Google’s blind-spot: wild ideas

Google’s badly named Google Image Labeler made us scratch our head this weekend. The clunky title is one more reminder that Google remains vulnerable despite its invincible image — something that start-up expert Paul Graham also points out in a recent interview.

One advantage that big companies have is leverage. Google, with its many millions of users, is now attempting an end-run around other Web 2.0 photo search sites, by launching a game (Image Labeler) that might end up getting a good portion of Google photos “tagged.” In this game, Google pairs you with another random user, and asks you both to label a photo it shows you both. If you both select the same word to label it, it is a good sign that the label is accurate. You get points, and you move on to the next round. Now, Google isn’t using the conventional word “tag” for this project, perhaps a sign that it fears being sued for using the term (popularized by competing photo site, Flickr, now owned by Yahoo).

But what a name. Google has had a problem building an interactive “community” around its features, and this name reflects that weakness. It may come as no surprise, then, that the game is licensed from outside of Google — specifically, from Luis von Ahn, the inventor of the “ESP Game,” which the Google game is modeled after. Ahn says the game could tag all of Google’s photos within a couple of months. So this game may work.

So where are Google’s blind-spots, if it can simply license technology from outside to do the sorts of things it can’t do internally?

Paul Graham, who leads Y Combinator, a incubator that invests in start-ups, had a noteworthy answer in an interview with Techcrunch a couple of days ago.

He said one of his companies, calendar start-up Kiko, failed in going up against Google, because Google Calendar is the kind of application that Google hackers use at work. Another application that they use is GMail. But other, social networking companies, such as Orkut aren’t used by Google’s internal folks, and so they tend to not to do well.

So a startup could compete with Google if they had an idea so wild that it would freak out the internal [Google] gatekeepers, no matter what area it was in.

And the more wild and freaky, the better, which is why Graham says he is funding an “insane” idea to be pursued by the same Kiko team.