It's pretty obvious from OpenAI's corporate structure that they always knew they were a moonshot. If their products reached the point where they were reliable enough to fully offload human labor onto, the increase in productivity would be extremely valuable. At the level of reliability where a human still needs to check the machine's work, the utility is extremely limited.

And I don't think there's anything wrong with that. Compared to most things that we burn respurces on, at least the AI investment has produced something that doesn't unambiguously make the world a worse place.