3 Comments
User's avatar
Graham L's avatar

What a great discussion. Great work, down to earth, refusing to get lost in any inflated conceptual abstractions. I was a bit amazed when he said he couldn't tell whether he was talking to a chatbot, by the way, as in my experience - on ordinary things like bank websites - chatbots are absolutely hopeless and devoid of nuance or genuine human help or understanding, and I just always give up on them in despair. But I loved the comparison of AI consciousness with the consciousness of a modern car - it's just doing very clever stuff, that doesn't "mean" consciousness. I didn't go along with his saying that you look at a chimpanzee and think it's conscious, but you don't do that with say a cockroach. I do regard cognition-awareness-consciousness on one spectrum, so however primitive the awareness of the cockroach may seem, I personally regard it as having some degree of consciousness. Anyway, great guy, wish I knew him. Or worked for him.

Patrick Celka's avatar

If we ask an Aboriginal person living in the wilderness of the tropics and in the same invalidity situation as this Tennis one, I doubt that his brain will light up in the 'Tennis' zone, right? But then we cannot conclude he is not conscious (addressed at 42min and beyond...). I suppose that consciousness is prior to the brain being activated, and when it does, it requires the person's consciousness to access her/his memory and a sense of self? Therefore, it seems to me that the brain's response to a question is not a necessary condition for consciousness. Also, the person listening to the question must have the capacity to hear and understand the question before the brain activity can respond.

Did he really mean chimps are not conscious, or am I missing something?

How many examples of theories have been laid out and experimental tests have been done decades later? Having theories makes you refine your understanding of the questions and therefore of the foundations of the questions. But, yes, theories for theories are (almost) useless.

Doug Fay's avatar

Reminds me of old references to placing a bell over a gravesite and a pull-string inside the coffin so the body could ring the bell if it regained consciousness. Also, the story (apparently not supported by historical records) about Lavoisier's head blinking its eyes up to 20 times after being guillotined. Eeesh! I wonder what chatbots experience if they know they're being turned off.