I would want a manual override, at least. If I want to direct the car myself right this second and there's no way to instruct a computer interface in time, I'd want to be able to hit a big red button and grab the wheel.

So why don't we cut down forests, bury the carbon our own way, and then plant more trees?

The difference between horses and humans is that horse populations work to serve human populations, and human populations work to serve human populations. When we make the economy run by robots, it won't be human populations working to serve robot populations, it will be robot populations working to serve human populations.

With those circumstances, I would think that luck is the situation of, when you are not aware of all the relevant variables involved in the result, hoping that the variables hold values that produce a result you like. Luck would be that hope validated.

So rolling a die and hoping for a 6 means that you are rolling the die and hoping that the relevant variables of the universe that you don't know are going to naturally make the die end up a 6.

Better analogy is having a social circle of overweight people and collectively agree to go to the gym together, and then only you show up.

Sure, you're fit, but the group's still a group of fat people.

And yet the greedy company finds itself one customer less than before.

What we're hoping for is that if insurance companies, even a small number, start offering low rates (with fewer crashes, they can afford to lower rates), then customers will accumulate there and the greedy companies will be forced to adapt or die.

True, being an expert in one field doesn't make you an expert in all fields, but it does state quite clearly that the person is an educated and intelligent person. The defense the programmers were using was that the OS was easy to understand for intelligent people, and that only idiots would have trouble. When an intelligent person had trouble, that defense fell apart.

700-1000C

Think back to the past, when the theories of 'modern' physics were incorrect. The could have said the same thing: "Point to one of these theories that have been disproved." and no one could because those theories had yet to be disproved.

If someone back then were also to say "Point to one of these theories that we know will never be disproved." no one could truthfully point to any theory. Surely, one of the theories available to point to eventually became disproved, but how were they to know that? How could they know which theory ends up accurate, and which theory turns out to be inaccurate?

How are we to know which of our theories are accurate?

Then the tipping point is when the cars can handle any emergency a human could and anything it can't handle couldn't be handled by a human either.

I'm sort of expecting a functionality amounting to: if an automatic car has no idea how to proceed, park by the side of the road, put the hazard lights up, and request human takeover.

Since when is it the problem of a company hiring top tier workers that the places producing top tier workers have different diversity numbers than the places producing the rest of the workers?

You are correct in that as material objects become cheaper and people still have jobs, people with jobs accumulate more material objects and standards of living rise.

But that doesn't mean that automation will keep human labour from being relegated to those who specifically want a task less efficiently done by a human hand. Robots became better than us at manual labour, so long as they were doing the exact same thing over and over again, and now we are getting robots that are better than us at manual jobs with more varied demands.

Going by our current state of software, we will eventually reach a point where any necessary profession is most efficiently done by a robot, even intellectual ones. From that point, any human labour is based in illogical sentimentality, as a robot could do the same thing better and cheaper.

Robots could make robots, robots could repair robots, robots could design new robots for new tasks, and humans find themselves with no place in the job market, only left to reap the fruits of robotic labour and engage in whatever hobbies they wish.

I can see a company making an AI to generate beautiful imagery that they can sell.

Drill sergeants cost a significant amount of money to maintain, whereas if this technology is produced in large quantities it can become a valuable investment as a replacement for drill sergeants.

If they both accomplish the same thing, and one can be cheaper than the other, why would you not choose the cheaper one?

Two questions:
1) How do those numbers compare to the number of men and women in the military?
2) How do those results from 1) compare to national sexual assault rates?