NEWSWEEK - Nov 22 - Dating sites today are just marketplaces. Over the years, many have used algorithms to predict matches, but none have improved the odds much. OkCupid co-founder Christian Rudder, is skeptical about algorithmic attraction. “One person doesn’t kick out enough data,” he said. “Someone will look great on paper and then have an annoying laugh.” Researchers are inventing a new generation of computers based on the way human brains work. Instead of processing data through algorithms, these machines can watch data streaming past and see patterns. Numenta watches activity for Amazon Web Services and looks for patterns that can give early warning about a hacker attack. “Our systems are like brains at birth,” says Numenta founder Jeff Hawkins. “They just learn from the stream. You don’t have to tell them anything.” Another step will be to apply machine vision to analyze Tinder-style swipes. Tinder doesn’t learn anything from users' swipes—it doesn’t come to understand what physical features they like. But with advances in machine vision, it could.
by Kevin Maney
See full article at Newsweek
See all posts on Tinder
See all posts on OkCupid
Summarized by 
