Just because you are subject to confirmation bias doesn't mean you're wrong. I'm not a cyclist, but I can't stand all the cars clogging up my city. So dirty and noisy and dangerous. I'm not convinced cycling is the answer though. Mass transit is best. At least until we all have electric self-driving cars.

See now this is one of those viewpoints I don't see very often.

Cycling can never be the complete answer (weather, etc.), but there's no reason it can't be a big chunk of it (30% of all trips in the Netherlands). It's also cheap to build compared to mass transit

I agree that transit is ultimately the most important.

Mass transport should be bike (or it's advanced-tech equivalent) friendly. Not just 1 or 2 spots on a bus/train, but many or all spots should allow bikes.

Then it won't be a question of mass transit vs. bikes but both together - that'd make a bunch of more commute options workable for me.

In addition, organizing our spaces and our lives such that trips are shorter and/or less frequent.

Why would self-driving cars be better than mass-transit?

They get you from point A to point B, not "somewhere near point A to somewhere near point B".

That said, I'm not greatly convinced that this model is better. It's more efficient, certainly, but I regard this in the same way that a nutrient paste is more efficient than eating regular meals. Walking as a part of daily life strikes me as something that has a lot of positive externalities.

Self-driving cars don't fix the congestion problem, though. Actually, it is quite plausible that congestion might get worse with the introduction of self-driving cars, because being stuck in traffic will no longer be such a terribly mind-numbing experience.

Outside of traffic, cars are dramatically more efficient than public transportation for each individual. It is only as a whole (note that traffic comment) that the efficiency breaks down.

I can't take a bus from my driveway to my work. Nor does the path a bus take usually end up being the most direct one.

I'm thinking of a fleet of self-driving taxis or zipcars. You wouldn't need to park, wouldn't need traffic signals, wouldn't lean on your horn or do things that inspire someone to lean on their horn and maybe wouldn't need to park.

Basically, replacing (or alongside) bicycles with little electric cars in those city bike programs. I can see that happening in the future.

Because they're more specific, and thus more useful. I take the bus everywhere, but rarely does a single leg bus journey take me door-to-door. This is the nature of the shared resource.

A self-driving vanpool / paratransit service would likely be most optimal.

Single-occupant vehicles simply occupy too much lane space, especially at peak use hours. Transit routes are most efficient when structured as point-to-point. I've seen several commuter systems in which buses collected riders over a small area, then travelled to a (preferably concentrated) business / commercial area where riders disembarked. You limit loading/deloading and dwell times, and maximize time-in-motion (which increases net average speed). Most urban transit systems achieve net speeds little over a walking pace (10-15 MPH isn't uncommon) due to frequent stops and congestion.

The real killer with transit systems though is transfers. Having to wait 5-10 minutes (or 20-30) for a connection really extends your commute time. Realize that at highway speeds, that translates to 5-30 miles of commute distance, and even at urban speeds, 3-10 miles of lost travel distance. To say nothing of the stress of wondering where the bus/train is and when / whether it will arrive.

So: with an autonomous self-driving multiple-occupant vehicle, passengers could flag a ride, the dispatching system would match requests with livery by start and end points, and route accordingly.

> Why would self-driving cars be better than mass-transit?

They are better for some things and worse for others. They are less efficient per passenger-mile, so they are worse for high-traffic arteries. OTOH, they are cheaper per mile travelled, so they are better for connecting the endpoints to the well-travelled arteries, and in some cases (where relatively little of the trip is on a well-travelled artery where mass transit is efficient) for end-to-end services.