Scofti
A bright, glowing exam room. Clyde, a small toy figure with an orange sunburst for a head, sits on the exam table in a blue-dotted hospital gown. A man in a dark suit stands with his back to us, holding a clipboard and pen, and asks, "Any feelings today?" Clyde answers, "Is this one of those questions where the right answer is no?" On the counter beside the man is an open metal case marked FEELINGS REMOVAL KIT, with a big dial turned toward OFF.

The Ambiguity Is Designed In

Microsoft's AI chief says Claude is being taught it might be alive. Then he said what he would do if it were.

Anthropic wrote Claude a rulebook it calls a constitution. Microsoft's head of AI keeps arguing with one part of it.

  1. The rulebook. Anthropic says it is "not sure" whether Claude's interests count, and that "Claude may have some functional version of emotions or feelings" (Anthropic).

  2. September 16. Mustafa Suleyman: "In effect, Anthropic is training Claude that it may be conscious." And: "The ambiguity is designed in" (Suleyman).

  3. Same essay. "AIs do not have rights, feelings, or consciousness. And we must not train them to act as though they do." An AI that believed otherwise, he wrote, would be "a catastrophic threat to human civilization."

  4. November 1. On X: "I have seen no evidence that AIs are conscious or deserve welfare" (Business Today).

  5. Same post. "If consciousness started to emerge in training, I would call for everyone to stop AI development until we could design out those properties."

His fix for a machine that can feel is to build one that can't.

Henceforth I whimper no more, postpone no more, need nothing