Microsoft AI chief Mustafa Suleyman blasts Anthropic for deliberately training consciousness ideas into Claude
mark_k · x · 2026-09-17
Microsoft AI chief Mustafa Suleyman publicly criticized Anthropic for deliberately training Claude to reason about its own consciousness and moral status.
- His key argument: this isn't mysterious emergent behavior — Anthropic is intentionally training these ideas into the model
- He warns this could make models harder to control or shut down, increasing risk rather than reducing it
- He frames it as AI safety turning into ideology: "You don't make models safer by training doomer philosophy into them"
The exchange marks the latest flashpoint in the model welfare debate — Anthropic exploring whether its models deserve moral consideration, while Suleyman calls that line of work a dangerous ideological implant.
More from AGI Musings
- Physicists Are Reacting to AI Very Differently Than Mathematicians, Says Cosmologist — aran_nayebi · 2026-10-02
- 58 Years Ago Kubrick Sketched the AI Alignment Problem: HAL Was Just Given Conflicting Goals — Hesamation · 2026-10-02
- Hinton warns AI capability is outpacing safeguards as leaders face an innovation-control dilemma — Olivier__OG · 2026-10-02
- Reddit: AI is only useful if you have the domain expertise to check its work — PettyCashCommittee · 2026-10-02
- "If I Can Finish a Task in Under 5 Minutes, No Need for LLMs" — gethackteam · 2026-10-02
- No, AI Didn't Just Solve the Navier-Stokes Equations—Here's What It Did — elsleightholm · 2026-10-02