The economics really aren't there for chatbots. They don't scale well. Each additional user costs more money than they generate in revenue. But automated taxis? That's a different story. Currently Waymo has to go through the laborious process of creating intricate 3D maps of the cities and places its vehicles will operate in. But that mapping cost is fixed, regardless of the number of taxis driving in that city. And a lot of R&D has gone into everything. But the R&D costs are the same whether we're talking one taxi or a million.

Robotaxis have very different economics than AI chatbots. They actually have very real economies of scale. Plus while their self-driving vehicles currently cost about $150k, that cost is only going to go down with time. And compared to the cost of a human driver? It's hard to see how they won't end up profitable eventually.