Self driving cars won't make financial sense in a lot of regions.

The driver plus bring your own vehicle model is a mostly operational cost model where the cost slides based on region, which makes it perfect for rolling out on a global scale.

The self driving taxi service model has a significant capital cost associated with it and that cost won't slide much based on region. It would take far too long to see meaningful returns on self driving taxis in India for example.

which is maybe why the losses are a little less dramatic than they sound. once they eliminate their network of independent contractors I'm sure the financials look a whole lot better. ultimately they are just another boring logistics company.