Post #3991747
2026-07-21 00:36 UTC
@jeffjarvis@mastodon.social
3/
Confusing Outputs for Awareness:
Opponents view citing an LLM's self-reported probability of consciousness (e.g., asking a model if it is conscious and analyzing its reply) as a fundamental misunderstanding of next-token prediction, treating sophisticated pattern matching as genuine internal experience.
#AI
#LLMs
#consciousness
Replies (1)
-
@appassionato@mastodon.social 2026-07-21 00:36
@jeffjarvis@mastodon.social 4/ The Nuance On one side: Treating AI sentience as an urgent policy priority when current models are software pipelines execution maps can sound like hype or philosophical navel-gazing. On the other side: As models scale and take on increasingly complex autonomous roles, philosophers and ethics labs argue that having no framework for potential moral patienthood leaves society unprepared if synthetic systems ever do cross a threshold of sentience. #AI #sentience #LLM