NEXT: Can We Gauge Social Mood?May 13, 2019 1:11 pm David Staley
In the iconic graphic novel The Watchmen, the evil genius Adrian Veidt gazes at a wall of scores of television screens, each tuned to the world’s media, news reports, entertainment. (This is set in the 1980s, don’t forget.) As Veidt scans all the visual chatter, his mind sees patterns across the screens. At one point he pronounces to his henchmen:
“Hm, let me see…First impressions: oiled muscleman with machine gun…cut to pastel bears, valentine hearts. Juxtaposition of wish fulfillment violence and infantile imagery, desire to regress, be free of responsibility…” And then he pronounces “This all says ‘war.’ We should buy accordingly.”
We’re never quite certain what goes on in that brain or how he arrives at his conclusions or how he has the mental capacity to so easily identify signal in all that noise, but this is all part of his persona as an evil genius.
The Defense Advanced Research Projects Agency (DARPA) are today envisioning using artificial intelligence to accomplish something like Veidt’s feat. As described in an article in Tech Crunch, “A new program at the research agency is aimed at creating a machine learning system that can sift through the innumerable events and pieces of media generated every day and identify any threads of connection or narrative in them. It’s called KAIROS: Knowledge-directed Artificial Intelligence Reasoning Over Schemas.”
In psychology, a schema refers to the cognitive framework by which we organize and interpret information. DARPA, in effect, want to replicate this cognitive process through artificial intelligence. The goal, like Veidt’s, is to identify the zeitgeist of the moment, to capture the mood of the world as a way to anticipate events. DARPA are interested in this technology as a way to anticipate hot spots and emerging security threats around the world before they occur.
Obviously, as a futurist, I am at once intrigued, and perhaps even a little threatened, by the potential of this technology. One of the things I do as a futurist is to scan the environment, to take in as much information about the world as I can, looking for clues to potential trends or signals for what might be happening next. My “Next” columns are one such effort to scan a lot of information and to determine the meaning of that information.
Both Veidt and KAIROS are attempting to understand “social mood.” When Veidt is scanning all of those television screens, he is gauging the social psychology of the population. The concept of “social mood” comes from socionomics, the science of social prediction coined by Robert Prechter. Adherents of socionomics believe that mood influences events rather than events influencing mood, that the collective mood creates a certain kind of environment that influences actions. If we can accurately measure social mood, so the thinking goes, we can anticipate events.
Socionomics looks to stock markets as the best measure of social mood. When the stock market is rising, for instance, this reflects a more positive social mood. I have problems with using the stock market as the gauge. Investing in stocks reflects only a small segment of the population, and anyway so much trading today occurs through algorithms and automatic processes that I wonder if the mood being gauged via the stock market is the mood of AI.
I do not discount the importance of social mood, however. I just believe we need better ways to capture it. We could look to what is trending on Twitter or what people are liking, loving or hating on Facebook as one way to measure social mood. Using social media in this way might be the 21st century equivalent of Veidt’s 1980s television watching.
I have questions as to whether or not it is even possible to replicate human schemas. Do we truly understand how schemas work in our own minds? Even if we were to develop an artificially intelligent means of uncluttering the information to reveal patterns of social mood, we still require a human intellect to make sense of those patterns. Adrian Veidt scans the TV screens and identifies the pattern: that process could very well be automated. But then Veidt tells us what those patterns mean. I’ve no doubt AI can be developed to help us see patterns that might otherwise go undetected. But it is still only a human mind that can make sense of the patterns.