Sunday, December 7, 2008

This scares the crap out of me

If I have learned anything as a social scientist (and as a contractor, frankly), it's that human behavior is so widely varied and, frankly, individual, that attempting to generalize on any level besides the most basic is pretty much a lost cause.

But of course, the scientists keep plugging away.

Now, we have a computers that supposedly "detects happiness, disgust, fear, anger, surprise and sadness with 85 percent accuracy." Okay, sure, some innovative computer work, good for them. Maybe we can get some better CGI art out of it.

The problem is that "other applications of emotion recognition software might be to detect terror suspects on the basis of their emotions, not just on their physical characteristics."

Look, some of the shit America has done in GWOT has been beyond criminal. Are we going to surrender this activity to a computer? I've seen some very scary ideas floated around the various IC firms in the past couple years, and this ranks up there with the best (worst) of them. Quantification has its place - but never in the determination of an individual's current mood. Just ask my girlfriend, who constantly complains about my "happy face."

No comments: