A piece of artificial intelligence software created to mimic the “wisdom of crowds” correctly predicted that the #MeToo movement and its silence breakers would win Time magazine’s Person of the Year award.
A group of researchers at Unanimous AI (UNU), a start-up based in Silicon Valley, devised a machine that acted like a hive or swarm, and literally called swarm intelligence, to predict the result from the 33 shortlisted options.
The machine worked on the principle that a large group is capable of solving more complex problems than individuals – the theory is used partly to explain birds flocking and fish creating schools.
The theory, and the algorithm created by Unanimous, worked on the idea that swarms are more efficient because they tend to converge towards the best answer.
It was also inspired by bees, who use swarm thinking to come to decisions regarding their hive based on converging to one answer.
The algorithm was combined with real-time human input that included forecasts based on events in the real world to create what the developers called an “emergent hive mind”.
It worked, with the UNU correctly predicting the winner of Person of the Year for the second year running. Having initially been leaning towards a Vladimir Putin victory, the swarm altered its choice just days before the announcement, settling on the #MeToo campaign, which was confirmed as the winner earlier this month.
Now scientists and researchers are debating how this type of algorithm could be used in other scenarios – including in medicine to help come to treatment or other decisions in a more calculated manner.