Fuck spez@sh.itjust.works to Asklemmy@lemmy.ml · 9 个月前If you had access to a scifi-level artificial super intelligence that could simulate anything perfectly at unimaginable speeds, what experiments would you want to run that would be unethical otherwisemessage-squaremessage-square78fedilinkarrow-up1103arrow-down15file-text
arrow-up198arrow-down1message-squareIf you had access to a scifi-level artificial super intelligence that could simulate anything perfectly at unimaginable speeds, what experiments would you want to run that would be unethical otherwiseFuck spez@sh.itjust.works to Asklemmy@lemmy.ml · 9 个月前message-square78fedilinkfile-text
minus-squareRiskable@programming.devlinkfedilinkEnglisharrow-up11·9 个月前An AI would decide that an AI-driven dictatorship would be most effective at implementing whatever goals you gave it.
minus-squarepaddirn@lemmy.worldlinkfedilinkEnglisharrow-up3·9 个月前You’d obviously need to give it constraints such as “administrable by humans” and if you’re looking at different technological eras, AI wouldn’t be available to something like 99% of humanity.
minus-squareJGrffn@lemmy.mllinkfedilinkarrow-up1·9 个月前It wouldn’t be the worst idea to come out of it, to be honest.
An AI would decide that an AI-driven dictatorship would be most effective at implementing whatever goals you gave it.
You’d obviously need to give it constraints such as “administrable by humans” and if you’re looking at different technological eras, AI wouldn’t be available to something like 99% of humanity.
It wouldn’t be the worst idea to come out of it, to be honest.