What's so special about natural intelligence?

To be blunt, I think that a lot of the mainstream arguments for why human beings are not like digital computers miss the point.  I'm not one to pontificate about this since I am just an amateur, but it is only recently (in the last 6 months) that I have been exposed to some really solid arguments on this topic.   In this post, I want to single out three works that really get at the meat of the issue, which I'll talk about first.

All three of these books emphasize what I think is the crucial and criminally overlooked distinction between artificial and natural intelligence.  The essence of this distinction is that, in computers, the meaningfulness of different symbols is specified externally by the programmer.  In animals, the meaningfulness of a specific part of the environment is the outcome of the developed structure of  the animal itself.  The animal has developed within the environment in ways that make specific aspects of that environment meaningful.  This is a subtle point that I resisted at first.  Another way of putting it is to say that NOTHING (or very little) is meaningful to a computer, because the syntactic operations that occur therein have been specified by a human, and are therefore only meaningful to the human.  Content on computers may be part of meaningful loops, but those loops always include a human being.

I should mention that there is an exception to this.  Feedback loops within the computer that are monitered and are important to the computers functioning are certainly like parts of the human body.  However, these are not typically the reason why people equate computers with human, and the same types of feedback loops can be found in many devices, like thermostats.

As I said above, three books/articles touch on this issue (or related issues) in much better ways than I am capable of. They are:

One of Peter Cariani's articles on semiotic systems.  In this article he touches on the Peircean ideas of secondness and thirdness, stating that all syntactic operations on computers only attain thirdness when a human observer interprets them.  In other words, the computations solved by the computer are not meaningul to the computer, and that is the crucial fact that separates computers from people.

The Embodied Mind by Varela, Thompson and Rosch.

I will most likely make a posting devoted to this fantastic book.  This is one of the most carefully written, clear (once you get used to the subtle language used by the authors), and deep books I have ever read.  The amount of ground covered is astounding, and the authors give a good introduction to the incredible developments done in the field of cybernetics that preceded cognitive science, and foresaw many of its shortcomings by decades.

Bright Air, Brilliant Fire by Gerald Edelman.

This is a fantastic book that touches on biological findings and incorporates them into a view of cognitive science that does not ignore experience.  Edelman is a fantastic writer who explains the basics of neuroscience with a clarity and logical organization that is immensely satisfying to those with very little knowledge of biology.

There are many other sources for ideas similar to those that I have touched on here. This is by no means a thorough list.  Instead, my point is to show that there is a crucial difference between computers and people that is all too often ignored, or not spoken of clearly.