What's the theory of change here? FSD fundamentally is going to make keeping a car moving on the road cheaper, and making something cheaper makes it happen more.
If you wanted something like dynamic routing, then what you'd do is have it as a supplement to standard routing. Effectively taxi to the nearest bus stop.
There's absolutely no reason for dynamic routing to not reliably be faster than or as fast as static routes.
You could even design vehicles for transporting multiple individual passengers in totally separate spaces.
There are endless improvements to the status quo you can achieve with self driving vehicles, things which will simultaneously improve quality of life for literally everybody.
When 100 cars hit a decrease from 4 lanes to 3 you are going to have a jam no matter how the cars are driving.
When we continue using individual transportation to allow suburbs to expand further and further from the places people want to go roads will continue to fill and traffic will slow and jam.
We can make the roads straighter with more gentle curves and robotaxis can drive 100mph you say!
Keep going…
Cars traveling at a safe speed do not need to slow down in a turn. If the friction on the 4 wheels doesn't result in oversteer nor understeer, you just keep going through. So what are you referring to? Laws of nature? Laws of the road?
.
Traffic jams most often occur during periods of peak usage (rush hour). How much of this is attributable to driver attention vs the inherent inefficiency of many independent vehicles attempting to coordinate speed, position, exiting, entering, etc? I don't take for granted that independent (vs synchronized) FSD would be significantly better.
FSD cars that coordinated with each other could leave lights even more quickly and better collaborate to zipper, observe/communicate/predict lights and modulate speed to prevent excessive stopping, etc.
I can’t see how that won’t negate and even be worse than any benefit from improved reaction times.
Any attempt to solve for repositioning just takes you closer and closer to buses or trains.
Not necessarily. FSD services would probably optimize for occupancy, which would mean limited and unpredictable availability for customers. Since the fleet is a capital investment rather than being floated by gig workers, supply would presumably be more constrained than Uber etc are now, and we see what happens at peak demand with them. I don't see that business logic changing between Uber drivers and FSD.
The really big win here is parking and time spent looking for it. Self-driving trips can just be shorter
Parking is a problem because of our dependence on cars. You also don’t need to park if you take a bus or a train and the things you want to get to are built near the bus and train stations.
> You also don’t need to park if you take a bus or a train and the things you want to get to are built near the bus and train stations.
It is nonsensical to compare self-driving cars against buses or trains in a situation where we still have tons of private cars on the road.
FSD can totally improve buses too, and well-designed multi-passenger vehicles can certainly easily end up making up most of the fleet. You could for example seat 4 people in completely isolated spaces without needing outrageously large vehicles. That'd be great for serving commutes where loads of people are going to the same areas.
Should, but if you really want to sell it to the public, you want to keep the service as close as possible or better than what's currently available. That's more feasible with fewer passengers on a vehicle, too many stops will easily double your travel time.