Robot Consciousness (A book review of How to Create a Mind)

Robot Consciousness (A book review of How to Create a Mind)

I’ve just finished the absolutely fantastic How to Create a Mind by Ray Kurzweil.  It is masterfully broken in to three basic parts.  First is a discussion of how the mind operates from a logical perspective.  The second is a mapping of that on to the physical constructs of the brain.  Finally, there is a discussion about what artificial intelligence means, the dangers of it, and some really interesting thought experiments on what it means to be human if machines can think like we do.  I highly recommend the book to anyone who’s curious about such things.

Rather than my typical book review though, I wanted to spend time on a thought it sparked with me.  Kurzweil talks at length towards the end of the book about consciousness and free will and the extent to which Artificially Intelligent beings would have them compared to us.  He points out that eventually (he estimates 2029) we will create a robot that is essentially indistinguishable from humans… it will learn like one, process information like one, and make decisions like one.  He then walks us through a set of thought experiments intended to show that there is no difference between humans and this new being.

I would like to propose a difference… One of the things that makes human free will interesting is that we don’t yet know if it exists.  We are unsure that if you could complete a brain scan of a human in one instant AND understand all of the sensory inputs that brain has available to it if what the brain would do next could be calculated.  There are operations of the brain that we cannot yet explain and those are where the argument for a “soul” still exists.  With an artificially created being, we would have a log file; a record of everything the brain has seen and how it impacted its “personality.”  We would be able to go back after a decision was made and see EXACTLY why.  We could go through the logs and say something like, “6 things happened to Robot A that registered as aggravating and that is why he snapped at James.”  There’s no room for “maybe that’s just Robot A’s personality/soul”, we could derive his actions entirely based on the logs and the formulas we know are in his brain.  THAT is what makes Robot A not “conscious” and without “free will”.

God help us if we’re ever able to mine our own brains for a similar log file.