Three Views on Conscoiusness
With the rise of generative AI we've all seen a lot of discussion about whether AIs could be conscious, now or in the future. While many of these discussions are productive I've seen others that look more like the proverbial youths arguing about the question "If a tree falls in the forest and no one is around to hear it, does it make a sound?" That's not a question you can just answer, but by realizing that "sound" can mean both an auditory experience and a sequence of vibrations in the air and so dissolve the question, satisfying our curiosity without answering it directly.
I don't see us having any hope at this moment of dissolving the much harder questions around artificial consciousness, but I do hope that recognizing that the term "consciousness" can mean different things in different philosophical and scientific systems can make discussion more productive. I see three major schools of thought there which I'll call reflective consciousness, qualitative consciousness, and temporal consciousness though others may slice things differently and this involves some lumping together of theories that aren't entirely compatible.
Reflective consciousness is perhaps the most famous philosophical position due to Descartes's famous phrase, cogito ergo sum or "I think therefore I am." Descartes himself distinguished between conscious awareness in general and the sort of self-consciousness pointed to by his famous phrase but it's clearly the later we're worried about when thinking about AIs potentially being conscious and what this means for them being morally significant. In modern science you see this approach in zoologists and child psychologists investigating things like the ability to recognize one's self in a mirror at its most basic level. You also have planning involving future mental states, e.g. "I'm not thirsty now but I will be later so I'll bring some water on this hike." And you have investigations of the ability to have beliefs about other people's beliefs like the Sally Test.
Qualitative consciousness is the property of having subjective experience that feel like things, that there's a certain "redness of red" that isn't just a matter of abstract information. C.I. Lewis used the word "qualia" for this in 1929 and this view is most often found associated with the philosopher David Chalmers who uses the phrase "phenomenological consciousness" for it. There are a large number of philosophical thought experiments associated with this school such as Mary the Color Scientist or Philosophical Zombies but I'm not aware of any scientific research programs that bear on it and some adherents contend that consciousness is outside the realm of scientific investigation by its nature.
Temporal consciousness is the view identifying consciousness most with the formation of memories. In modern times its most associated with a research project in reaction to behaviorists by trying to probe how people's behavior changes when they are aware of some stimuli versus not aware of it. The most famous experiments involve subliminal messages where some stimuli is flashed on a screen for some very short period of time. There is a threshold time where long enough flashes can affect a person's behavior for long periods of time including having them report what they saw, a conscious experience in this school. Flashes can still affect behavior but in fewer ways and only for about a second, a subliminal experience in this school. Recently experiments have expanded to running these experiments on people in FMRI machines to see which parts of the brain are involved in conscious versus subliminal experiences.
I'm not going to insist that anyone else divide things this way but in the future we'd all benefit from not arguing whether AIs are "conscious" but instead arguing about whether they have qualia; or whether transformers are more analogous to sensory memory or short term memory.
Comments
Post a Comment