I probably still have some friends and/or acquaintances out there that know a thing or two about the life I led before I focused this site on writing about the performance of music. Since I spent most of my time at the Artificial Intelligence Laboratory at the Massachusetts Institute of Technology between the beginning of my junior year as an undergraduate until the approval of my doctoral thesis, my thoughts about that discipline are more than abundant. Nevertheless, I have tried to hold my metaphorical tongue about that background when writing about the performing arts. I had no trouble doing so until this past Sunday, when the pink Datebook section of the San Francisco Chronicle devoted two articles and four pages to the topic, along with a headline on the cover: “AI is opening new horizons in music. Computer systems are pulling off previously impossible innovations.”
I have enough experience to assert with a fair amount of confidence that, since my student days, the very idea of what artificial intelligence was has been abused as much as it has been used, if not more so. I would therefore like to attempt a “reality check” to “clear the air” about how the concept was conceived and what has happened to it since then. When I was a student, there was only one book on the topic, Computers and Thought, a generous anthology of previously published articles edited by Edward A. Feigenbaum and Julian Feldman, both at the University of California at Berkeley at the time. (Feigenbaum would go on to launch the Artificial Intelligence Laboratory at Stanford University.)
The anthology began with an essay by Alan M. Turing entitled “Computing machinery and intelligence,” which was published in the October, 1950 issue of the journal Mind. The title of the second section header was “Is It Possible for Computing Machines to Think?” The very first word of the first paragraph in that section was “No!” Nevertheless, he did not dismiss the possibility of artificial intelligence research if one could set at least one realistic goal. The goal he proposed was this: “to construct computer programs which exhibit behavior that we call ‘intelligent behavior’ when we observe it in human beings.” In other words, software might not be able to “make it;” but there might still be some value in software that could “fake it!”
Well over half a century has passed since Turing’s words were published. During that period, there gradually grew a faction that was less interested in pursuing the challenging questions behind artificial intelligence and more interested in “selling the brand.” By now that salesmanship has confined research to pursue challenging questions about intelligence to a very small corner, almost impossible to find in the thick fog of what amounts to little more that hucksterism delivered in frequently elegant and often opaque language.
Now, to be fair, technology has been responsible for a plethora of changes in how the performing arts are now practiced. Unfortunately, there are no end of promoters that like to hang the term “artificial intelligence” on many of those changes without even the slightest idea of how to justify the claim. As a result, the old adage “There’s a sucker born every minute,” usually attributed to P. T. Barnum, may have devolved into “There’s a sucker born every microsecond!”
Sadly, we now live in an age in which it is almost impossible to tell the difference between credible claims and sheer balderdash. My guess is that neither of the Chronicle writers were particularly well-versed on what technology can and cannot do (and whether what it can do has anything to do with “intelligence”). The good news is that the eyes of most Chronicle readers glaze over from any encounter with technology. Nevertheless, in my current capacity, I fear I shall have to be on guard when I encounter a contemporary performance (even one by a writer) that, beneath the surface, turns out to be little more than one of those nineteenth-century medicine shows!
No comments:
Post a Comment