I’ve often been asked whether there is any relation between my interest in art and my work in artificial intelligence (AI). At a practical level, where my work is aimed at making computers more “intelligent,” I’d say the connection is tenuous at best. But at a more philosophical level, AI is essentially cognitive science, which like art is intimately concerned with human perception. Questions about how visual perception works, from retinal stimulation to conceptual understanding and emotional response, are not only central to cognitive science, but probably its best-studied example. I think learning about the psychology of perception can be of value to artists, though certainly many care for it not at all, even if, through training or intuition, they are using its teachings anyway. I think a more explicit knowledge becomes especially useful in applied areas, such as how to design a pleasing and useful object of some sort, or how to alter a graphic or put words with it to achieve some desired effect.

An example of something that fascinates me from both art and AI perspectives was described in a recent New York Times article, When Language Can Hold the Answer. Several studies–more careful and convincing than previous ones–have shown a significant connection between language and visual perception. In this particular case, that translates into a different capability for color perception depending on the side of the visual field (left or right) of the stimulus, the reason being that some aspects of language are predominantly left-brain. One study found that Russian speakers, whose language makes a distinction between blues that English doesn’t have, are more sensitive to distinctions in blue in the right visual field only. Such a finding would probably not influence the way any “expressive” artist paints, but it might well lead to improving the effectiveness of an advertisement, which typically involves an image paired with certain “priming” words.


If it weren’t for artificial intelligence, I’d sometimes have no intelligence at all. Last Sunday I followed these bear tracks for miles, not exactly deliberately, but unwilling to give up my intended trail just because a bear was using it, too (and also hoping for a good sighting). I was evidently gaining on him or her, though not fast enough to catch up before the light began to fade and I turned back. On consulting several tracking books yesterday, I concluded that my quarry was most likely a grizzly out of its usual range, rather than the black bear I’d imagined. Not too smart, especially at this season.


Photographing tracks reminded me of Birgit’s duck dancing–letting animals make the marks–so I also photographed a set that I believe are bobcat, though the prints weren’t very clear in the shaded, looser snow. The black hole is the stream I had followed on the way up, and which had attracted most of my attention. Just as well.


In preparing the images above, I’ve been aware brightness and contrast and how the eye follows these cues. I experimented with slight color toning to influence the overall feeling, though in the end I preferred neutral shades of gray. These are things anyone with any background might do. In your case, does a theoretical understanding of perceptual psychology play any role in your working process? Or are you working entirely intuitively, or according to habits or methods you’ve acquired?