'Pace yourself' vs 'Stop Frankenstein': Nividi CEO Jensen Huang, OpenAI chief Sam Altman and US VP JD Vance clash over AI safety <br> The debate also covers independent AI evaluators, model alignment, safety testing and the use of computing resources. Mark Zuckerberg has backed independent assessments and Meta’s decision to delay Muse for safety work, while JD Vance has urged AI companies to build defensive tools against cyber threats created by advanced models.