One does wonder if the self-driving promise here was:
1. A cynical con job to get investors on board from the beginning
2. Not quite so cynical but nonetheless recognized by management as a hail mary pass
3. Uber just got swept up in the overall industry overoptimism/groupthink around self-driving dates and capabilities
but there was basically zero chance that self-driving cars of the sophistication and scale needed by uber would be widely deployed within 10 years (i'd wildly guess a 50% chance of its existence in 30 years), as anyone employing a critical mind for more than a few minutes could surmise.
However, you've only had to read this site over the past few years to see the huge number of people who bought the Level 5 is right around the corner story hook line and sinker. In all fairness, I attribute a lot of this to people who so wanted to believe they'd never need to own or drive a car that they weren't as critical as they should have been.
I find it "interesting" how quickly the narrative has shifted. A lot of the same people who were more or less making the aggressive promises pretty much all changed their stories in near-unison. One suspects they all knew or had come to known at some level but felt they had to wait for others to back off too lest they be seen as laggards.
once uber/lyft cut out the dispatcher, reduce no-shows a tiny bit, increase utilization a bit, off-loaded cap ex/risk to drivers, where do the killer profits come from?
when you do a five forces analysis, the only place they have some leverage is on the drivers (supplier-side), whom they've already squeezed down to below minimum wage (on average).
walmart can eek out huge profits from a low-margin business because they provide essentials that are hard to subsitute at their price point. ride-hailing is not nearly as essential and has a number of substitutes available in most places.