3 ms·

It's just a matter of whether it's 40/60, 50/50 or eventually 80/20. The point I think is still valid: some fixed cost gets you half the way from where you are

It's just a matter of whether it's 40/60, 50/50 or eventually 80/20. The point I think is still valid: some fixed cost gets you half the way from where you are, towards 100. You will never get to 100, and the place where you might stop and think "This will be more expensive than a private driver for the coming decade" is probably lower than we think (say 90/10 or even 80/20). Making a car that can do 99% of my highway driving or 90% of my driving time will be easy, possibly even in my lifetime. But neither of those makes for truly autonomous taxis, and they won't drive me home from the pub when I sleep in my back seat, so I really don't see the point.

ghaff 5y ago
I actually think fully autonomous driving for some subset of highway driving would be a really big deal--both for reasons of convenience and safety. But you're right. It doesn't buy you a taxi service which is what a lot of people care about.
- Digit-Al 5y ago
  That's what I was just thinking. You could potentially have autonomous geofencing. You enter your destination into the satnav and then start driving; when you get to a geofenced highway that is considered "AI safe" the car will give you a signal. At that point, if you wish you can set the car to autonomous mode and your seat will automatically move back a bit and tilt back so you can relax; maybe even take a nap (not sure about that though). When you are about 10 minutes away from exiting the geofenced section the car will return your seat to the normal driving position and sound an alarm so you have time to get your brain back into gear and take over control again.