Using radar, cameras, GPS, sensors and wireless technology, cars will be able to “talk” to each other and navigate safely by knowing where they are in relation to other vehicles.

"Reduce traffic" means "fewer cars on the road." "Smoother traffic flow, an end to traffic jams and greater safety" isn't talking about reducing anything. The advantage of fleet autonomy isn't about a reduction in traffic, it's about a coordination in traffic. Better coordinated traffic means denser traffic, means more cars on the road.

This is where the hand-wavey shit starts to get out of hand: the way you "reduce traffic" is by increasing the attractiveness of driving alternatives. 30 people on a bus takes 30 cars off the road. 30 people on light rail takes 30 cars and a bus off the road. 30 people on bicycle lanes takes it all off the road. But 30 people in driverless cars is still 30 cars on the road.

So I reduce congestion by requiring autonomous transport through the middle of London, for example. Now all the delivery drivers know that the majority of traffic goes through the middle of London. So they go around and take their traffic elsewhere, where they can still get ahead of the pattern. You haven't reduced traffic, you've moved it. Yeah, there's ten thousand people surfing Facebook on their phones while their Priuses sit in traffic... but what was a knot in the middle of town is now a knot going around that knot.

Meanwhile, the cost of "congestion reduction" has been fobbed off on the individual consumer, rather than the tax base. Build a road? Public works. Require autonomous vehicles? Well, now you're dealing with a new car. Are you going to subsidize lower income levels the way they supposedly do with PACE lanes and the like? 'cuz now instead of your tax dollars going to local contractors building local roads for the local community, you're sending money to Toyota.

IMO the first step will be making our public transportation driverless, which should reduce the cost.

By what logic? What percentage of the cost of a bus ticket do you really think goes to the driver? Better yet, what real cost of a bus ticket is paid for by direct subsidy?

Furthermore, various Zipcar-like services should be able to offer "subscription" services for people.

They do already. Friend of mine has had a Zipcar subscription since 2005. "Driverless" changes this how? 'cuz I guarantee - they're going to want you to have a license in case the computer fritzes.

Subscriptions will obviously vary in cost, but they should prove fantastic and cheap alternatives for people like me that A) live in big urban areas, B) can't afford parking/insurance costs, C) find it inconvenient to own a car in the first place and don't like driving in congested downtown areas.

So why don't you have a Zipcar subscription already? Because it won't schlep itself to your door? Because driving downtown is that much of a hindrance? How does "self-driving" affect the equation here?

Although there isn't likely to be a massive reduction in cost unless we can find a way to efficiently exploit renewable energies for transportation.

Has nothing to do with vehicular autonomy.

But we have cars on this planet right now that have driven for hundreds of thousands of kilometers in real-conditions without getting into an accident once.

And millions more with similar brags being driven by humans. I'm not saying "autonomous cars are a stupid idea" I'm saying they're still cars.