Researcher: Treat Consciousness Definitions as Testable Hypotheses, Brain-to-AI Link May Be Testable Soon
kanair · x · 2026-10-05
One view holds that saying "consciousness hasn't been defined" isn't very useful — we haven't even defined what it means to define something. In natural science, a definition is better thought of as a hypothesis: a useful one should constrain a theory that makes non-arbitrary predictions, and ideally allow us to derive theorems.
The author also argues we are reaching the limits of current approaches to consciousness. Real progress requires much more neuroscience, especially experiments testable in humans. For example: what qualia would a person experience if we connected the human brain directly to an AI system? The technology may already be close to testing this empirically.
More from AGI Musings
- Lee Kai-fu brokers Kazakhstan's school AI pilot: 10 schools now, 500 by April 2027 — kaifulee · 2026-10-06
- LLMs poisoned the online Q&A paradigm — and are now the new Q&A paradigm — max_paperclips · 2026-10-06
- Designer's Steve Jobs tribute: from hacking a beige PC into a Mac to lamenting tech's refinement loop — ryolu_ · 2026-10-06
- Functionalists point to MoE vs dense models to argue brain compute debate — JoshPurtell · 2026-10-06
- Rob LeClerc backs the case that AI could let humanity retake bigger risks thoughtfully — robleclerc · 2026-10-06
- iamtrask boosts token-attribution idea to fix enterprise AI's incentive misalignment — iamtrask · 2026-10-06