This seems obviously wrong? Any system whose name includes the word "forecast" was built to predict the future in some domain / over some time horizon / to some level of granularity.
It's an interesting thought, but isn't that still a statistical response to stimuli based on learned experience? Albeit one more advanced and subtle
It no more requires reasoning about the future as such than does stopping when someone or something is actually in the way (and thus the car will hit it in the future)
I’m not sure about that, I mean this is something that client-side prediction in games is doing all the time, so why wouldn’t a self-driving car do it?