Thursday, March 16, 2023

Superhuman artificial intelligence can improve human decision-making by increasing novelty

This is very promising.  New knowledge depends of identifying novelty and examining it when it arises.  Still need memory but not a trained memory forcing past retention into the mix.  The problem has always been real access to novelty.

That is why i read thousands of words every day since i was nine.  You need to be able to access the data and then allow you brain to decide by checking the future if you will ever need it.

That, by the way is the prime limitation of AI.  As a trained human brain, I am able to subconsiously remember the future.  I do not see AI doing that.

Superhuman artificial intelligence can improve human decision-making by increasing novelty

Edited by Michael Gazzaniga, University of California Santa Barbara College of Letters and Science, Santa Barbara, CA; received August 31, 2022; accepted December 19, 2022

March 13, 2023
120 (12) e2214840120

Although advances in artificial intelligence (AI) created superhuman AI systems, little is understood about how such AI systems will affect human decision-making. We examine historical changes in decision-making by professional Go players over the recent seven decades, focusing on changes after the advent of superhuman AI (e.g., AlphaGo). We find that superhuman AI may have improved human decision-making, and that this improvement was associated with increased novelty in decision-making as human players were encouraged to make decisions previously unobserved in history. Our findings illustrate that superhuman AI can encourage novel decision-making by humans in certain domains and suggest that innovative thinking can spread from machines to humans and among humans themselves, possibly improving human decision-making in those domains.


How will superhuman artificial intelligence (AI) affect human decision-making? And what will be the mechanisms behind this effect? We address these questions in a domain where AI already exceeds human performance, analyzing more than 5.8 million move decisions made by professional Go players over the past 71 y (1950 to 2021). 

To address the first question, we use a superhuman AI program to estimate the quality of human decisions across time, generating 58 billion counterfactual game patterns and comparing the win rates of actual human decisions with those of counterfactual AI decisions. 

We find that humans began to make significantly better decisions following the advent of superhuman AI. We then examine human players’ strategies across time and find that novel decisions (i.e., previously unobserved moves) occurred more frequently and became associated with higher decision quality after the advent of superhuman AI. 

Our findings suggest that the development of superhuman AI programs may have prompted human players to break away from traditional strategies and induced them to explore novel moves, which in turn may have improved their decision-making.

No comments: