safetyopinionbearishThe entire future of humanity could rest on whether AI safety measures work as hopedGary Marcus03 Aug 2026https://x.com/GaryMarcus/status/2084027104082235770