Subjective Experience in AI Systems: What Do AI Researchers and the Public Believe?
This study investigates divergences and convergences between AI researchers and the general public regarding “subjectively experienced AI”—artificial systems possessing phenomenal consciousness or subjective experience—addressing critical gaps in AI ethics and governance. Method: Employing a stratified mixed-methods survey, the study integrates probabilistic elicitation, Likert-scale assessments, and statistical significance testing to quantitatively compare perceptions across groups. Contribution/Results: Both groups estimate a 60–70% probability of subjectively experienced AI emerging by century’s end (median year: 2100), yet researchers anticipate earlier emergence and exhibit lower ethical acceptability. Over half of respondents endorse immediate developer-level safety safeguards; strong consensus exists on requiring accountability mechanisms and ethically constrained behavior for such systems. The findings identify robust cross-population agreement on governance fundamentals, while highlighting the central normative tension surrounding subjective experience in AI policy discourse.