Will this ever happen for OpenAI, i have no idea, but that is the bet.

More than a bet. It's an engineered outcome, as Microsoft tries to force people into their AI walled garden.

Only question is how many people go along for the ride.

Uber was running a deficit for a long time until it turned profitable.

Uber is a great example of negative externalities, as the average Uber driver doesn't earn money once you depreciate the value of the car being used.

That said, if things improve both in terms of the quality of responses you can get from models as well as reduced costs to run them, then there is definitely huge economic potential.

The line I've seen on AI boils down to this. AI won't meet human economic potential. But it will run cheaper, which means paper growth, which means the investment is "worth it" at an industry level.

But at a macro level? Economy wide? Big Number may go up, but real productivity is going to slide the more AI attempts to replace human labor.