AI and Intelligent Design

by Neil Rickert

I’m a bit late adding my two cents to the blog debate between PZ Myers and Ray Kurzweil.  If you want to review the debate, then a good place to start would be with PZ’s August 21 post on “Kurzweil still doesn’t understand the brain“, and follow some of the links from that report.

I was reminded of the debate by a recent John Wilkins post.  And when I reread Kurzweil’s response to the first PZ post, I noticed how well that raises some of the design versus evolution themes that I have been raising in this blog.  This post will comment on some of what Kurzweil has posted.

For starters, I said that we would be able to reverse-engineer the brain sufficiently to understand its basic principles of operation within two decades, not one decade, as Myers reports.

I’ll go on record as doubting that the brain will be reverse engineered within 100 years.

I presented a number of arguments as to why the design of the brain is not as complex as some theorists have advocated.

There we see a key point.  Kurzweil is talking about the design of the brain.  He is looking at the brain as a designed thing rather than as an evolved thing.  We generally see “reverse engineering” as a way of retrieving the underlying design from an designed thing.  However, the brain is not a designed thing, it is an evolved thing.  And evolved things are very different from designed things.  If the brain is not a designed thing, then there is no underlying design to retrieve, and thus the planned reverse engineering is bound to fail.  Or, to put it differently, the brain was not engineered in the first place, so there is no engineering step that could be reversed.

A little later, Kurzweil says:

To summarize, my discussion of the genome was one of several arguments for the information content of the brain prior to learning and adaptation, not a proposed method for reverse-engineering.

And there is another illustration of the problem.  For the developing embryo interacts with its environment from the start, even before the brain begins to form.  Adaptation of the developing foetus is well underway before there is a brain.  So talk of “content of the brain prior to learning and adaptation” is seriously confused.

As best I can tell, Kurzweil seems to be thinking of the brain as a fixed designed thing that is engineered from a detailed blueprint in the genome.  And he takes this fixed designed thing to be an information processing system.  He sees adaptation as an information processing detail.

By way contrast, I see adaptation as distinctively biological.  If I look at a tree in my back yard, I see how it adaptively grows, and thus modifies its own shape, so as to better gain access to the available light.  And if I were to look at the roots, I expect that I would find the same kind of adaptive growth to better find nutrients in the soil.  This cannot be a matter of information processing in the brain, for the tree has no brain, no neurons.

The goal of reverse-engineering the brain is the same as for any other biological or nonbiological system – to understand its principles of operation.

Fair enough.  But Kurzweil assumes that those “principles of operation” are information processing principles.  It is far more likely that they are biological principles, not information technology principles, and that those biological principles are already being intensively studied within biology.  In terms of biological principles, my brain probably works in much the same way as Kurzweil’s brain.  In terms of information processing principles, the chances are that any two brains are very different.

%d bloggers like this: