On Thu, 22 Dec 2016 02:45:39 -0800 (PST), wrote:

No, we don't tell it what to do (beyond the high level control of
providing a destination- no different than we do with a human driver).
We obviously provide rule classes,

Rule classes = word salad, peterson.

It demonstrates that you are a fool when you speak about things you
are wholly uneducated about.

such as ethical guidelines, road
rules, etc. Again, no different from human drivers. But there still
isn't going to be a rule that requires a wall over a human, for
instance.

A choice still has to be made, but which human gets to make it?

The people who design the machines, observing the commonly understood
ethics of their society as well as the recommendations of people who
study ethics. This is how things operate now in many areas, and
there's no reason to think the situation will change.

The deep learning is going to result in complex and largely
unpredictable behaviors based on the actual experience of millions of
autonomous vehicles.

The machine isn't learning, that's just hype.

It demonstrates that you are a fool when you speak about things you
are wholly uneducated about.