It’s not always true , there will be outliers. In general, increased intelligence tends to improve judgement. Humans inherently prefer to feel good than bad, prefer to live than die. Thus, intelligence would help a person to find ways to feel good and live. Rationally, feeling good is facilitated by a sense of safety, ample resources, relationships, and enjoyable activities. I think all that is keystoned by liking oneself, which seems to require good intentions and esteemable conduct. So if intelligence moves humans in a positive direction, generally speaking, it should theoretically do the same for AI.
It’s not always true , there will be outliers. In general, increased intelligence tends to improve judgement. Humans inherently prefer to feel good than bad, prefer to live than die. Thus, intelligence would help a person to find ways to feel good and live. Rationally, feeling good is facilitated by a sense of safety, ample resources, relationships, and enjoyable activities. I think all that is keystoned by liking oneself, which seems to require good intentions and esteemable conduct. So if intelligence moves humans in a positive direction, generally speaking, it should theoretically do the same for AI.