Opinion: Uber was supposed to help traffic. It didn’t. Robotaxis will be even worse

There is an argument, and it's pretty straightforward and reasonable. In congested cities, a large percent of drivers (I've seen as high as 1/3 for downtown Manhattan pre-pandemic) are within a few blocks of destination and are actively seeking parking, and Uber doesn't need to park. Which does imply an actually-correct argument -- the amount of parking you need in a city to support cars being the primary transportation mode is ridiculous.

But the entire sales pitch of Uber is that it will be so quick and easy that you'll use it without a thought. Their entire business model is to generate more trips, not consolidate or reduce trips. There is no incentive in them to reduce how long their users spend on the road absent competition with bikeped/public transit. There's a reason Uber pools never caught on. Obviously the easier car trips get, the more people will make trips by car. And while the parking geometry is insane, the road geometry of non-parking vehicles being the primary transportation mode is barely any better. Especially when you consider you've simply shifted a large number of vehicles-seeking-parking into vehicles-seeking-fare.

Absent any investment in better bikeped and public transportation, it is possible Uber can have a marginal improvement to congestion. But not enough to really make a difference and improve anyones' lives. And it's also possible it slightly worsens it by generating trips that otherwise would've simply not happened because the traffic sucked too bad.