AlphaGo Pushes Human Go Players To Be More Creative

A recent study published in the journal PNAS found that since AlphaGO’s landmark victory in 2016, Go players have become less predictable and have embraced newer strategies, assigning each move a “Decision Quality Index” (DQI) with the help of “superhuman” Go AI. Prior to 2016, the quality of professional play improved relatively little from year to year, and in some years there was even a drop in overall quality.

In 2018, median DQIs changed at a rate above 0.7, and professional players were using newer strategies. The researchers suggest that the development of superhuman AI programs may have encouraged human players to abandon traditional strategies and explore new moves, which in turn may have improved their decision-making.

“Our findings suggest that the development of superintelligent artificial intelligence programs may drive human players to move away from traditional strategies and explore novel moves, thereby improving their decision-making,” the team wrote.

“It’s not surprising that players who train with a machine tend to make more moves that the machine recognizes,” said Stuart Russell, a professor at the University of California, Berkeley.