The bottom limit of consciousness
Here and at PF there have been quite a few discussions about consciousness in computers. THIS ARTICLE takes up the question of whether honey bees, flies, and similar insects can 'feel' -- have some degree of consciousness.
They make their case this way:
? Other scientists have argued that a part of the human brain called the midbrain can, on its own, give a person lacking more advanced parts of the brain simple awareness.
? The insect brain does something similar to the midbrain in absorbing information from the environment, from memory and from the body to organize its activity.
? If the insect brain does the same job as the vertebrate midbrain, then the insect has the capacity for awareness.
If this line of reasoning is correct, Dr. Barron and Dr. Klein say, a robot built with artificial intelligence that could integrate sensory data, memory and body awareness would have the capacity for the minimal level of consciousness they describe.
I was surprised. They presented a good argument for bees and flies having some (a smidgeon, not a lot) consciousness. C. elegans, with its 300 neurons, probably can't. But a bee has nearly a million neurons, and the bee-brain circuitry, they said, is more complex than ours.
They make their case this way:
? Other scientists have argued that a part of the human brain called the midbrain can, on its own, give a person lacking more advanced parts of the brain simple awareness.
? The insect brain does something similar to the midbrain in absorbing information from the environment, from memory and from the body to organize its activity.
? If the insect brain does the same job as the vertebrate midbrain, then the insect has the capacity for awareness.
If this line of reasoning is correct, Dr. Barron and Dr. Klein say, a robot built with artificial intelligence that could integrate sensory data, memory and body awareness would have the capacity for the minimal level of consciousness they describe.
I was surprised. They presented a good argument for bees and flies having some (a smidgeon, not a lot) consciousness. C. elegans, with its 300 neurons, probably can't. But a bee has nearly a million neurons, and the bee-brain circuitry, they said, is more complex than ours.
Comments (12)
It could also be that there's a minimum number of neurons you need for consciousness to become manifest.
I think this is the logical error we are making here with consciousness, because for me it's quite obvious that a smart animal can have some levels of consciousness while humans are more conscious. And likely the reason is the ability to learn things. It's not a question of either being conscious or not. Are your either smart or an idiot? Not much use to put humans / animals / whatever to groups either a) smart or b) idiots.
Regarding consciousness, we do have some definitional problems, so we'd have to explain what we mean. If I swat a bee and it lies there stunned on the ground not moving but then it gets up and flies away, I think it'd be accurate to say that it lost consciousness momentarily but then regained it. It was less than dead, but not quite the same bee we all previously knew and loved when it was lying there, so I think it's correct to describe it in terms of its consciousness.
The sort of consciousness I find interesting which causes serious current limitations in AI is the idea of conceptual awareness. It's clear that computers don't know "about" things or conceptualize them in their CPUs. I'd say the same of bees, I guess, although I've never been a bee, so it's hard to know.
Scout bees fly around looking for flowers. When they come across flowers they fly back to the hive and transmit the distance, direction, and maybe something about the kind-of-flower they found. In order to do that they have to at least sustain a purpose on their own, and once fulfilled, remember the location, distance, direction, and type long enough to get back to the hive and report. That requires a comparatively small amount of consciousness--but it is vastly more than a hunk of fractured concrete has.
Whether the latest chip from Intel has enough complexity to register to itself that it is up and running, I don't know. Does an Intel chip "feel" a surge or lag in the power supply? Don't know. Bees had a long time to develop whatever consciousness it has; An Intel chip has been through maybe a decade or two of development; there are always more transistors per chip and faster operation speeds. I don't think Intel is spending a lot of money on trying to make it's chips conscious of being a part of a computer doing nothing better than visiting philosophy sites.
Some animals apparently have more complex circuitry than human brains possess. Bees aren't big enough to carry a big brain, so they developed denser, more complex circuitry than we have. I'm totally out on a limb here, but my guess is that smart birds like parrots and crows maybe have more complex circuitry than we do as well, since their brains are quite small -- especially for what some of them are able to do. We are big enough to afford the Intel approach -- just keep adding more neurons to keep up with demand.
They do exactly what you tell them to.
HAL 9000: "Dr. Chandra, will I dream?"
Dr. Chandra: "I don't know HAL."
I think the binary system is something very fundamental to us. And perhaps other species too. With it we have made much sense to the World.
But then things that cannot be put into the binary mold are problematic to us. Should be no wonder.