FULL STORY
Gary Marcus Downplays Extinction Risk, Then Clarifies Agent Pause Call
Gary Marcus said AI's odds of extinguishing humanity are near zero, while urging guardrails and a pause on deploying internet-connected general AI agents. He later clarified he was not calling for a blanket AI pause.
2026-09-15 ~ 2026-09-15 · 2 episodes · 6 posts
Episode 1 · Gary Marcus dismisses near-term AI doom but warns of disasters without enforced guardrails (2026-09-15, 4 posts)
In an India Today interview, Gary Marcus said the odds of AI wiping out humanity in the coming years are near zero, yet warned of serious AI-caused disasters, arguing guardrails must be externally enforced rather than left to profit-driven AI giants.
- Gary Marcus: We must think about AI in terms of collaborating to make the world better — GaryMarcus · 2026-09-15
- Gary Marcus takes p(doom) and 'pace the frontier' debate to India Today — GaryMarcus · 2026-09-15
- Gary Marcus: AI wiping out humanity in a few years is "close to zero" probability — GaryMarcus · 2026-09-15
- Gary Marcus on TV: AI guardrails must be enforced, not self-imposed — GaryMarcus · 2026-09-15
Episode 2 · Gary Marcus Clarifies: Pause Only Networked General AI Agents (2026-09-15, 2 posts)
Gary Marcus clarified that he is calling for a temporary pause only on large-scale deployment of networked general-purpose AI agents—which he says have proven unreliable and potentially harmful—rather than a halt to AI as a whole.
- Gary Marcus clarifies: he wants a pause on internet-connected AI agents, not all AI — GaryMarcus · 2026-09-15
- Gary Marcus clarifies: pause internet-connected AI agents, not all AI — GaryMarcus · 2026-09-15