This study is an effort to provide a behavioral analysis and social simulations on the liability rules and their effect on behavior and social efficiency, in the consideration of artificial intelligence (AI). AI can be categorized into weak AI and strong AI, according to the extent of autonomy and capabilities. This study will cover their legal definitions and limitations, with the discussion of self-driving AI as an example, as well as investigating the current autonomous and agency theories, for the purpose of coding AI-related regulations and criteria for systematic analysis.
By utilizing Agent-Based Modeling, this study uses different rules and criteria as variables in the economic analysis and social simulation, such as levels of liability rules, data protection, and AI controllability. Through simulating how agents (including customers and producers) act in different types of jurisdictions, this study is intended to explore how these variables affect the behavior of AI-related parties, and how to achieve better fitness to a particular society or improve efficiency. In addition, this study will also provide a discussion on whether such results can be applied to strong AI, under the consideration of threshold effect, enforcement error, joint and several actions, contract stages, and so on. We aim to explore the aforementioned issues with behavioral and empirical perspectives, and to offer suggestions, in the hope of promoting the advancement of AI, and at the same time, ensuring the general welfare of the public.

Authors:
Jing-Huey Shao, National Cheng Kung University, Taiwan

---

Conference Comments & Feedback

Place a comment using your LinkedIn profile

Comments

Share on activity feed

Post with LinkedIn

Powered by WP LinkPress

Share this Presentation

Share this...

Facebook

Pinterest

Twitter

Linkedin

Posted by Clive Staples Lewis

Last updated: 2023-02-23 23:45:00

All Posts