GM seeks US approval to deploy self-driving car without a steering wheela
arstechnica.com
arstechnica.com
1. The car is now under the complete control of a computer - not a human driver. There are only a limited number of things the passengers can do to stop the vehicle or improve their situation in case of an incident.
2. It _requires_ a computer to make complex moral decisions about the safety of the passengers, dictated by the preferences/views/biases of the manufacturer/regulators.
It seems that we have crossed the line between "should technology solve this issue" and "could technology solve this issue".
I think Tesla has it right with their implementation of self-driving capabilities - a human is present and logically fully responsible for the vehicle at all times.
As a point against this specific vehicle: Waymo’s vehicles occasionally get “stuck” and have to be recovered manually. (Waymo dispatches a driver who comes and takes over the car.) How does Cruise plan to deal with situations requiring manual intervention, where remote labeling and pathfinding isn’t enough?