The Most Dangerous Type of A.I. Company / ai, AI (Artificial Intelligence), AI & Machine Learning, AI & ML, AI & Society, AI 2025 trends / By hi@aiweekly.co.in Meta may be an example of one of the most dangerous types of A.I. companies — one that believes it’s building tools, when it’s actually building superintelligence. Kevin Roose and Casey Newton discuss it on the “Hard Fork” podcast.