[ExI] Do robot cars need ethics as well?
patrickkmclaren at gmail.com
Thu Nov 29 15:37:24 UTC 2012
On Nov 29 02:31 PM, BillK wrote:
> So it is not just making sure the car doesn't crash (as at present).
> Humans will require the robots to make judgement calls. And that is
> definitely not easy.
> Consider the bus crash example. What if it is a prison bus with
> convicted murderers in it? Does each human carry a value tag for the
> computer to use? Or is it personal survival that always wins? As the
> writer says, human morality is still a work in progress, so
> programming ethics into robots is problematical.
I would expect that in situtations wherein circumstances deviate from an acceptable level, a "minimise effect on environment whilst maximise human safety" approach would be satisfactory. Eg. Clip the rear end of the bus, deploy airbags at appropriate time.
Most other approaches would result in a race condition, with unpredictable outcomes. For example, personal survival might force other (unexpected) vehicles of the road. Enumerating the value tags of the passengers would reduce the time available for response.
More information about the extropy-chat