And *who* exactly are they going to 'look to' to determine what humans 'really want', given that we all want different things..fundamentally different things. From the individual+clan level all the way to the cultural+ societal. Also temporal scales? 10yrs? Centuries? Millennia?
-
-
-
In particular on the infinitely unique individual level
Kraj razgovora
Novi razgovor -
-
-
Agree with Russell instead of machines pursuing goals of their own they should satisfy human preferences and behavior. These major challenges but can be addressed and focus for one of my AI gigs 1) behavior not rational 2) preferences change depending on mood or circumstances
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Big step in the right direction but ultimately, it should instead do what we *should* want, like http://metaethical.ai
-
But again,what 'should' we want? eg. what's best for: our species as a whole?/a sector of our species?/biological life in general on earth?/intelligent life (including artificial) in the universe?What's best for: those living now?living 500yrs from now?/in 1M yrs?And so on..
Kraj razgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.