3 ms·Anthropic says they've found a new way to stop AI from turning evil6 points by Brajeshwar 1y ago