This is the self-driving robotaxi fantasy. More profitable for businesses and cheaper for consumers. I don't know if Lyft and Uber are viable continuing business models at their current prices but I really doubt that robotaxis, even if they worked, would really be able to go much below those prices. I mean, their current business model is making labor provide its own capital and barely paying them enough. I'm not sure how removing labor but adding more expensive capital (with a lot of backend labor, also, too) really changes this much.
My general point is that Lyft and Uber have pretty good coverage and pretty good prices even in many suburban areas. You can already replace a lot of non-commuting trips with them at reasonable prices. One barrier to this is the marginal costs of driving are hidden so when you own a car a trip seems almost "free," but many households with one or more adults who don't need cars for commuting could probably subtract one and fill the gap with Lyft trips and save money.
If these services aren't replacing private cars, I'm not sure why robotaxis would. Just how cheap can a 4 mile trip to the grocery store be, and if it's $8 instead of $12 (numbers entirely made up) is that really going to make the difference to people? And I do think in urban hellholes like mine Lyft/Uber have made getting around without using a private vehicle more appealing, but I don't see how robotaxis improve that...