I love to discover new authors, filmmakers, genres, artists, facts, and viewpoints. I love to explore the world, its experiences, and its search for meaning through a wide variety of eyes and ears, filters and assumptions, joys and hardships.
With the advent of the Internet, digital media, and extraordinary personal technology I now have more access to more titles across more categories and genres and viewpoints than ever before. More people can produce, and more people can consume without the prohibitive barriers to entry that kept small and independent voices from entering the market.
It should be nirvana for someone like me, but that’s not quite how it’s worked out. It turns out there are too many titles, too many authors, and too many voices to get a real handle on the vast diversity available. It’s hard to find a particular thing.
Likewise, with that massive democratization of content has come a increase in easy accessibility to truly awful stuff—poorly written, poorly told, poorly produced, over-attenuated, overwrought. There may be ten times the number of available titles, but Sturgeon’s Law still holds true—ninety percent of it is still crap.
Services like Pandora do an interesting thing with music. By applying a battery of technological deconstructions of various musical elements they’ve developed a Music Genome Project that can actual analyze digital music files for core elements and determine affinity attributes, giving their service an ability to find similar music and add it to a dynamic playlist.
The effect is that the services gives me more of what I already like, from a wider variety of artists. In other words, it helps me discover new artists already in-genre, but it does little to reveal complementary genres or new styles (at least it hasn’t for me so far).
Amazon uses buyer behavior analytics combined with ordinary product categories to show trends in both browsing and purchasing habits. This helps me see what other people are looking at and provides a chance for me to look at products or titles that have little direct affinity to a currently selected item, but that are connected in the minds of other buyers by non-intuitive or complex means. It leverages the minds of customers as revealed through browsing habits to open previously unconsidered avenues to me.
Netflix uses simple objective categorization in combination with massive opinion-gathering from customers to create both absolute and relative approval rankings. By further analyzing an individual subscriber’s selections they create a more complete affinity map that can create dynamic categories never intended by the original artists or distributors. The effect is that they can map my affinity for dark (yet hopeful) stories set in foreign (or fantastic) settings that reveal philosophical or spiritual explorations in odd or absurd ways without concern for realistic media.
In other words, Netflix is—through massive analytics of both observed behavior and polled preference—coming closer to predicting interests I didn’t know I had, but that were suggested by the films I both watch and like. Combined with reader reviews that express specific reasons for rankings, Netflix gives me the most effective set of recommendations that help me reach beyond my current interests to new and varied interests of any other service so far.
(The fact that my children watch off the same account messes that up a bit; the blend of children’s animation; moody, eastern-European or Asian absurdist drama; and science fact/documentary/Top Gear by various family members must be stressing their algorithms a bit.)
But even Netflix comes up short because it doesn’t capture subtlety well. It’s amazingly powerful and effective, but still a bit of a blunt instrument.
Ultimately, the only deeply effective tool for recommendation I’ve found is simple conversation with people who know something whereof they speak. Not just the high critics, but pop critics and ordinary consumers with a modicum of self-awareness and communications skill.
In other words, the blogs, community sites, conferences, and other discussion forums that bring together a diverse audience that share some elements of affinity, but are not limited by that affinity. The intelligent discussion of ideas, preferences, and hopes that makes a previously uninteresting title more intriguing.
As always happens with the advent of new technologies, people worry that the new thing will supplant the old. But just as eBooks is expanding to total number of titles sold (the reduction in print sales is less than the increase in eMedia sales), technological means of identifying affinities and recommending titles is only another tool in the quiver of the arts explorer—just one more way of being exposed to new titles and new visions.
With the rapidly expanding availability of eMedia works, we need the best of both—human and technological analytics—to sift the vast possibilities and expose the jewels, the pearls of great price waiting unnoticed to be found by those who can be most served by them.
Suggestions for tools, applications, blogs, forums, etc. that touch on film, music, art, literature, essay, history, poetry, and more would be greatly appreciated to help the rest of us discover those jewels.