Exploring Laws of Robotics: A Synthesis of Constitutional AI and Constitutional Economics

摘要

Abstract As today’s world increasingly harnesses ever more powerful AI systems, policymakers and developers must recognize the need for effective regulatory frameworks to ensure that the underlying LLMs are used ethically and responsibly. Integrating ordoliberal constitutional economics with AI ethics helps to create such frameworks through system prompts, reinforcement learning, and non-fine-tunable learning. Constitutional AI aims to embed ethical principles and robust safeguards into AI systems to ensure they operate within pre-defined boundaries, prioritizing safety, legality, and human rights. By embedding ethical considerations and compliance requirements directly into the operational core of AI systems, a focus on regulation and transparency of system instructions can proactively shape AI outcomes. Such a system prompt approach is consistent with ordoliberal ideals and offers a preventive strategy to ensure that AI technologies operate responsibly from the outset, as we demonstrate with an empirical experiment involving Llama 2. However, it requires democratic legitimacy via so-called “mini-publics” and ongoing research to ensure that AI systems adhere closely to these imperatives.