3 ms·
Making the "bad AI" vs "good AI" distinction pre-training is not feasible, but making a "bad use of AI" vs "good use of AI" (as in bad/good for the people) seem
by jacobgkau 2y ago
Making the "bad AI" vs "good AI" distinction pre-training is not feasible, but making a "bad use of AI" vs "good use of AI" (as in bad/good for the people) seems important to be able to do after-the-fact (and as close to during as possible).