>I think you're overselling the extent of the difficulty here. People from out of state usually manage to get through New York, pedestrians and all, without getting into an accident.

At this point, "it's easy for a person to do it" is probably better evidence that it's a hard problem in AI.

In this case, I believe it's an inverse learning problem. We observe other agents acting in an environment, deduce their policy, then implement that policy ourselves.

A self-driving car that can reconfigure it's software to mimic the behavior of other cars it observes, sounds difficult to me.

> We observe other agents acting in an environment, deduce their policy, then implement that policy ourselves.

This is essential, especially internationally. If you're an American you think it's perfect normal to be allowed to turn right even when the traffic light is red. You don't get told not to do that when you hire a car in Italy. Now an automatic car can be programmed with different rules depending on which country it's in, but even in the same country, you wouldn't enter a box junction when the exit isn't clear, at least in most places. Turn up in rush hour London and you find there's no way to actually get through the junction unless you do that, because others fill the junction so when your light is on green, the exit isn't clear.

I remember pulling upto a red light late an night. Empty road, but the red light only goes green when it detects a car. It didn't. What would a fully automatic car do in that case?

I always drive differently, more cautiously and slowly while near colleges. Especially during the night. It is much more likely for college students to be intoxicated and even if they are paying attention they might not be capable of realizing a car is coming etc... This is also true near elementary schools, but not because they might be intoxicated but because little children are more likely not to pay attention. I really jope self-driving cars are trained for such nuances.

You don't need it. Companies will program the cars for the various regions where it's profitable to do so, and block the self-driving features elsewhere. Much like geo-restrictions in streaming platforms, and such. Will anyone in the US not buy a self-driving car because it doesn't work in, say, Portugal?

That sounds much harder than just stopping if a pedestrian is on a trajectory to enter your lane. I mean, obviously it would be better to deduce their intentions and act according to that but I don't see any reason to say that that's necessary to have useful self driving cars.

A car that defaults to braking without any deduction beyond detection will never make it through an intersection, let alone a whole trip, in some parts of the world.

New York looks like a well regulated army training course compared with somewhere like Cairo.

Never is a long time, but I don't see self driving cars coping well when traffic is coming the wrong way down the street, let alone when cars are driving along the sidewalk

Yes, Cairo will probably have self-driving cars over a decade after New York has them. But that doesn't mean self-driving cards aren't a commercially viable product before they work in Cairo. Very few technological innovations work better than the thing they replace in all circumstances initially. Back In The Day there were a lot of places a horse could go that a Model T couldn't but that didn't prevent cars from replacing horses.

Works for one SDV in a stream of human-driven cars. Doesn't scale well.