Chapter 5.                         (continued) 

Many empiricist philosophers see AI as our best hope for defining, once and for all, what thinking is. AI offers a model of our own thinking that explains it in ways that can be tested. A program written to simulate thinking either runs or fails, and every line in it can be examined. When we can write programs that make computers converse with us so well that, when we talk to them, we can’t tell whether we’re talking to a human or a computer, we will have encoded what thinking is. Modeled it so that programmers can explain it with algorithms. Run the program and observe what it does. Repeat.

 

With the rise of AI, cognitive scientists felt that they had a real chance of finding a model of thinking that worked, one beyond the challenges of the critics with their counterexamples.⁵

 

Neurology, behavioral science, or AI – these are currently the best paths along which we might get to a purely empiricist explanation of what thinking is. They begin from evidence that all can witness. That fits Empiricism.

 

Testability in physical reality and replicability of the tests, I repeat, are the characteristics of modern Empiricism (and of all Science). All else, to modern empiricists, has as much reality and as much reliability to it as creatures in a fantasy novel. Ents. Orcs. Sandworms. Amusing daydreams, nothing more.

                                                                     

 

                                  

                                           

                                      Kurt Gödel (credit: Wikimedia Commons)

 

For years, the most optimistic of the empiricists looked to AI for models of thinking that would work in the real world. Their position has been cut down in several ways since those early days. What exploded it for many was the proof found by Kurt Gödel, Einstein’s companion during his lunch hour walks at Princeton. Gödel showed that no rigorous system of symbols for expressing human thinking can be a complete system. Thus, no system of computer coding can ever be made so that it can adequately refer to itself. (In Gödel’s proof, the ideas analyzed were basic axioms in Arithmetic.) Gödel’s proof is difficult for laypersons to follow, but non-mathematicians don’t need to be able to do formal logic in order to grasp what his proof implies about everyday thinking.⁶

 

 

 

                                        Douglas Hofstadter (credit: Wikipedia)

 

If we take what it says about Arithmetic and extend that finding to all kinds of thinking, Gödel’s proof says no symbol system for expressing our thoughts will ever be powerful enough to enable us to express all the thoughts about thoughts that human minds can dream up. In principle, there can’t be such a system. In short, what a human programmer does as she fixes flaws in her programs is not programmable.     

 

What Gödel’s proof implies is that no way of modelling the human mind will ever adequately explain what it does. Not in English, Logic, French, Russian, Chinese, Java, C++, or Martian. We will always be able to generate thoughts, questions, and statements that we can’t express in any one symbol system. If we find a system that can be used to express some of our ideas really well, we discover that no matter how well the system is designed, no matter how large or subtle it is, we have other thoughts that we can’t express in that system at all. Yet we must make statements that at least attempt to communicate all our ideas. Science is social. It has to be shared in order to advance.

 

Other theorems in Computer Science offer support for Gödel’s theorem. For example, in the early days of the development of computers, programmers were frequently creating programs with “loops” in them. After a program had been written, when it was run, it would sometimes become stuck in a subroutine that would repeat a sequence of steps from, say, line 79 to line 511 then back to line 79, again and again. Whenever a program contained this kind of flaw, a human being had to stop the computer, go over the program, find why the loop was occurring, then either rewrite the loop or write around it. The work was frustrating and time consuming.

 

Soon, a few programmers got the idea of writing a kind of meta-program they hoped would act as a check. It would scan other programs, find their loops, and fix them, or at least point them out to programmers so they could fix them. The programmers knew that writing a check program would be hard, but once it was written, it would save many people a great deal of time.

 

However, progress on the writing of this check program met with problem after problem. Eventually, Turing published a proof showing that writing a check program isn’t possible. A foolproof algorithm for finding loops in other algorithms is, in principle, impossible.⁷ This finding in Computer Science, the science many see as our bridge between the abstractness of thinking and the concreteness of material reality, is Gödel all over again. It confirms our deepest feelings about Empiricism. Empiricism is useful, but it is doomed to remain incomplete. It can’t explain itself.

 

Arguments and counterarguments on this topic are fascinating, but for our purposes in trying to find a base for a philosophical system and a moral code, the conclusion is much simpler. The more we study both theoretical models and real-world evidence, including evidence from Science itself, the more we are driven to conclude that the empiricist way of understanding what thinking is will probably never explain its own method of reaching that understanding. Empiricism’s own methods have ruled out the possibility of it being a base for epistemology. (Define the word meaning?) (In Algebra, solve x² + 1 = 0).