Personally, I don't even think the two are in conflict. Many of the comments on this article vastly underestimate self-driving's impact. Self-driving creates motives to use transit because it allows the economics of automobile fleets to consolidate around shared use instead of ownership - the pull of "the robotaxi service is cheap", which can be pulled off with a regular, modest YoY cost reduction, will ultimately make people give up the car, and if they give up the car, they will then save a few more pennies by using the bus on occasion - the bus will be more frequent if it goes self-driving too, because it's the labor cost of drivers that determines a lot of the operational economics of transit fleets. Take out that cost and you can run a lot of tiny vehicles as well as huge trains, meaning transit, deliveries and taxis will blur together as modalities, no longer needing to provision for what pencils out with a human driver or to adopt an ownership model where the user has to buy for their largest use-case and eat the associated cost. They can go smaller without much issue and rent larger as needed. It's a different vocabulary of "what transit is" once you add that dynamic.
The political element follows directly from that: if you aren't a car owner, but you are a bus and taxi user, you stop caring about parking and the specifics of your commute route. The robocar service will lobby the city for their own efficiency in your place - and what benefits their efficiency also benefits the users. So in one stroke you end up with an urban population that is aligned to empty out huge swathes of road space and garages for other uses such as bus lanes, bike parking, etc. People will feel safer on a bike because the cars will drive safely, and this is found in studies to be the main bottleneck to bicycle usage, so investment in dedicated bike infra could flatline in a self-driving environment, and bike usage will still rise!
It's hard to find an angle where there is truly a downside - the doomer response is kneejerk.