Dave Allen

Guest post: Web curation and filtering: Defining new roles for digital artists

posted by , 1 Comment

By Maurice Boucher:

In our real world lives we collect objects and ideas. Those who do this professionally (curators) tend to have highly developed organizational skills and the rest of us . . . well, the most skill we use to manage our ‘stuff’ is little better than composting for the garden (admit it). The internet world consists of massive amounts of data and media that, lets face it, can barely be called anything as generous as a collection. The terms we apply to this super-state (the cloud, the stream, the digital pie-hole if you like) express our complete capitulation to the chaos of the Internet and to disregarding any notion of trying to solve the problem.

We have Google of course and a few other search engines to help us navigate our way through the vast ocean of flotsam and jetsam. It is too bad search engines are really just the Internets’ equivalent of hanging a pretty picture over a gaping hole in the wall. Google (and a few others) have some pretty nifty algorithms to filter keyword searches so that we get a nice page of links with enough near-hits and not too many misses to give us either the illusion that; a) we have the necessary keyword query sharp-shooting skills or b) there is something or someone out there in the great googly-Mowgli internet that is paying attention to our needs. For those of us who like to have a peek behind the picture to see what’s inside that hole in the wall, we have the promising technologies like HTML5 and CSS3 to see if we can plaster that hole with ‘the semantic web’ which basically amounts to the Internet having some cotton-picking clue about what it is actually talking about.

Of course the Internet does not ‘talk’ to us. It is more like a glorified TV that politely pretends to listen to you now and again; which is better than current TV technology that just doesn’t ever shut up until you turn it off, but I have higher hopes for the internet. The Internet should eventually develop into a technology and a data resource that can understand us at least to the level (and that’s a low level) that we understand ourselves. Digital curation is a significant piece of that strategy.

The people behind the scenes at Google and other enterprises working on this problem focus on technologies like Artificial Intelligence and other software and processor intensive solutions. I think this is a mistake and in Dave Allen’s earlier post, The filtered Internet, curation and algorithms a video of a TED talk by Eli Pariser suggests there is at least some logic in the search algorithms devoted to parsing results according to perceived relevance that amounts to a ‘filter bubble’ myopia. I would go further than that and say there is an intrinsic tendency in developing software logic (especially if there is a marketing department representative in the development meetings) as a digital gatekeeper/filter (beyond sponsored links and personalization) that iteratively manipulates and manages user expectations over time creating an internet ‘point of view’ or perspective on the world that expedites a false sense of customization and appropriateness of search results.

Recognizing that bias in digital media is a major part of where we are going, and that there is a limit to an algorithms ability to parse meaning and intention out of a handful of keywords, even if it references the online personal history of the user, means we should see this as a challenge and an opportunity to redefine traditional roles and invent new solutions to aid in creating an Internet that more accurately reflects human aspirations and goals.

As I remarked in response to Dave Allen’s post I know of no algorithm that can work out the difference between what people ask for and what they actually desire. That is the philosophical question that really is the core software requirement of a music recommendation engine, and music curation is an ideal testbed case to see if we can build a layer on the internet to act as verification of the search process. I remarked in Dave’s post that communicating socially and informally (with strangers) and sharing music is not enough to build a bridge between what people ask for and what they desire. People have to have a sense that some agency is acting at least semi-exclusively for them and has some insight into who they are.

This is already happening almost incidentally on YouTube. I subscribe to YouTube channels that offer me a constant feed of videos that really amount to music recommendations. Some channels are overtly commercially invested in getting me to listen to some new music (I mostly avoid those). The channels I like best curate to develop a tribal idiom across a particular style of music or artist discography. This ability for anyone to create their own YouTube channel combined with the modest commercial incentives through advertising revenue YouTube offers is not a replacement for real world artistic activity, it is probably not yet even a realistic supplement for physical distribution and promotion (the PomplamooseMusic example aside). It is an outlying indicator of the road ahead, nothing more, but when we are finally ready to view the digital artist/musician as an inheritor of traditional roles that could only be fulfilled by 20th century musicians (music curation both in composition and song catalog) and D.J.’s (remix of both stylistic context and musical artifacts combined with creating a tribal personality) then the movement will gain momentum.

At the heart of the online music curation role is the possible solution to the expressed need verses unexpressed desire problem that permeates the Internet and prevents us from developing internet culture beyond purely commercial interests. For curation to work it requires that whatever algorithms are created, they at the very least must expose an interface to someone who has insight into music’s influence on the human psyche and give them some level of economic incentive to devote the time and energy into using the tools. The professional musician down through the ages is THE specialist in dealing with the expression and satiation of human desire. It would be wise of companies like Google to pay attention to the experiments of outliers and the outlying audience that use their tools like YouTube in unexpected ways instead of relying solely on artificial intelligence to ‘shape’ the data mining results into more digestible forms. The artists have to be included in the equations that run the algorithms of curation and filtering for the internet to have a future beyond being just another compendium of useless facts and trivia.


Comments