首页|Fine-tuning of artificial intelligence managers' logic in a supply chain with competing retailers
Fine-tuning of artificial intelligence managers' logic in a supply chain with competing retailers
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Wiley
Today, with the advance of artificial intelligence, companies in the real world are using AI as managers to make operational decisions, who can respond quickly to market shocks and whose logic can be fine-tuned to programmed pessimism/optimism, that is, underestimating/overestimating the market. The introduction of AI managers poses new challenges to supply chain management, and how to manage AI managers warrants further exploration. We investigate the optimal AI manager fine-tuning strategies in a supply chain consisting of one manufacturer and two competing retailers, each operated by an AI manager in the face of an uncertain market shock. We establish the manufacturer-retailer AI manager fine-tuning game, where the manufacturer and two retailers endogenously decide whether to fine-tune their AI managers' logic. The market may suffer an uncertain shock, and once the shock occurs, the AI managers' logic settings and price decisions can be quickly adjusted. We find that the manufacturer would never fine-tune the AI manager, while the retailers may fine-tune their AI managers to programmed optimism. Notably, AI manager's fine-tunability only benefits the retailers and harms the manufacturer, entire supply chain, consumers, and social welfare. To make AI manager's fine-tunability beneficial to all participants, that is, to reach a win-win-win situation, we design two incentive mechanisms, retailer pessimism incentive mechanism and mutual pessimism incentive mechanism (MPIM), where MPIM can lead to the win-win-win situation. Further, we endogenize the compensation, endogenous retailer pessimism compensation and endogenous mutual pessimism compensation, both achieving the win-win-win outcome. We also make several extensions and provide suggestions for supply chain firms to fine-tune their AI managers' logic.
NETL
NSTL
Wiley
