“If the problem you are trying to solve requires a magnum opus, you are solving the wrong problem.” -Aza Raskin
Developing a clear understanding of a problem assumes we can agree 1> there is a problem. The next part is 2> refining that problem, so 3> we can begin to draft a scalable solution.
In scouring the proposals that initially came along with the call for ideas, I was struck by the number of ideas attempting to create highly personalized news models. The initial proposal I came to the table with: “Popping the Filter Bubble” was a brief write-up (note: if the term “Filter Bubble” isn’t ringing a bell, I’d suggest 1> this TED Talk, and 2> Evgeny Morovoz’s book review/analysis.) Pariser recently helped to reignite a conversation surrounding potential dangers of personalized news and information streams, highlighting a concern that a narrow junk-food based media and information diet can have negative political and cultural ramifications.
What the problem isn’t:
Personalized search in and of itself.
What this isn’t an attempt to do:
Break or reveal super-secret algorithms.
Force “information vegetables” all day everyday.
But, it would be awesome to:
1> Bring more transparency to the filtering process. 2> Promote an understanding that gatekeepers in news and information dissemination haven’t disappeared. The gatekeepers have moved, and the gatekeepers look different, and sometimes they look like algorithms.
Let’s bring more of the known factors filtering our search results to the forefront in a way that is obvious, and see if we can’t manage, play with, change and easily share our results when we alter that information. So, see below how it’s remarkably easy to test how our our top results are altered when we change our location data - for instance,
My search results this morning (7/18) for “chocolate” (Washington, DC):
And “chocolate” (Manually changed location to Minneapolis, Minnesota):
What would be exciting:
For there to be away to do this with more than just location, so some of the other 57 factors featured into the search process. (Factors like: what sort of browser you’re using and what version, how large your text is, etc…) then: it would be excellent to make those results easily and immediately shareable.
Problems in execution:
I call myself an “aspiring coder” because the term is both at once full of promise and impossibly vague. To be clear, I would have no idea how to accomplish any of this, but would absolutely be interested in hearing your thoughts as to feasibility. If it’s impossible, maybe there’s some more refined way to make it work.
In terms of practicality:
It’s a bit easier to discuss this problem conceptually, and more difficult in terms of practicality. While there have been some not altogether scientific experiments examining differences in search, the borders of the problem/issue remain fuzzy.
soon later today, 7/18:
Visuals, bearing in mind the “just build it” mantra we’ve been hearing and attempting to internalize.