Google Search Moving Web 2.0
Michael Arrington previews what purports to be a prototype of a new Google search experience, incorporating user commenting and voting such as we see on Digg and other social media sites:
The video above shows a user interface being bucket tested by Google to select (probably randomly determined) users. Earlier today we showed a screen shot of the interface and a video of the search history, recorded by Adrian Pike, the CTO of startup Tatango. This new video, however (also recorded by Pike), shows the full Google search experience with a very Digg-like interface. Users vote search results up or down – a down vote makes it dissapear with a “poof,” an up vote moves the result to the first page.
I’m not at all sure this is a good idea. What we’ve seen repeatedly at Digg-type sites is that a handful of users — fewer than 100 early adopter power users — will band together and dominate the voting. Indeed, as few as 10 users can totally skew the results if they know what they’re doing. Surely, we don’t want the same thing to happen to search?
Google’s electronic algorithm is too subject to gaming and the splogs are showing up far too much in the results. But it’s basically an excellent search engine that needs constant tweaking to combat these moves. Moving in a totally different direction, especially one proven to be even more easily gamed, would be a major blunder.
In general, yes, not useful. But what if you could link up with friends that you trusted and have your search results only be affected by their actions?
Lijit, whose search engine I’m trying out (top right of site) is attempting to do precisely that. It’s an interesting concept although incredibly difficult to implement.