Wednesday, March 06, 2019
REMoleth
Article considers the possibility of a Turing test for consciousness.
It's a surpassingly hard question, intrinsically impossible to answer objectively.
None of the paths proposed in the article are likely to get there. I think the best starting point is dreams.
We're reasonably sure that mammals and birds have dreams like ours. Cuttlefish also show the same external indicators of dreaming.
Why is dreaming a good leth? Because there's NO POINT in dreaming unless the dream is happening within consciousness. A dream is an internal play on an internal stage. If nobody is watching the show, there's no purpose in running the immensely complex mechanism of scripting and narrating the show.
Would this apply to AI? I don't know, but it's a precise objective measure for living animals and plants. If a creature has dreams, we know FOR SURE that the creature is conscious. The converse is less definite. There may be conscious creatures who never dream.
Later, continued the theme and expanded the question.
Why is dreaming a good leth? Because there's NO POINT in dreaming unless the dream is happening within consciousness. A dream is an internal play on an internal stage. If nobody is watching the show, there's no purpose in running the immensely complex mechanism of scripting and narrating the show.
Would this apply to AI? I don't know, but it's a precise objective measure for living animals and plants. If a creature has dreams, we know FOR SURE that the creature is conscious. The converse is less definite. There may be conscious creatures who never dream.
Later, continued the theme and expanded the question.Labels: Answered better than asked, Leth, Metrology
¶ 8:17 AM
