Media as Software: A Conversation With Doug Turner

Kiki Mills at MITX introduced me recently to Doug Turner, whose past includes eight years as a member of the 3D graphics research team at Apple's Advanced Technology Group.  Doug and I met for breakfast and talked shop about digital media.  One of Doug's ideas, which I found particularly interesting, is (his words) the concept of "media as software".  Right now rich media streams are largely analog audio and video once they are "published".  (If you've composed or edited a digital video "project" and then converted it into its final form, you know what I mean.)  Doug describes this  as publishing digital media as platforms on which other people can add/edit their own stuff. 
 

Green-screening the base content, as Stephen Colbert has done, is an "analog +" approach to this idea.  Second Life is a good example that provides tools in which people can build stuff using Linden Labs' language and API's, but Doug has something a little more accessible in mind.  He's written some code that allows him to embed static images in a variety of interesting contexts.  One can imagine mashup services that might provide users with libraries of iconic images, and tools like Doug's which would allow them to embed/ mix/ mashup their own content.  Here's an example of Doug's work, which he calls "Mood Shifted Photos", at Flickr:

http://www.flickr.com/photos/36521967972@N01/sets/72057594142043889/show/

Comments

[Douglass Turner]

Cesar,

The mood shifted photo concept is actually only part of a larger trend of media with software underpinnings. The roots of this trend go back to a fundamental theme of digital media: model vs. content.

GooTube video is content. iTunes music is content. Flickr photos are content. Flash animation is content. All bits. Yes there is meta-data (user tagging/comments/link) surrounding the bits. But at the end of the day bits is bit.

Contrast this with: The ASCII text on a Weblog. The built spaces and avatars of Second Life. These are models. Representations. Software loves models and representations. Google page-rank (and thus Google itself) wouldn’t be possible without the ASCII representation of text. Every avatar within Second Life can be unambiguously located, tracked, and identified because it is a model. Models are easily transmitted from a source and faithfully replicated at a destination. Models can be easily altered, remixed, and replicated in a variety of ways and level of granularity that audio and video content can’t approach.

I have lots more to say on the subject. Suffice it to say we are just scratching the surface of the emerging era of software as media.

Cheers,
Doug Turner