After the recent scandal regarding NSA PRISM program alternative search engines such as DuckDuckGo.com got a significant increase of attention and an inflow of new users (see Suren’s post for interesting statistics). Central issue of the big hype is the question of privacy on the Internet.
While I believe that this is an important and complex question which should be addressed by our society, I am not very concerned about the topic personally.
First, as Vladislav mentions in his post, the question of which search engine to use is both irrelevant from privacy standpoint (as there are so many ways your privacy could be compromized on the internet) and also reasonably controlled by using incognito mode and surfing from an internet cafe.
Second, I think that the issue of information bubble is a more evident problem which has visible consequences on the global scale (check out this mindful TED talk on the topic). Getting unfiltered results is possible through incognito mode / special browser settings or even using the “big switch to turn on/off the personalised search results” which actually exists on the top right corner of search results page.
But to me the key question is not the availability of untailored information by request, but rather the impact on population at large as the majority consumes filtered information by default!
And here comes the first example of a behavior-shaping technology. What if Google/Facebook/Twitter create an algorhythm that recognizes the “intent” of your search and differentiates between:
- searching for something specific and concrete (majority of search requests)
- searching on topics that have multiple aspects / points of view / recommended options
and in the latter case automatically switches to a mode which presents the user with summary analysis of all relevant results. For example, if you search for a cabriolet car and even most of your friends like BMW you will receive a sidenote with the list of all popular cabriolet cars with an indication on which are more popular among your friends or among people with similar demographics etc. Same logic could apply to more sensitive topics such as viewing comments on political topics, controvertial events etc.
Through this enhanced experience search engine will influence users to
- Understand that there is a variety of positions / options / issues
- Give an indication on which are more accepted / preferred among his social graph or demographics
- Which in turn will make the person think and maybe make some smart conclusions.
While this is a blurry example, I think that such an approach may resolve the issue of information bubbles if applied correctly. And this feature may be viewed by users as a competitive advantage for a search engine who has it.