A lot of buzz is beginning to circulate about Powerset, purported to be a natural language search engine that will draw upon AI technologies to make the searches more intuitive. I doubt this is going to be all that big, even if PARC is who's behind this particular search engine. The success of search engines like Google is not just based on the accuracy of its results but on the speed as well. having to type entire sentences in order to meet the natural language criteria will only add delays to the search process. Plus, how long will it actually take to refine the system to provide accurate results?
Powerset seems more a toy rather than a practical search engine that might challenge the big boys like Google and Yahoo. It'll probably help develop leaps and bounds for AI research since essentially they need to master how to teach language to a computer system. However everyone else will still stick to what they know.
We're already conditioned to keywords when it comes to searching online. The search engine technology behind Google has capitalized on this by creating a system that supports "keywordese" and improves accuracy as more queries are submitted. While it's not a smart search enging like Powerset wants to become, it's an effcient one that meets the current behaviors of the millions of people online right now. Sure, we could learn to use Powerset in time, but if the process becomes longer rather than shorter or if the results don't seem all that different, why will people venture off into something new?
I guess at the end of the day, we still need to wait and see how this thing will actually perform. Perhaps we're all getting our hopes up all for nothing.
Related Links:
No comments:
Post a Comment