jont- wrote: As for court, I thought we were supposed to get a jury of our peers, so really it shouldn't be a jury of anyone unless they're also advanced trained.
Good Gods no! Could you imagine, based what it's like on here "Blah blah BGOL blah blah rotational blah blah" - cost a fortune in legal bills . . .
jont- wrote:It's an interesting precedent for autonomous cars that "better" isn't good enough (if we're being sold on the idea that they are "better" than human drivers but can still be fallible).
To be fair, the article is 'better but did it wrong' [eg maintaining offside cornering position in the face of oncoming traffic].
Autonomous/automated cars will only 'learn' and use the rules they're given. Currently they may struggle with some aspects [understatement?].
For example: try to explain in simple AV friendly terms, the difference between a safe following distance and the 'close up / contact' pre-overtaking position
You're getting into the realms of teaching AVs to judge and balance risk - and researchers are working on that (I saw a video from MIT CSAIL on exactly that sort of thing, for an AV in lane 1 with a vehicle emerging from an on-slip, with the AV having to decide whether to slow to make space or accept a shorter than ideal gap in L2).