whose baby is it?). This way - AI work is like a slot machine - will this work or not? Either way - casino gets paid and casino always wins. Nevertheless - if the idea or product is very good (filling high market pain) and not that difficult to build - it can enable non-coders to "gamble"…

Under this definition, everything is gambling, including commenting on HN (will I get upvoted or downvoted?).

There used to be forums without voting. It was discovered that forums with voting attract more engagement because of the emotions produced by the voting.

Re: Are the costs of AI agents also rising exponentially? (2025)

#88

Earlier quoted context omitted.

No, burning models into hardware won't make them faster or reduce the cost. It will cost way more for similar performance as what you would get with a gpu. I am not telling you why, you can go figure that out on your own.

But isn't this happening here https://taalas.com/ already. They have a demo of llama running at 17000 tokens per second https://chatjimmy.ai/

With some research, that chip appears like it would cost about $300-$400 to manufacture, die only.

For an 8B parameter model.

Opus is estimated at 500B-2T parameters. At that scale you’re past reticle limits and need HBM and multi-die packaging, which means you’ve essentially built an inference ASIC (like Groq or Etched) rather than something categorically cheaper than GPUs. The “burned into silicon” advantage mostly evaporates at frontier scale.

Re: Are the costs of AI agents also rising exponentially? (2025)

#89
post #59

I think an interesting counterpoint, is whether the value obtained is reducing.

Exactly. AI is not really replacing people but it's definitely allowing them to do more and more interesting things. You should offset the cost of having an AI do something against the cost of doing that manually. Your mileage may vary of course. But I am definitely getting things done that I wouldn't even have started without AI assistance. And that stuff is valuable to me. Although you could argue that anything AI…

Value feels pretty relative to me. If anyone can do a 'thing', is that thing worth less?

Re: Are the costs of AI agents also rising exponentially? (2025)

#90
post #60

No, but the AI labs would love to frame it this way so they can continue to nerf models and increase prices while they use the cheap, highly performant, highly powerful models internally to replace all of your businesses.

Sure is looking that way. What can't Claude do at this point?

> What can't Claude do at this point?

Writing maintainable code that scales.

Post reply on HN