Wed 14 May 2008
Systems of glaciers, music, and algorithms
Posted by Betty Barrett under Just for Fun , New TechnologiesComment Here
If you like some of my other posts I hope you will enjoy this one as well. Systems are such a fascination for me - they are complicated, not always obvious, but always full of the unintended. It intrigues me when I stumble on someone else’s writing and they are using a systems lens. Recently I found several examples of people describing events and giving us a glimpse of the systems the events exist within. For example, Sharon Begley in Newsweek is worried about global warming. Her piece starts with the false consolation she received reading a website sponsored by the coal industry which “painted climate change in Edenic terms, promising that the atmosphere’s rising levels of carbon dioxide would act like airborne fertilizer, boosting crop yields and turning marginal regions into breadbaskets.” This provided her a segue to the Gangotri glacier. Here’s a paragraph that captures a great chunk of a system quite nicely beginning with the glacier.
One of the Himalayas’ largest, it has been shrinking since the late 18th century. But over the last 25 years it has shrunk about half a mile, a rate three times the historical norm. The retreat threatens more than the loss of a panoramic background for tourist snapshots. The Gangotri supplies 70 percent of the flow of India’s Ganges River during the dry season, when farmers depend on it for irrigation. Glaciers on the Tibet-Qinghai Plateau, which are also shrinking, feed major rivers in China, which are also crucial for irrigation. “Without the ice melt, the Ganges and the Yellow rivers could dry up in the dry season, shrinking harvests,” says Lester Brown, president of the Earth Policy Institute. “If the Ganges flows only part of the year, double cropping [in which farmers plant rice and wheat in back-to-back growing seasons, and which underlies India's green revolution] breaks down.”
In one fell swoop Begley goes from the coal industry to food shortages caused by the shrinking of a glacier. That’s nice work.
The New Yorker carried a piece by Alex Ross that also seemed to capture an amazing view of another type of system or perhaps, the links between a couple of systems, the natural and the technological. Ross interviews John Luther Adams, an American composer living in Alaska. Adams writes music in collaboration with Alaska. Using technology, Adams has created a performance installation - a room that “translates raw data into music.”
At the Museum of the North, on the grounds of the University of Alaska in Fairbanks, the composer John Luther Adams has created a sound-and-light installation called “The Place Where You Go to Listen”—a kind of infinite musical work that is controlled by natural events occurring in real time. The title refers to Naalagiagvik, a place on the coast of the Arctic Ocean where, according to legend, a spiritually attuned Inupiaq woman went to hear the voices of birds, whales, and unseen things around her. In keeping with that magical idea, the mechanism of “The Place” translates raw data into music: information from seismological, meteorological, and geomagnetic stations in various parts of Alaska is fed into a computer and transformed into an intricate, vibrantly colored field of electronic sound.
Basically Adams has created a technological system that uses the natural system to create a real time musical experience. The sounds are generated by what is happening outside the museum and around the museum. The sounds are generated by earthquake activity, sunshine, and aurora activity. I wish I could hear this music because it must be amazing. At the end of the New Yorker article there is a connection to another Adams piece entitled Dark Wave - although it was not rock and roll to me, it was strangely mesmerizing. Adams says in The Place piece, he is trying to capture the “unheard” sounds of a place that he loves; “My music is going inexorably from being about place to becoming place.”
Michael Schrage discusses the system that allows merchants like Amazon to make recommendations to customers on their websites. In his piece in Technology Review, Schrage says “The truth is that I now get more good recommendations about more things, more often, from Bayesian algorithms than from my best friends. Perhaps this should make me wistful, but it doesn’t. Better technology doesn’t mean worse friends.”
[Bayes' theorem or law uses probability theory to assess the probability of two random events being related. It is heavy mathematics and forms the basis for lots of work in artificial intelligence. ]
Schrage describes the recommendation system at the Amazon site — you know the one that say if you liked this book then you might also like all these other books that others who liked this book also liked. Really. it is just another way to get me to buy more books, which I have to say is not all that difficult. Here’s Schrage’s more elegant description.
Click-throughs are the currency of the recommendation nation. The more choices you make (or decline to make), the more finely tuned the recommendations become. The more your peers interact with Amazon, the better Amazon’s engines can infer which recommendations will make the most sense for you and the most dollars for them. The result is that recommendations can become breathtakingly profitable examples of what economists call “network effects,” where a network’s value is proportional to the number of its participants.
Because Amazon’s customer base is so large and the algorithms have so much data to work with, the recommendation system provides a service that I enjoy by pointing out books that I might otherwise not have seen or discovered. It is also astounding to me that the system works so well. It makes my shopping experience much more fun and I don’t need to get out of my pajamas. I only wish it were as easy to meet the others with similar interest in real life - the conversation would be terrific.
Using Bayesian algorithms to help me find books is only one way this mathematical system is changing our lives. The NYT just featured Daphne Koller, an artificial intelligence researcher, whose work is leading to advances in fields as diverse as traffic control and cancer diagnosis. In essence this work harnesses the power of systems to create new insights that are generally useful to all of us as we deal with the increasing uncertainties of our time.
Systems power is a force that not only gives us insight, it can also give us headaches as we recognize that we just do not yet have the capability to see the whole structure of any system. Certainly we can’t avoid unintended consequences. It really does teach us that we can’t control anything but we can appreciate its complexity.