Prohibit Cannot Regulate: On the Question of Adjusting the Use of AI in Everyday Life of Society

Abstract

This article examines the framework governing artificial intelligence (AI) instruments that are becoming an integral part of society’s everyday life. The core question framing authors’ reflections is: Should AI technologies be prohibited or regulated from the outset?

The paper begins with a general overview of the central question surrounding the development of artificial intelligence (AI), setting the stage for a deeper exploration of critical issues. It examines specific challenges related to AI, including deceptive practices that may arise in machine learning algorithms, the functionalities and implications of autonomous agents in various sectors, and the concept of the "second transition," which refers to the developing state of AI capabilities and their integration into society. Throughout this discussion, the authors highlight the substantial regulatory challenges that arise from these advancements. Building on this foundation, the authors then analyze the existing regulatory context in detail, emphasizing the distinct roles that developers, consumers, and regulators play in the oversight of AI technologies. They discuss the responsibilities of developers to ensure ethical practices in AI design, the need for informed consumers to navigate complex AI systems, and the imperative for regulators to establish effective frameworks that keep pace with rapid technological changes. Additionally, the paper explores the valuable contributions that social scientists can make in shaping AI regulations by bringing insights into social behavior, ethical considerations, and public policy. This collaborative approach is essential for developing robust regulatory mechanisms that address both the opportunities and challenges of AI integration into society. In conclusion, the authors outline four regulatory principles: 1) The key is not to regulate AI’s design and manufacture but the relationships among those who create, deploy, and utilize it; 2) The focus of AI regulation should not be on its making but on addressing the negative effects of using AI tools in everyday life; 3) In the first stage, regulations should focus on the development of AI tools in specific areas where the benefits of replacing humans with AI are most apparent; 4) There is a fundamental difference between regulating AI in areas that relate to existential issues (where we need to introduce prohibitions right now) and other areas (where we should first regulate and then prohibit).