Lower the chances of, not prevent.

Sure, but in the above user's hypothetical, that would mean that in such an area a concerned human driver with a greater ability to predict general human behavior would have a statistical safety improvement on the autonomous vehicle, which might not understand, for instance, that since it's a Friday night and the big game just ended and I'm in the city center I should be more careful than usual because there will be more intoxicated people.

Which is a lot of high level reasoning and inference with information from a variety of sources which aren't on the face of it strictly related to the driving task.

That's the type of thing that an excellent driver thinks about, but about 10x more thinking than I believe the average driver does.

Exactly, you can never reduce accident rates to zero, even if you slow to a crawl or stop (now you're endanger traffic behind you). Debris could fall on your car and make the steering non-responsive. A flat tire could burst at any speed and cause a bicyclist to veer into traffic/off a cliff/whatever.

When it comes to human drivers we deal with probabilities, but for some reason people want absolutes with autonomous ones.

This seems like the opinion of someone who drives too fast?

Sounds like the opinion of someone who knows minimum braking distances still exist even when going slowly.

If pedestrians or especially children are present, one should be driving very slowly. 15mph sounds about right. At that speed one is unlikely to kill, since even if a pedestrian "jumps out" (an event that is vanishingly less common than drivers not paying attention), one has enough time to stop.

Most drivers drive too fast. (Me too!) Most drivers have not killed anyone. All drivers who have killed someone, were driving too fast at the time. Many drivers will disagree with me, but they are simply wrong about appropriate speeds, and we will all be safer when robocars are driving for them.

Even when the minimum braking distance is not small enough to avoid hitting a pedestrian, a lower speed will impart a lower kinetic energy vastly reducing the odds of fatally injuring the victim.

This statement is either very poorly thought out, or needlessly accusatory and inflammatory.

I would say that if you claim that a massively complex system like public traffic, with thousands of participants, many of them badly or completely untrained, can be organized in such a fashion that any and all accidents can be prevented, the burden falls on you to show how.

This subthread is not about the massively complex system. Rather we are discussing the good judgment of individual drivers. Even if my statement were not literally true (the best kind, and it literally is), it would still be good practice for all drivers to commit themselves to driving slowly enough to prevent collisions with pedestrians.

This is the reason robocar firms will fight to keep all their data private. If it were public, researchers could show and personal injury attorneys could argue persuasively that there are some speeds that would never cause fatal collisions. Since those speeds will be slower than most passengers wish to travel, this mode of travel will be more vulnerable to lawsuits.