Foundation and Twitter
In the 1940s, Isaac Asimov began publishing the shorts stories that would eventually become the landmark Foundation series. In it, he created a science called psychohistory that was able to mathematically predict the future on a large scale. In other words, even though individual human actions or small-scale events couldn’t be predicted, the actions and destiny of an entire population of people could be.
And, now, that fantasy might be starting to show the first glimmers of a reality…thanks to Twitter, of all things.
According to a recent article in GigOM, an MIT researcher has created an algorithm that can predict Twitter trends hours in advance with 95% accuracy:
Associate Professor Devavrat Shah says his model has been 95 percent accurate during testing and has been predicting trends hours before they appear on Twitter’s list. The algorithm incorporates a new approach to machine learning that compares real-time data with historical data and predicts outcomes based on past events that most closely align with the current situation. So, rather than analyzing a topic’s chances of trending equally against the entire historical corpus of topics, it will assign more weight to topics whose paths followed similar trajectories up the ranks of top trends.
Of course, since this is social media that we’re talking about, the immediate context for this type of capability is around advertising possibilities, although the article mentions its relevance for stock prices, tickets sales, and “other dynamic quantities.”
And what’s more of a dynamic quantity than the entire human species? I mean, in Asimov’s fiction, psychohistory was abstract. Like theoretical physics. However, with Twitter and other social media, we suddenly have billions of real-time data points over a global populace covering the entire range of human experience and action. And, as time goes on and supposing social media doesn’t go the way of penny loafers, we’ll eventually have decades worth of real-life data points. Enough, in fact, to start modeling the entirety of human behavior on a large scale, especially since that behavior is becoming increasingly globalized and homogenized.
We could be predicting changes in mores, knowing the qualities of future world leaders, population changes, society collapses, technology advances, maybe even to the point that we could make course corrections when the data predicts disastrous outcomes. Although we might need a secret society like the Second Foundation for all that.
Basically, it’s your responsibility as a citizen of this planet to continue posting about everything you eat, what you think of every Adam Sandler movie, the funniest thing you heard on the Subway today, etc. We’re talking the future of the human race, here.