Comment by pier25
6 days ago
I'm mostly skeptical about AI capabilities but I also think it will never be a profitable business. Let's not forget AI companies need to recoup a trillion dollars (so far) just to break even [1].
VCs are already doubting if the billions invested into data centers are going to generate a profit [1 and 2].
AI companies will need to generate profits at some point. Would people still be optimistic about Claude etc if they had to pay say $500 per month to use it given its current capabilities? Probably not.
So far the only company generating real profits out of AI is Nvidia.
[1] https://www.goldmansachs.com/insights/articles/will-the-1-tr...
[2] https://www.nytimes.com/2025/06/02/business/ai-data-centers-...
Some of the revenues are very real. A few million subscriptions at tens-hundreds of dollar per month add up to non trivial revenue pretty quickly. I think most software engineers will have such subscriptions fairly soon. We're talking about a market where companies drop 5-20K/month on software engineers and hire whole teams of those. Of course they are going to spend on this. Maybe not 500$. But 20-100$ is a lot less controversial.
And this is quickly spreading beyond software engineering. Software engineers are just being guinea pigs for agentic AIs eventually popping up in all sectors. Basically, while security and quality issues are being sorted out, it helps having users that are a bit more clued in about what they are doing.
That's why AI investments are so hot right now. Of course there are a lot of AI companies that will fall short. There always are. And companies like Nvidia that will make a lot of money selling GPUs.
But there is some non trivial amount of revenue potential there. Anybody still in denial about that is probably penny pinching.
I refuse to pay for any product where I'm the product. Run it locally, maybe I'll pay for it but never online.
My guess would also be that at the 100$ price point only one company can be profitable but that is just a very wild guess.
For what I've seen SE's are some of the last to adopt it. My marketing colleague has been overflowing in generic AI crap produced by external writers for over a year now.
> I refuse to pay for any product where I'm the product.
Are you "the product" if you're paying $50 but the company also gets $0.35 of value from your data?
If yes I think you're overreacting, if no then I don't think your worries apply to AI subscriptions. (Other worries do, but not that one.)
1 reply →
Right now Google, OpenAI, Anthropic, etc are playing a game of chicken. But even after that it's not clear if the remaining players will be able to generate a profit.
Even if AI companies can recoup the billions or trillions of dollars invested, how long will that take and what will be their margin after that?
Hence the flurry of articles, controversial interviews and aggressive marketing and the pitch to replace all white color jobs.
What about the free open weights models then? And the open source tooling to go with them?
Sure, they are perhaps 6 months behind the closed-source models, and the hardware to run the biggest and best models isn't really consumer-grade yet (How many years could it be before regular people have GPUs with 200+ gigabytes vram? That's merely one order of magnitude away).
But they're already out there. They will only ever get better. And they will never disappear due to the company going out of business or investors raising prices.
I personally only care about the closed sourced proprietary models in so far as they let me get a glimpse of what I'll soon have access to freely and privately on my own machine. Even if all of them went out of business today, LLMs would still have a permanent effect on our future and how I'd be working.
How do you plan on running those open weight models then? You need a bunch of very expensive hardware to even slightly "good" performance.
Will you be buying/renting the hardware needed to run these models with a decent performance?
I can guarantee +99.999% of users won't be doing that.