Turing Machines and Brains…

The Church-Turing Thesis states that any function that is computable can be computed on a Turing machine. Basically, this boils down to computers being able to solve any algorithmic function. Supporters of so-called ‘strong’ AI claim that a robot could be built that is ‘sentient’ or self-aware. This claim really says something not only about robots and computers, but also about the philosophy of mind. For those that support strong AI this boils down to saying that consciousness resides in the brain, which in turn is something of the nature of a universal Turing machine (computer).

Without really disputing the claims of strong AI, I wonder sometimes if man’s thought ‘processes’ can really be adequately described as algorithmic. I suppose that if one wanted to postulate, the non-algorithmic seeming of human thought could boil down to a hidden underlying algorithm or system of algorithms. The only problem I have with this line of thought are those mathematical problems which cannot be solved algorithmically. [see the Busy Beaver problem] There are solutions, however. These are also discoverable. If they are discoverable non-algorithmically, then doesn’t this say something of thought?

I wonder also, since language is at least philosophically linked to reason, about word-forming. Sometimes in conversation I tend to pause, searching for the correct structure or word that best typifies my meaning. But I don’t think of it as an algorithmic process of word-selection, although upon reflection it almost seems like different words are ‘tried’ and discarded. Language, to me, seems to be a more slippery thing than computation. However, I guess that one could simply say they differ only in degree of complexity.

I guess that raises a final question: when does complexity increase itself constitute a different kind of thing?