A.I. Predicted Oscar Winners with Stunning 94 Percent Accuracy
For third year in a row, swarm A.I. outperformed industry experts.
Unanimous AI is known for accepting high-profile challenges from journalists, testing the power of its Swarm AI technology across a range of venues. After correctly predicting the World Series and Kentucky Derby — and perfectly forecasting Trump’s 100-day approval rating before he even took office — Unanimous AI took aim at the 90th Academy Awards, publishing its prediction here at Inverse last week.
How did Unanimous do? Its Swarm AI technology was nearly perfect, achieving 94 percent accuracy across the 16 major award categories, with only one award not coming out as predicted. The technology was 100% percent accurate in forecasting winners in the six major categories, including Best Picture. In that category, Swarm AI disagreed with most industry experts – and the Vegas odds – all of which deemed Golden Globe winner Three Billboards Outside Ebbing, Missouri the favorite over Swarm AI’s predicted winner, The Shape of Water.
Swarm AI combines real-time human input with artificial intelligence algorithms, optimizing a group’s combined knowledge, wisdom, insights and intuition into a unified output. In other words, Unanimous builds artificial “hive minds” that amplify the intelligence of human populations to create an artificial super-expert that can outperform traditional experts.
That’s transformation from average fan to “super-expert” is exactly what happened for this year’s Oscars. Researchers at Unanimous AI formed a swarm of 40 movie enthusiasts connected in using AI algorithms in real-time. The chart below shows Swarm AI predictions compared to industry experts and major publications.
“Unanimous A.I. gathered movie nerds together to predict Oscar winners far better than most critics could,” says Inverse culture editor Grace Lisa Scott. “Using artificial intelligence, the technology successfully created eerily accurate results.”
Below you can see a full list of the Swarm’s predictions for the Academy Awards. As you can see, the contenders in each category has been ranked according to the likelihood of victory according to the AI and the winners have been highlighted in green. In 15 of the 16 categories, including all of the six major categories, both screenplay categories, and even the documentary and foreign language categories, the AI was able to perfectly predict the winner of the Oscar. In the single category where the AI was incorrect, Best Production Design, the eventual winner was deemed 2nd most likely to win, and the AI did note that its choice was only 50% likely to win.
Thus, the chart below reflects just the latest example of a group of people combining their wisdom and intelligence with AI algorithms to beat the experts at their own game.