AI agents manage a Premier League fantasy soccer team without beating humans yet
A developer deployed a swarm of AI agents to build and manage an experimental Premier League fantasy roster while flying to visit family.
Initial model suggestions lacked real data
The first attempt used a tool from the UK's Turing Institute that ignored injuries and management changes. Those early suggestions were remarkably bonkers because they did not incorporate real-world information. A later setup collated more information but still produced surprisingly poor team selections.
New agents read news and run models
The developer asked Codex to create a harness that guides different agents through various steps. Several agents were tasked with reading news and blog posts while others analyzed recent results. More agents ran models to make predictions based on the gathered data.
Results remain below human performance
The initial team suggestions left star players like Erling Haaland on the bench. The developer expressed surprise that the system could not outperform humans at this stage. Frontier models are expected to get better at making predictions in the near future.
Reported by one outlet
Only one outlet has published this. Nothing here has been checked against a second report, so read it as that outlet's account and follow the link below for the original.
Reported by
1 independent outlet. Headline as published. Links open the original report.