Artificially Intelligent Search: A Look at Recent Developments

Posted by Theron Lalla, Associate Account Manager

It almost sounds like a twist on an old meme—“In Soviet Russia, web browses you!” But it’s not a joke, and while it may not be Soviet Russia, the web has been browsing you—your habits, preferences, likes, dislikes, filters and so on.  It feels Orwellian in nature, and it’s quickly becoming the next big trend in the search engine land.  Yet it doesn’t stem from evil plots to control the world (that I know of)—it stems from a growing recognition that too much information has quickly replaced the problem of not enough information, and every search engine is racing to get it “just right.” So let’s take a look at some of the recent developments in this emerging tech trend.

On September 8th, Google launched Google Instant, the first incarnation of a predictive results page.  It was intended to “transform” search as we know it.  Yet a preliminary report suggests that, for the most part, search behavior hasn’t changed.  This may be due more to user behavior rather than the technology itself—most people cling to search habits until forced to change.

A month later, on October 13th, Bing announced a partnership with Facebook with the new feature unofficially dubbed “social search.”  In a recent post, Performics explored a few of the key implications, both positive and negative.  In short, it taps the power of recommendations by using Facebook “Likes,” but brand alignment will be hard to control (a die-hard Apple fan will likely ignore any “Liked” brands or products from a PC user, thus contaminating the message for the original user).

Then, point-counterpoint-point.  Four days after Bing’s announcement, Google announces plans for “serendipitous search” technology—where artificial intelligence mixed with previous search behavior now suggests results for you—before you even type them in.  It finishes what Google Instant started, barely a month before.  This is the real definition of predictive search, generating recommendations based on your past behavior and current location (among other factors).  What has done for years with books, Google will be doing for… everything else, albeit with a slightly more complex algorithm.

Out of left field comes, taking a manual rather than automatic approach to filtering results: merging what they call “slashtags” (filters embedded in your search query) with an open, anonymous community of taggers.  Only a few days in the ring, however, and there have already been some usability issues reported.  If they handle the UX, however, the new engine will be poised to be the first to merge the social elements of or Wikipedia with regular search engine results.

In the upcoming year, we will likely be witness to even further development in the realm of artificially intelligent search.  But is this a good thing or a bad thing?  Privacy concerns have accompanied every tech development that steps in this direction, especially in regards to government access to user information.  As search grows and user information becomes increasingly public—especially with the other big trend, Cloud-based storage, user protection will become a political hot-button, calling for the establishment of web-based civil rights.  And just as we once declared a separation of church and state, eventually there may be a need for separation of search and state—because artificial intelligence on it’s own is amazing, but when guided by political agenda, is absolutely terrifying.

Bookmark and Share

Comments are closed.

Performics Newsletter