Armando Da Silva Vieira: The Illusion of the Neutral Machine
Armando Da Silva Vieira, Lecturer of Applied AI In May 2026, researchers from Oxford, Google DeepMind, OpenAI, Anthropic, and Stanford published a paper (https://arxiv.org/html/2605.10310v1) with an unusually candid premise. AI alignment, they argued, has been dominated almost entirely by preventing harm — safeguards, controllability, compliance. Necessary, but incomplete. Like a psychology that only studies illness […]