Tao's blue team stands for generative "AI", the red team stands for critical/auditing "AI". I have not seen any independent claim that generative "AI" makes pr

Tao's blue team stands for generative "AI", the red team stands for critical/auditing "AI". I have not seen any independent claim that generative "AI" makes programs safer or that generating supervising features as you suggest works. For auditing "AI" I have seen one claim (not independent or using a public methodology) that auditing "AI" rakes in bug bounties.