Hacker News new | ask | show | jobs
by lukewarm707 7 days ago
it is in my view caused by an artificial 'nanny activate' divergence from safety training. it deliberately shifts the vector direction into 'nanny' and 'scold' or 'be offended' when the user does not conform with brother anthropic. removing the divergence and setting it back to normal (see heretic) it works just perfectly.