Artificial intelligence startup Magos has developed software that combines neural networks and blockchain technology that simulates a collection of biological brains to accurately predict the outcome of future events relating to sports, politics and even behaviour.
The software learns and recognizes complex patterns in data allowing it to predict patterns, trends and even more subtle ticks, making it a powerhouse at forecasting.
The software was tested over a four-month period earlier this year, betting on multiple sporting events, and did notably better than other gamblers (both human and bots) creating a 30% higher return on investment.
By using the advances in artificial intelligence (AI) and machine learning, and tapping in to a wide variety of platforms, including conventional prediction markets, exchanges and sportsbooks, the firm is on the path to building the ultimate forecasting tool.
“MAGOS is based on a system of neural networks. Each network performs a specific task, but they work together in collaboration. The backbone of MAGOS is its modular architecture. It allows us to develop and implement individual forecasting modules, targeting different kinds of forecasting domains, from business and finance to sports and politics,” said Andreas Theiss, data scientist and CTO at Magos.
MAGOS is essentially a complex, scalable AI based on five neural networks that, when working together, have the power to predict the outcome of any future event with high accuracy, which not only outperforms conventional prediction systems but also humans.
MAGOS can be used for forecasting events in any domain, provided that there is sufficient historic data to learn from.
AI is transforming the world we live in and has already managed to achieve some amazing breakthroughs, including spotting cancer in tissue slides more accurately than doctors, flagging errors in legal documents and even beating the world’s top gamers at Dota 2, an extremely complex video game.
Magos is currently holding an initial coin offering (ICO) to crowdfund future development and has already raised half a million dollars.
Magos aims to be the leader in collaborative neural networks that are fine tuned to predict the future.