[ExI] Autonomous car ethics

Anders anders at aleph.se
Sun Jun 26 12:47:28 UTC 2016


On 2016-06-25 19:16, Chris Hibbert wrote:
> > Should a self-driving car kill its passengers for the greater good ?
> > for instance, by swerving into a wall to avoid hitting a large number
> > of pedestrians?
>
> This is entirely an academic philosophy question, and not at all a 
> question that needs to be resolved before self-driving cars should be 
> put on the road.

Yup. I often end up having to explain to engineers that they mainly need 
to think about this kind of ethics if their safety engineering is too 
bad. There has to be an ethical decision (or non-decision) when all 
safety has failed and bad things will happen, but the importance of 
ethics is to explain to the inevitable court case why the code looks 
like it does.

Trolley cases happen in real traffic and need to be resolved, but they 
are rare enough that they are not show-stoppers for autonomous cars. The 
real issue will be things like how they handle rain and snow, or a 
packed inner city. But long before that they will take over in mines and 
factory facilities.

(I am just back from an AI meeting hosted by a big German car 
manufacturer. Go figure.)

-- 
Dr Anders Sandberg
Future of Humanity Institute
Oxford Martin School
Oxford University




More information about the extropy-chat mailing list