Update: Clearly, things have changed from when this was originally written.

So I’ve just been watching the ustream of Blekko by founder and CEO Rich Skrenta.Blekko has both an amazing idea and a product that still has a long way to go.

I love the idea of the slashtags to a degree – the ability to easily refine and drill down through data makes an amazing amount of sense, especially the date slashtag. What I see as an issue is adding some manner of intelligence to them. For example, the /foo tag is a great idea in theory – make a tag that searches particular sites – but a big risk is spam. After all, if one person makes the tag then what happens when they forget a site? What if /foo only hit techcrunch but not GigaOm? If that happens and Om makes a tag, suddenly the utility of the tag is diminished because you get “tag spread” where thousands of overlapping tags hit the same sites, with no clear way to distinguish between them. Tags have to be clearly valuable or they will not be used.

Now, this opens another can of worms when you ask the question “what about other users editing tags?” Spammers could add their sites to the tag; if I add an adsense scraper site to the tag and people do /foo/date and see my links at the top, I can get them to click and hit my site with little real value. You can argue that the site won’t make it into the index because it is low quality, but that is never a sure thing (not even with Google). In order to make it useful, a balance needs to be struck.

Now, these are theoretical questions that will only be answered over time, but I do see the immense potential of this system. The question is, how well can it be refined so that a truly usable product emerges? I look forward to finding out.