Predictive Search: There’s a downside with the upside
July 30, 2013 § 1 Comment
Back in March, Fast Company’s Co.Create staff writer Joe Berkowitz published a short, humorous piece about the suggestions displayed by Google Search’s auto-finish based on other people’s searches. He titled the article “The Accidental Poetry of Google’s Predictive Search,” and opened it with, “Half the fun of finding what you’re looking for online is finding what you’re not looking for.”
When Google updated search to include “sentence-finishing capacity,” a user was given the gift of an unanticipated selection of potential searches to complete. Some of these are funny, in that they are in no way what you were looking for when you set out on your search, and there is no way on earth you would ever have thought to. For example:
As Joe Berkowitz points out, “A new thread called Google Poems opened up on Reddit recently, showing off the accidental poetry of four different variations on a theme.” It still has legs.
But what if, instead of opening up your world to unforeseen possibilities, all of this auto-suggesting is actually limiting it?
There is a piece in today’s New York Times, “Apps That Know What You Want, Before You Do,” that details the next phase of “predictive search — new tools that act as robotic personal assistants, anticipating what you need before you ask for it.”
These applications, from the likes of Google (Google Now), Evernote, reQall and Mindmeld, to name a few, pull data from your smartphone, calendar, email, social network activity, search history and sites visited, and merge it with “public” data, such as the weather forecast and movie showtimes, to make suggestions as to what you should do, where you should go, and when.
Glance at your phone in the morning, for instance, and see an alert that you need to leave early for your next meeting because of traffic, even though you never told your phone you had a meeting, or where it was.
The intent is essentially a good one: take the massive amount of information on the Web and present it to you in a manageable way that helps you make decisions and better manage your life, day-to-day. Journalist Claire Cain Miller digs into the current shortcomings of the technology, for example, suggesting movie times for a movie whose trailer you viewed but did not like. But these glitches are of minor concern; challenges in functionality will be overcome in time. What I find disconcerting (or “creepy”) is this:
The technology is the latest development of Web search, and one of the first that is tailored to mobile devices. It does not even require people to enter a search query. Your context — location, time of day and digital activity — is the query, say the engineers that build these services.
The companies providing these new tools make some sweeping assumptions: that you want to be defined as “your context;” that you want to be presented only results that are determined to be of interest to you by some bit of code that bases your potential options on your previous online activity and the activity of those you have never met; that you want all of that activity tracked to make your life easier.
What if “your context” has changed? What if you’re not the same person that you were yesterday and want new opptortunties? At what point does “helpful” cross the line? And, when it does, will you have the opportunity to opt out?
Something to think about next time you Google “why is…“