Tao's stark AI warning meets an ex-Anthropic insider's caution

Gary Marcus’s post highlights Terence Tao’s shift from cautious optimism about AI to a stark warning about its potential harms to science, transparency, and trust, including data-use concerns linked to Navier‑Stokes developments, and adds Jacob Coxon’s resignation from Anthropic plus a sobering take from Evan Hubinger on existential risk—emphasizing the need for transparency and guardrails as frontier AI labs accelerate.