2026
I asked 538 people if AI can feel. One question wasn't enough.
I asked 538 people to judge five fictional AI system scenarios. Emotion-like processes, subjective experience, and welfare turned out to be rather different questions.
October 2026A benchmark score is not a model
Three models can share a benchmark score and still require completely different supervision. Two tiny Go tasks showed why.
August 2026How do you ask people whether an AI can feel?
Explain the topic too well and you've taught respondents the answer. Don't explain it enough and they may not understand the question.
August 2026❦