Re:I think Anthropic and OpenAI have found product-market fit
by antonvs on Thursday May 28, 2026 @11:03AM
(#48309968)
> ... against the actual work their company cares about doing. [...] stuff that matters
This is a key point. Some engineers are having fun doing e.g. greenfield stuff with AI that they never would have had time for otherwise. Whether the company cares about that is another question.
It's related to Goodhart’s Law. If AI token usage is a target, then you're going to get a lot of token usage, but it's not likely to correlate well to improved business outcomes.
This is a key point. Some engineers are having fun doing e.g. greenfield stuff with AI that they never would have had time for otherwise. Whether the company cares about that is another question.
It's related to Goodhart’s Law. If AI token usage is a target, then you're going to get a lot of token usage, but it's not likely to correlate well to improved business outcomes.