Science and Technology Daily reporter Zhang Mengran
A proof-of-concept study published in the British magazine "Nature·Human Behavior" found that the artificial intelligence algorithm may be able to propose a new mechanism for allocating resources among the crowd.
For many years, there has been disagreement among philosophers, economists, and political scientists about how the benefits of human cooperation should be distributed.
AI's resource allocation decisions are more popular. Picture source: "Nature·Human Behavior" online
This time, Christopher Summerfield, a scientist at the British "Deep Mind" company (Deep mind), and his colleagues trained an artificial intelligence (AI) system to design a new method for the redistribution of public goods. mechanism. The research team first asked thousands of volunteers to participate in an investment game in groups of four. In the game, each person will get different amounts of money and need to decide whether to keep the money for themselves or share it for the benefit of the entire group and let the shared funds return to themselves with interest. The
research team next trained the AI system to find a policy that reallocated funds to individuals, with the policy's popularity determined by the votes of human players when deciding which policy to choose to play again. The AI system successfully found a policy that received more votes than the baseline policy, such as redistributing funds equally to everyone or returning them in proportion to individual contributions. When the research team asked other human volunteers to serve as reassignment deciders, none of their strategies were as popular as the AI's.
Although the study focused on a special version of the public goods game involving four people, the researchers believe that future research can continue to expand the concept of AI and study whether it is effective for larger groups and more complex game scenarios.
Editor: Liu Yiyang
Review: Yue Liang
