New study from researchers at Unanimous AI and MIT shows that AI algorithms modeled on biological swarms can significantly amplify the accuracy of human teams forecasting financial markets
A new study to be presented at the 2019 IEEE International Conference on Humanized Computing and Communication (HCC 2019), showed that teams of financial traders, when connected by AI technology modeled on biological swarms, could significantly amplify their accuracy when forecasting financial markets. The paper, Human Swarming Amplifies Accuracy and ROI when Forecasting Financial Markets, was a collaboration between researchers at Unanimous AI and Massachusetts Institute of Technology (MIT).
Using a unique form artificial intelligence called Swarm AI, researchers tasked groups of financial traders with predicting the weekly change in for common market indices – the price of gold (GLD), the price of crude oil (WTI), the S&P 500 (SPX), and the price of gold stocks (GDX). The participants were tested either (i) working alone, (ii) working together by majority vote, or (iii) collaborating in using Swarm AI technology, which enables networked teams to converge, together in real-time, on optimal solutions.
Over a testing period of 19 consecutive weeks, the results collected and compared the average individual trader achieved 57% accuracy when predicting the weekly trend across all four indices. The average group, forecasting by majority vote (often referred to as the Wisdom of the Crowd) performed moderately better, achieving 66% accuracy when predicting weekly trends. The group using Swarm AI technology to converge on optimal solutions, did substantially better, achieving 77% accuracy when predicting weekly trends. In terms of financial gains, the groups using Swarm AI achieved 13% ROI across 19 weeks, significantly outperforming individuals and traditional majority vote.
These results are consistent with other recent studies showing that networked human teams, when working together using Swarm AI technology, can significantly amplify their collective intelligence, enabling more accurate decisions, forecasts, evaluations and prioritizations. In a study recently conducted at Stanford University School of Medicine, groups of radiologists used Swarm AI to collaboratively diagnose chest x-rays. Published results showed a 33% reduction in diagnostic errors when using Swarm AI. In another recent study, business teams were tested on a standard IQ test using Swarm AI and were shown to increase their effective IQ by 14 points.
“Unlike most AI researchers, we started from the firm belief that people are really smart,” said Dr Louis Rosenberg, Chief Scientist at Unanimous AI. “Our technology connects people, using AI to amplify their knowledge, wisdom, insights, and intuition, enabling groups to perform together at super-intelligent levels. We’ve seen this work effectively for political forecasting, sports forecasting, sales forecasting, and now with this latest study, financial market forecasting. “
About Unanimous AI (http://unanimous.ai)
Unanimous AI is the maker of Swarm®, the world’s first forecasting platform that amplifies the intelligence of networked teams, and Sportspicker™, a premier sports intelligence service. In 2018, Swarm AI technology from Unanimous was recognized as SXSW’s 2018 “AI Innovation of the Year.” Unanimous’ pioneering work in this area is funded by National Science Foundation (NSF) #NSFFUNDED.
Link to paper: Pre-Print
Want to learn more about our Swarm AI technology? Check out our TED talk below…