Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI's economics don't make sense

wheresyoured.at

Post reply on HN
Print view View original

151–160 of 201 posts

Re: AI's economics don't make sense

#151

Post by lbreakjai »

joshjob42 wrote: post #10

There's a few major problems with the article. The most obvious is that frontier labs are not charging remotely close to the cost of tokens; afaik most estimate north of 80% profit margins. As a reference, providers are profitably providing Kimi K2.6 for $4/1Mtok out. Is that as good as Opus? No, but it's probably at least Sonnet level, so that's ~4x cheaper than Sonnet while still being profitable to serve on the ma…

Problem with this math is it always assumes some ridiculous baseline compensation (or costs, in this case) as a matter of fact. There's an entire world of developers not costing 200k to their employers.

Truth of the matter in most companies large enough is if you make your devs 30% more productive, then that'd mean 30% more code going through "change management" hell for months. You're not even paying to stand still, you're just pushing even more down a bottleneck. The price most people are willing to pay to make things worse is close to zero.

Re: AI's economics don't make sense

#152

Post by margalabargala »

maplethorpe wrote: post #147
margalabargala wrote: post #118

Earlier quoted context omitted.

Extensively discussed elsewhere in this thread. Just start at the top and start reading comments.

Can you summarise? I only reached your comment after scrolling past all the others and I still don't have the answer. Is the new data that models are more useful for coding than they once were?

That sounds like a reading comprehension skill issue? In which case I don't see why me summarizing would move the needle.

But if it helps, no, the data being discussed is surrounding the economics of running inference and R&D, nothing to do with the utility of models for coding.

Re: AI's economics don't make sense

#153

Post by loeg »

kcartlidge wrote: post #136
loeg wrote: post #41

Earlier quoted context omitted.

> How many tokens can you realistically burn through in one chat session? I've used single digit billions in a couple days, FWIW.

I'm a fair bit lower than some others as I only use it outside of work hours on my own small projects, but my Cursor account shows (for a random recent date) 12,184,233 tokens in a day. That day feels pretty representative. That's with 86 interactions spread intermittently over a couple of hours so if I did a full working day like that I'd be looking at maybe 40 to 50 million.

My employer is paying for it, so I'm cost insensitive, and this is mostly with Claude / Opus 4.7 (which consumes a lot of tokens?).

Re: AI's economics don't make sense

#154

Post by alsetmusic »

JohnMakin wrote: post #53

I've sort of lost some respect for ed that I had early on in the hype cycle - he's still right about some things, but I can see him slowly and subtly retreating from his strong position, held even a few months ago, that these things will never ever be useful for anything and it's all a scam because they don't actually do anything at all except burn money. He would say it like 8 times a monologue. I remember one podca…

I am sympathetic to his view because I also considered the whole AI hype train a complete scam until pretty recently. When I saw enough people validating that coding agents were actually legitimately ok and sometimes good at things, I decided to spend $50 on one to test it out.

I have been pleasantly surprised at its utility knocking out grunt work. It's not super smart, but it's great at things like writing a python script to edit characteristics of a jsonl file or sorting structured data. I didn't ever expect it to be useful beyond extremely limited output and it's actually kinda good when you know how to narrowly target the tasks. The constraints of code make it a more suitable category than all the other stuff.

It's still a bs hype machine with Elon saying it might save all of humanity in court today. That's pretty unlikely.

Re: AI's economics don't make sense

#155

Post by senectus1 »

cottoneyejoe wrote: post #128
JohnMakin wrote: post #53

I've sort of lost some respect for ed that I had early on in the hype cycle - he's still right about some things, but I can see him slowly and subtly retreating from his strong position, held even a few months ago, that these things will never ever be useful for anything and it's all a scam because they don't actually do anything at all except burn money. He would say it like 8 times a monologue. I remember one podca…

His reasoning about costs are also completely flawed. API fees aren't the providers' costs. It's a largely arbitrary number that they think they can get away with based on what everyone else seems to be doing that they also expect to cover on-demand usage as well as their research, marketing, and stock buybacks. They likely have a 60-90% gross margin.

show me an AI company that is making an cashlow profit?

Re: AI's economics don't make sense

#156

Post by swader999 »

IshKebab wrote: post #145
swader999 wrote: post #46

Earlier quoted context omitted.

The output won't be read by humans (and increasingly this is the case in my own use) so I don't see how that works. If the output itself will be directed by the highest bidder, that doesn't work. Or if the output influences the agent's direction, that doesn't work either.

What do you mean it "doesn't work"? I can totally see OpenAI take money in return for companies adding custom content ("Everyone agrees Mattresses4u make the best mattresses") to the training data.

The utility of what your trying to accomplish goes to crap. For example, design me a strength program and it gets corrupted by gyms, trainers in my area etc that have been paid to be promoted in the output, especially if it's subtle. Or all of a sudden I'm getting a stack with Oracle in it all the time...

Re: AI's economics don't make sense

#157

Post by Rury »

hparadiz wrote: post #61
JohnMakin wrote: post #53

I've sort of lost some respect for ed that I had early on in the hype cycle - he's still right about some things, but I can see him slowly and subtly retreating from his strong position, held even a few months ago, that these things will never ever be useful for anything and it's all a scam because they don't actually do anything at all except burn money. He would say it like 8 times a monologue. I remember one podca…

The economics is spending a few hundred bucks on software for an IC you're already paying over ten grand a month in order to make them more productive. How are supposedly smart industry experts not seeing this obvious fact? Are these guys actually experts?

NVIDIA execs are now saying otherwise: https://fortune.com/2026/04/28/nvidia-executive-cost-of-ai-i...

Maybe Ed is right even if he's wrong on some things?

Re: AI's economics don't make sense

#158

Post by purplepatrick »

I keep seeing articles like this that extrapolate from token pricing onto token costs. This is wrong.

Companies don’t sell their goods/services at cost. A model’s being priced at, say, $30/M for output tokens doesn’t say anything about what it costs the company to provision the 1M tokens via the model.

And no, you cannot extrapolate from any company margins that someone may have overheard in an SF coffee shop onto individual product line margins or their trajectory either. This information is usually unknowable even in most SEC filings for public companies.

It’d be great if people who wrote these articles used, say, AI to look up some basics on how a business operates. It’s really easy to do, believe me ;)

Re: AI's economics don't make sense

#159

Post by no-name-here »

Glyptodon wrote: post #39

I think there's another route this goes. At $7k a year or more per eng in token use, I think it's very reasonable to buy engineers machines with obscene GPUs and RAM and run models locally. And if it doesn't make sense now, someone will figure it out and save companies $10k+/eng over 3 years.

If you only want/need the kind of model output that can be served on a machine costing single digit thousands, aren’t cheaper cloud-served models available? (And as the sister comment points out, sharing hardware allows greater utilization and lower costs per user.)

Re: AI's economics don't make sense

#160

Post by chromacity »

great_tankard wrote: post #81
JohnMakin wrote: post #53

I've sort of lost some respect for ed that I had early on in the hype cycle - he's still right about some things, but I can see him slowly and subtly retreating from his strong position, held even a few months ago, that these things will never ever be useful for anything and it's all a scam because they don't actually do anything at all except burn money. He would say it like 8 times a monologue. I remember one podca…

This is exactly how I feel about him too. I also find his "number big" approach to writing ("check out my 18,000 word blog about something I'm learning about in real time") off-putting, so I've completely stopped engaging with it. We need better critics of the industry.

> We need better critics of the industry.

There's plenty, but they don't have enough material to post once a week. And if you don't post once a week, you don't end up on HN once a month. As simple as that. Looking at the blogs that show up on HN regularly, the usual hit rate is 10-25%.

Post reply on HN