This week in AI — 2 stories

This week's stories share a quiet common thread: the growing tension between AI systems that move fast and the human structures - policy, cognitive diversity, democratic deliberation - that are supposed to keep them accountable. You'll learn why Bill Gates thinks the industry's self-imposed guardrails have already failed and what he's proposing instead, plus how researchers are tackling a subtler but related problem: the way large language models, left to their own devices, tend to flatten thought into a kind of statistical consensus that can masquerade as intelligence.