I think this is the key takeaway from this article:

<<Expect level 4 cars to take quite a while longer to reach private hands. "The economics simply don't make sense," [Ford CTO] Nair said.>>

Indeed: how much more are you willing to pay for a car where you don't do anything? Not much. How much more is a cab company prepared to pay for a car that doesn't need a driver? A lot.

For private cars I expect we'll see more and more refinements of driver assists in the premium market segment, but not much apart from the odd trickle-down to cheaper cars. Reason being LIDAR, tons of sensors and data processing power will still be very expensive, and in the cheap end of the market, people are still skimping on carpets and electric windows.

One example that immediately springs to mind is the flashing warning light in the mirror when another car is in your dead zone. Relatively simple, useful, "stupid" (as opposed to "smart") tech. I expect it will be standard equipment in five years.

Curious about what drives the cost difference vs. a "consumer" car. The sensor suite? (Is it just lidar vs. radar?)

I don't think the cost will be different for a consumer car or a fleet car. I'm just saying the cost will be pretty big, and for consumers the value you get is "I can relax while driving" (not worth very much), while for cab companies the value is "we don't need to hire any cab drivers" (can easily be worth $100 000 per car over five years).

Sensor suite, software licensing, and data licensing. If solid state lidar can get there soon enough, that will help. Data licensing for extremely accurate data for everything - traffic lights, stop signs, every single intersection, etc.

Really interesting; thanks.

I think the important take-away here is that all the major players are now in a public arms race. Which means autonomous cars are almost certainly going to happen.

I think we've all come to accept that Level 3 (return control to driver) is just not feasible.

Without the sensory stimulus that comes from driving the average harried American will be asleep in under 5 minutes.

Apparently, 'just five short years away' is the new 'just ten short years away' - an mythically ambitious number, remains irrefutable because nobody, including the people making the claim, can see that far into the research and development process.

,,...Fields said, adding that ride-sharing would make more efficient use of vehicles, with less time wasted for people and less pollution.''

If he would really care about pollution he would make sure that these shared cars are electric.

I think you'd do the self-driving market a disservice by bundling it with expensive range-limiting technology. If both are positive changes in their own right, then they're better off keeping them available separate for maximum adoption.

That's even more so the case with self-driving cars, since their ideal use cases (highways) are the least ideal use case for vehicles with short ranges.

It's like a Shit Elon Says quote, only Ford doesn't have any sorts of track record here whatsoever.

"Twenty launches a year, is not a crazy number at all. We expect that to occur without any miracles."

What in the name of anything are you on about? Ford has been doing cars for how long?