Re:Driver hits woman in S.F., then Cruise driverless car run...
by techdragon on Wednesday October 04, 2023 @12:28AM
(#37760850)
No the robot was left holding the bag after a disgusting person drove into this pedestrian and as a result of hitting the pedestrian with their car tossed the pedestrian directly in front of the driverless car. Which immediately freaked out and tried it’s best to pick a safe failure mode… in this case it picked wrong and that’s a problem… but a human is clearly the primary offender here and the driverless car was given an out of context problem because the human broke the law by doing a hit and run on this pedestrian
[ Parent
| on HN ]
Re:Driver hits woman in S.F., then Cruise driverless car run...
by watwut on Wednesday October 04, 2023 @01:07AM
(#37761041)
Generally I agree, but it was not out of context problem. It is sort of real world occurrence drivers (both human and non) are expected to handle.
Re:Driver hits woman in S.F., then Cruise driverless car run...
by techdragon on Wednesday October 04, 2023 @01:13AM
(#37761072)
It did try to handle it, by coming to a complete stop as fast as possible and avoiding further movement. I definitely think it should be smart enough to not stop “on top of an obstacle” under the car… but as far as everything before it stopping on the pedestrians leg… seemed like it would be hard to get a human driver to do better given the likely lack of a lot of “hit and run happens right in front of me” training data… zebras vs horses and all that. I expect one they will definitely add more “accident close by” data to the training data… and two they will get the damn thing not to stop on top of an unknown obstacle.