Thread by @NAChristakis

Super-cool @PNASNews study examines historical changes in decision-making by professional Go players from 1950 to 2021, focusing on changes in game play after the advent of superhuman AI (i.e., AlphaGo). pnas.org/doi/epdf/10.10… 1/

Human players in human-human matches began to make significantly better decisions in Go following the advent of superhuman AI. Players’ strategies across time changed to reflect more novelty (in the first 60 moves of a game). 2/

The development of superhuman AI programs may prompt human players to break away from traditional strategies and induced them to explore novel moves, which in turn may have improved their decision-making or even enjoyment of the game. 3/

AI systems may beneficially affect human behavior.

More generally, a deep effect of AI will be to modify the way humans treat each other in "hybrid sytstems" composed of both human and AI agents, "helping humans to help themselves": theatlantic.com/magazine/archi… | @TheAtlantic 4/

Some prior papers of ours (#HNL) regarding how AI can affect how humans treat each other (within "hybrid systems" of human and AI agents):

nature.com/articles/natur… (2017, @Nature)

pnas.org/doi/10.1073/pn… (2020, @PNASNews)

sciencedirect.com/science/articl… (2020, @iScience_CP) 5/