If you can’t join them: beat them.

Or at least scare them to death. Look at Google’s funny presentation of an interactive TV research paper. The gist: in a perfect world, iTV would be an ubiquos reality, any marketeer’s dream and the perfect pasttime for many a people who’s idea of entertainment isn’t watching a PC scanning itself for viruses. But as we all know, the world is far from perfect (Global Warming, Cialis spam, Cherry Coke – just to mention a few). So since one and a half decades and for the next foreseeable time, iTV can consistently be described as the future of tv (and future it stays).

The reasons are plenty: a hodgepodge of industry interests and players is frequently churning out standards which are practically nowhere deployed (MHP). The computing power of the targeted hardware (A.K.A. settop boxes) is dwarfed by any 10 Dollar wristwatch from a street vendor in Paraguay. And most walled garden of the operators aren’t landscaped after the hanging gardens of Semiramis, but obviuosly more likely with the desert Gobi in mind.

Yes, it’s a mess. But who’s better equipped to fix up a mess than Google? After all, they brought (some) order to the web (which is after all the mother of all messes). Introducing Google TV. Or at least a paper called Social- and Interactive-Television Applications Based on Real-Time Ambient-Audio Identification.

In a nutshell, the idea goes like that. Many people watch tv and surf the web at the same time. Combining those experiences in a somehow converged appliance is still a pipe dream. Connecting PC and STB is mostly not an option. Making people doubleclick to switch a program is the not-so-elegant status quo: first change the channel on the remote control, than switch web sites.

Google’s approach is somehow between big brotherish and ingenious. The only thing you need is a PC with a microphone and a tiny piece of software. The mike listens into the room, the app samples and irreversibly compresses the viewer’s ambient audio to summary statistics. These statistics are streamed from the viewer’s personal computer to the audio-database server for identification of the background audio (e.g., ‘Seinfeld’ episode 6101, minute 3:03). Meaning: transmitted won’t be your stupid talk about nuking your next door neighbor, but just the audio fingerprint of the background noise formerly known as television (and if 10 minutes after you talked about nuking your neighbor 20 square jawed hulks with blackened faces politely knock on your door, you’ll know that somebody has put a trojan onto your system).

Far fetched? Audio fingerprinting is already here. UK-based Shazam recognizes music you play to it, same goes for Fraunhofer Institut’s AudioID. This solution is quite elegant. You get rid of most of the existing complexities, don’t have to care about channel numbers, locations, distribution networks or else. You’re back to the program itself (which even could be a DVD you watch and Google TV will tell you that there’s a guy watching the same movie over and over just like you and you can start to talk, chat or whatever about your mutual obsession).

The paper describes four different types of applications: Personalized Information Layers, Ad-hoc Peer Communities, Real-Time Popularity Ratings, and Video “Bookmarks”. The impact on advertising is partially addressed, too. Well, networks, be aware. Will Google harvest the info clouds surrounding any piece of programming and at the same deliver targeted ads, based on context and consumer behavior (and thereby taking away your bread and butter)? Most likely, not. Even if Don’t piggyback isn’t totally covered by the famous Don’t be evil. But wait for those kind of apps and ads to show up combined with IPTV offerings. Additional lines of contextual content, combined with AdSense for IPTV. To be watched either picture in picture or on a second screen (PC, handset, …).

More to read: Official Google Research Blog: Interactive TV: Conference and Best Paper,
and Chris Riley: Clues on Google TV?

Addendum: The more I think about it, the more obvious it is. The mad professors at the Googleplex are good for many far fetched ideas. But let’s put it like that: the audio part is such an obvious privacy nightmare scenario, that they can’t be serious about it. No way. So we have to divide the paper in two parts. The relevant part for Google is the iTV-backend, with the app and ad serving.
But would have anybody in the whole world noticed just another iTV paper without the harebrained audio scheme attached? Right. And here we go. One paper. And Google positions itself as an interesting player in the TV/iTV market.