I’ve been thinking about search lately. Mainly the thoughts have centered on how poorly machine developed search actually works. Think about it, when you search, be it on Google, MS Live Search or anything, how relevant are the results? How many of the top results are simply gamed results from SEO tactics?
The problem is this: while an algorithm can do a good job, I think we really need a team approach – a Digg-like rating system that allows us to put some real human value into the mix. Of course those of you familiar with Digg will immediately say: “right, then the SEOs will just game that as well”, and you’d probably be right.
Think about this though: why not place weight on the ratings offered? Surely the opinion of an expert in a field is of a higher value than you’re Aunt Edna (unless she is that expert…)?
So it seems like it might be impossible…take a quick go through the now infamous Indentity 2.0 presentation Dick Hardt of Sxip did as OScon last year. If we had a ubiquitous means of identifying people on the web, then we’d be able to apply some value to their ratings. And we might even be able to do that one the fly, so that, say I might value the rating of my friend David more than my nephew who’s kind of weird…
My suspicion is that the next step for search is that we’re going to need to re-interject a human side to it. That’s why the web 2.0 stuff like Digg, del.icio.us etc. has become so popular – the human side.
We’ve seen the machine side, and frankly, it can only go so far. Bring on the meat sacks!