That's a bit like suggesting that autopilots and automatic landing systems are the wrong solution to reducing air accidents. The only difference is that automatic flight systems in aircraft were introduced in the past and self-driving cars will (presumably) be introduced in the future, but the validity of the approach to the problem is basically the same.
No, my thinking goes along the lines of society wanting/needing to use less cars not more. hence I would be more for mass transit type solutions rather than packing more cars onto the road more (space) efficiently.
It wasn't directly related to traffic accidents in the article, but to what the parent comment said at the end.
Well yes, using cars for urban transport is a bad habit that we should never have got into, and we should now be trying to get out of. But it's logistically and politically impossible to get rid of all cars overnight. There are a number of things we can do to cut down on car usage. Self-driving is something we can do to cut down on the fatality rate per car. Each of these is a partial solution. Partial solutions are complementary, not exclusive.
But self driving cars are a great way to cut down on the number of vehicles on the road via very cheap, highly utilised automated taxis. Vehicles should last a lot longer too because the autopilots would make fewer mistakes such as over-gunning engines, stalling, braking late and terrors more sharply, etc. they could enable huge efficiency savings.
