An AI cannot ‘know’ in every sense
THE latest bombshell about large language models is that they now have feelings. Specifically, the internal mechanisms of Anthropic’s model Claude Sonnet 4.5 have been found to possess “emotion-related representations” that…















