Re:US will say "no" to global ban on killer robots
by bryan0 on Monday November 16, 2026 @07:26PM
(#50525696)
I would say the books showed correctly, that it is impossible to avoid "trolley problems", but that is okay and expected. It's still directionally correct to imbue machines with the directive to not harm humans.
Reality cannot work in the way the books did, at least not with any type of computing architecture we've made so far. I'd go further and say it's antithetical to general intelligence and problem solving.