As other commenters have rightly pointed out - agents being a good product is orthogonal to agents being profitable. If you can start securing big contracts with enterprises by convincing them that the AI revolution is coming, that's enough for profit.
It will only be years down the road when people start realizing that they're spending millions on AI agents that are significantly less capable than human employees. But by that point OpenAI will already be filthy rich.
To put it a different way - AI is currently a gold rush. While Nvidia sells shovels, OpenAI sells automated prospecting machines. Both are good businesses to be in right now.
Absolutely. Although another sentiment also on HN is that it also doesn't matter which I find cynical when I see people use that as a pro-llm sentiment like "who cares if it doesn't work."
Granted, but you and I were both here in days when Terry Davis, may he rest in peace, was also. We have uncommon cause to know how broad a church "sentiment also on HN" may be.
Nvidia made $44B profit last year. They are functionally a monopoly and can charge whatever they want. That’s a good business.
OpenAI lost $5B last year. There are many competitors offering similar products, preventing OpenAI from extracting large profits. It isn’t a good business now. Sam Altman is promising it will be in the future.
It will only be years down the road when people start realizing that they're spending millions on AI agents that are significantly less capable than human employees. But by that point OpenAI will already be filthy rich.
To put it a different way - AI is currently a gold rush. While Nvidia sells shovels, OpenAI sells automated prospecting machines. Both are good businesses to be in right now.