Re:O3 beats a master-level GeoGuessr player, even with fake ...
by _heimdall on Tuesday April 29, 2025 @10:25PM
(#43840555)
That's the alignment problem. We intended a certain set of rules but didn't define them completely, or there were conditions we didn't consider.
An AI wouldn't have to maliciously break a rule to go wrong. The point is that the system could so exactly what it was supposed to do, it plays within the given rules but the outcomes aren't what we wanted or intended.
An AI wouldn't have to maliciously break a rule to go wrong. The point is that the system could so exactly what it was supposed to do, it plays within the given rules but the outcomes aren't what we wanted or intended.