Thoughts on Wolfram Alpha and Artificial Intelligence

I've been thinking about the announcement of the natural-language search engine Wolfram Alpha, which supposedly makes Google seem like a crude toy.

Humans have so many inherent limitations of perception. It is easily conceivable that within this century we will create a race of machines that are smarter (wiser?) than we are. (Sony is projecting 2011 for the year a computer can mimic human thought convincingly, the old Turing test.)

It all goes back to that famous saying of J.B.S. Haldane that the universe is not only stranger than we imagine, but stranger than we can imagine.

If we build thinking machines that have sensory capabilities far surpassing ours (like cell phones "understand" microwaves), they may eventually be able to see into other dimensions, discover things literally beyond our ken. Presumably they could translate those experiences for us, but the more advanced they get, the more the translation will be like formulating lessons for little children. For example, could anyone hope to explain the Theory of Relativity to a five-year-old?

The current ethical problems of artificial insemination, cloning, stem cells, etc. will seem like child's play when world leaders have to decide whether to trust a computer that says "You can't possibly understand, but I know I'm right."