People ask me about cloud computing sometimes and I’m usually ambivalent. Ok, distributed computing is useful, but it’s really just an efficient use of existing resources -a networking optimization. I don’t necessarily buy it as revolutionary concept. People try to democratize the system, but essentially the most-discussed paradigm is informationally asymmetric (the work of the many serves the needs of the few) so there’s not much immediate benefit, or guarantee of one, for most of us. Cloud computing is this year’s “timeshare”. But the idea isn’t a wash if you mix it up a bit. Let’s use music as a metaphor.

I used to listen to music on the radio. That sucked. What few stations I got were awful, and when I heard songs I was interested in, it was random. Ok, it was free… but there were ads and the songs I liked played too often, or not often enough. It was someone else’s playlist and that meant tons of crap to endure between the good stuff. Frustrated, I bought albums and made my own mixtapes. Eventually, my tapes became CD’s and my mixtapes were replaced with .mp3 playlists. The important thing is I went from a limited number of music channels, and someone else’s audio agenda, to unlimited channels and my agenda. Old news, I know. But what both of those shifts shared in common was the limited number of people responsible for creating the playlists. Cloud computing changes this.

These days I listen to music online. Pandora is pretty cool, but sites like The Hypemachine just might represent the future of entertainment. Hypemachine is curiously different than other online music stations because it scans and aggregates music blogs on the web and allows you to listen, from one central place, the contents of the blogosphere. Notice the audio agenda? The playlist is not generated by any single person. It’s generated by what all the people are talking and posting about. People can vote on their favorite songs of course, and seach data is tracked, so popularity searches inside the site are possible. But theoretically, with metadata, the search criteria are limitless. What’s interesting about this is that a computer, performing an unbiased statistical analysis of individual contributions, automates the distribution and exposure of relevant new media customized individually to you. Sorry, that was a mouthful, but that’s a form of cloud computing.

In mass communication, this concept is referred to as a “Gatekeeper”. In my music metaphor, DJ’s used to be the gatekeepers. They controlled which songs were played, they were the filter. If he had similar tastes as you it was great. But if he was taking payola, you got what some record company wanted you to get. As the Information Age progresses, and the sheer volume of data and media we need to sift though becomes larger and larger, we need assistance to find the signal in the noise. Increasingly, software has become the gatekeeper of choice. Amazon is scary in its ability to select book choices I might be interested in. I sift Google News using topical filters. My RSS tracks updates to only the few sites I’m interested in.

Where is this all headed? Some say the Semantic Web. We teach the World Wide Web to index itself, to classify our information automagically, and filter it for us. But the problem is computers don’t think like we do. As we become increasingly dependent on computer algorithms in the gatekeeper role, successful distribution of your ideas will hinge on your ability to be indexed and understood by a machine. In short, data on the web will have to comply to whatever indexing mechanics become popular. We see this now to a limited extent with Google. Search Engines index content by using simple keywords and a few other criteria to determine relevance. Exploiting this system and topping a search list can yield substantial financial gain as well as advantageous memetic fecundity. Complex ideas and metaphorical constructs, the cream of human expression, are largely incompatible with machine logic because they’re heavily reliant on human, not logical, experience. The risk here is that in a rush to accommodate convenience we’re potentially changing our communication behaviors. If the basic building blocks of ideas are formulated using language, linguistically speaking you can then argue this potentially changes the very way we think. I don’t know about you, but I already observe a dangerous amount of literal, mechanical thinking and not nearly enough of the creative thinking accountable for our most progressive insights. To be fair, the Semantic Web can encompass these values, but realistically an implementation will be in place before these technical limitations are solved.

I sometimes wonder what they’ll call the World Wide Web when the Moon gets its own IP address. A thinking human and a thinking computer can parse that statement, but only one might smile. The sad fact is humans are already short-sighted enough as it is, so let’s design informational systems that promote creativity and insight, not ones that limit it. I’ll concede this point when a computer can be programmed to understand my dirty jokes or explain to me exactly why I hate the term “World Wide Web”. But until that time let’s not let them control it.