Isn’t it about time we did content discovery differently?

Is anyone else a little tired of being pushed towards certain content because of mined personal information? It’s kind of like having that friend who resembles you a little too closely; eventually they become tiresome. Plus, there’s a fine line when it comes to personalized recommendations, which sometime WebWorkerDaily contributor Dave Clarke spelled out on Twitter earlier this morning:

That’s why I was happy to see new iOS (s aapl) app Video Time Machine ($0.99) take a different approach to mobile content discovery. As its name suggests, Video Time Machine allows a user to take a trip back in time, to check out audio and video clips organized by year, going all the way back to 1860, with the first sound ever recorded. It’s an app-based client for the content on YTTM.TV, an archive of videos from the past that surfaces content based on the year and content type (commercials, sports, etc.) chosen by the user.

Video Time Machine basically just pulls content from the website, and puts it in a very touch-friendly wrapper, but that’s all it took to make watching movies on my iPhone (and iPad) feel like a novel experience. Part of that is because the content I’m checking out is very rarely something I’d come across on Twitter, Facebook, or through any recommendation engines that plug into my regular daily consumption habits. I don’t want to “discover” things I already like, you see; I want to check out things I’d never come across on my own.

In a world where finding content consists mainly of services or people saying “you like this and this, so you should also like this” or “your friend likes this and you like your friend, so you should also like this,” trying out an app where the locus for what content gets surfaced exists outside of the individual doing the searching feels like a revolution. It reintroduced one key element that few recommendation engines can claim to achieve: the ability to surprise oneself. Good surprises might require a bit of input from the user along general lines, but as highly personal recommendations narrow the scope of what they’ll offer you, they also exclude a much larger cross-section of content. Striking a balance using input from the user, and criteria that are pre-selected through other means would make for more serendipitous finds.

I don’t think the appeal of Video Time Machine only works because of the type of content it offers, either. I’d love to see an app discovery tool that uses some other similar contextual criteria to group titles together, too. App releases organized by date would be cool, or by most frequently updated, or even by more irreverent coincidental similarities, like all apps that use a bird in their icon. It might not be as good at pigeonholing the average customer, but it would offer an alternative approach for users, who, like me, are tired of having their social graph mined for the purposes of being fed the same old content over and over again.