I just heard Jimmy Wales, founder of Wikipedia, speaking at Stanford Law School today. Wales is working on some new projects that he hopes will harness the community-driven collaboration of Wikipedia. He’s already had some success in branching out from the encyclopedia idea with Wikia, which is a “wiki farm” compiling information on a variety of different subjects (some of the most successful so far relate to video games).
What Wales spoke about today, however, is a new collaborative search project. The concept is still in its early stages, it seems, but the idea would be to harness the intelligence and dedication of human beings to produce search results significantly better than Google’s. This raises a few questions:
Is Google broken? It’s amazing what Google pulls up, but maybe we’ve all gotten so good at working with an imperfect system that we just tune out the spam and misinterpretations that still crop up.
Is a collaborative social model the appropriate solution to this problem? People are good at compiling encyclopedias, but they may not be good at emulating search rank algorithms. Also, Google is powered by millions of servers in dozens of data centers over the world managing petabytes of information. In other words, this may be a technology+money business, not a people+transparency business.
These issues aside, Wikipedia is one of the most amazing things to come out of the whole Internet experiment, so I’m excited to see what Wales comes up with. Has search become a basic service? Would it work better as an open-source system?
2 comments:
Not sure I'd say Google is broken but there's definitely room for improvements. Searching for topics like travel or flowers are horrible - there's so much advertising for hotels or sending boquets that you can't get much value (unless you're looking to make a purchase). It will be interesting to watch Mahalo since they're working on providing hand crafted search results. It's ambitious but it might just work for some of the common terms that don't work well with algorithmic results.
That is definitely the target area Wales is aiming for (down to the example of travel, which he used in his talk). I'm suspicious about a human-driven search engine's ability to keep up with the massive amounts of new data that come down the pipe, though. Another model that Wales mentioned was About.com--it doesn't cover everything, but what it does address comes with a little more depth and perspective than your average Google search results page.
Post a Comment