Here's a story about a useful application of artificial intelligence to the problem of searching the internet.

Dr Brill's question-answering system does something similar. Many question-and-answer pairs exist on the web, in the form of “frequently asked questions” (FAQ) pages. Dr Brill trained his system using a million such pairs, to create a model that, given a question, can work out various structures that the answer could take. These structures are then used to generate search queries, and the matching documents found on the web are scanned for things that look like answers.

The current prototype provides appropriate answers about 40% of the time. Not brilliant, but not bad. And it should improve as the web grows. Rather than relying on a traditional “artificial intelligence” approach of parsing sentences and trying to work out what a question actually means, this quick-and-dirty method draws instead on the collective, ever-growing intelligence of the web itself.

Note that such a system does not attempt to be "strong" in the AI sense -- that is, it doesn't accomplish its task in a way we'd consider to be similar to how a human would do it, and there's no consideration that the question-answering engine may actually be "alive". However, I think such research will probably lead to many more useful results than will "traditional" AI.

Computers do this sort of "quick-and-dirty" analysis very well, but they "think" -- in the human sense -- very poorly; in contrast, humans think very well and do aggregate analysis very poorly. Rather than trying to make a computer into a human substitute, we may do better to focus on the strengths of computers and how they can compensate for human weakness.

Still, of course, creating a computer than can think like a human is a fascinating endeavor and worthy of exploration in its own right.

(HT: GeekPress.)



Email blogmasterofnoneATgmailDOTcom for text link and key word rates.

Site Info