Despite his prescience, one aspect of Gelernter’s vision of the future has been slower to emerge.  Mirror Worlds envisions a web populated by “intelligent software agents,” essentially algorithmic applications which act and react, monitoring the inhabitants and users of the Internet, and feeding information back to their controllers.

I found myself thinking a lot about Gelernter and his intelligent agents while attending Variety’s first Entertainment Apps Conference.  As I listened to the array of speakers evangelizing on the power of these little software packages to change the way we view content, I began to realize that yet another prediction from Mirror Worlds was coming true.  But what does this new universe of apps mean for big media, and how will these intelligent agents alter the way we interact with content platforms?

The concept driving intelligent app production these days is the relationship between the home television, the viewer, and the “second screen.”  A Yahoo study last year found that 86% of people watching TV are simultaneously engaging with their mobile devices, a fact that has led TV advertisers to seek a way to leverage this second screen for a more immersive experience.  People in the business used to make a joke about interactive television, that it is the future, and always will be.  But with second-screen interactivity, that future is fast becoming a reality.

Silicon Valley has taken notice.  Venture capital firms are increasingly investing in operations such as GetGlue and SnappyTV.  In December of 2011, Google Ventures, Hearst Interactive, and Vinod Khosla announced a significant investment in Miso, a second screen pioneer with a quarter million subscribers, which provides a social media platform to people watching specific TV programs, such as Dexter on Showtime.

The issue of how intelligent agents factor into this new equation is a bit more complicated.  The second screen experience is attractive to users on many levels, allowing them to learn more
about their favorite programs, go behind the scenes, predict outcomes, and/or provide commentary with a community of fellow fans.  But what about advertisers?  After all, television networks exist to serve their audience, but even more, the brands hawk their products on the network’s platform.

The panelists at the Variety conference envisioned superior messaging potential by continuing
stories across platforms (TV to a mobile device) and delivering digital ads synchronized to real-time TV spots.  Recently, USA tested a synchronized Ford spot during an episode of Necessary Roughness and found that 23% of the viewers interacted with the advertisement when the ad ran simultaneously on TV, USA’s website, Facebook, and an iPad companion app.  Parks Associates in late November 2011 published a study asserting that 45% of US broadband households are comfortable seeing ads based on their TV-viewing habits and product/service preferences.

Sounds great, right?  But what about the pesky 55% who aren’t comfortable?  Let’s take a closer look though at the underside of what this all means, a through the glass darkly perspective on
Gelernter’s mirror world future.  The advertisers and networks pushing the envelope of second-screen interactivity and synchronized viewing are unlikely to stop at capturing your eyes and ears.  They want your data too, and here’s where it gets tricky.

Consider the 23% of viewers who interacted with the Ford advertisement.  Even as they receive information about the MKX car appearing in the show, the algorithms embedded within the programs will collect information about the viewers.  This is already happening in the online world every moment of every day.  Any time a consumer writes a comment, purchases
through a website, makes an eBay bid, plays a game, or takes almost any other action, that information is gathered and sold.  BlueKai, a supplier of online data, has “actionable” audience data on over 80% of the entire US Internet population and markets the information in over 75 million micro-transactions a day.  Lotame, another company in the space, goes further, breaking down 240 billion monthly collected interests, actions, and attributes into segments such as “Reach & Frequency – How often and when people express interests and/or actions,” and “Sentiment & Exposure – What people say, what they read, and when and how they say and read.”  So when you make a comment, telling a friend you’re pregnant or in the market for a car, an intelligent agent is recording your view, storing it in the cloud, and feeding it back to the advertiser.  So in the future, if not already, expect loads of ads for Pampers and CarMax to arrive after your “private” conversation ends.

In an Orwellian sense, as you watch the view screen, the view screen will be watching you.  In fact, every computer, mobile device, and set-top TV box has a unique fingerprint, a sequence of preference settings, and other attributes that make it distinct from every other device.  This will allow the advertisers to data mine their audience and virtually track their behavior.  As of today, there are no practical ways for users to block this function.  While Gelernter’s intelligent agents might not have fully arrived today, with the prospect of a new generation of apps for
the second screen, the interactive two-way mirror seems only a moment away.