نوع مقاله : مقاله پژوهشی
نویسنده
فلسفه هنر، دانشگاه علامه طباطبائی، تهران، ایران
کلیدواژهها
عنوان مقاله English
نویسنده English
With the rapid development of artificial intelligence (AI), aligning this technology with human values has become increasingly essential. Value-Sensitive Design (VSD) has emerged as a practical approach to value alignment, attracting considerable scholarly attention. This article evaluates VSD as a method for aligning AI with social values, drawing on the pragmatist account of sociotechnical systems as the evaluative framework. Using an analytical-critical method, both strengths and challenges are examined. The findings show that VSD offers a robust framework for value alignment. Its strengths include providing a practical pathway for embedding social values into technology, incorporating stakeholder perspectives, recognizing human agency, and addressing the full life cycle of AI systems. Nevertheless, VSD faces significant challenges, such as the protracted nature of alignment processes, difficulties in ensuring meaningful human control, and the lack of enforceable international laws and regulations. Our analysis suggests that treating alignment methods merely as self-regulatory tools may produce long-term negative consequences. To fully realize the benefits of aligning AI with human values, it is therefore crucial to develop enforceable international frameworks that complement and support local laws and governance structures.
کلیدواژهها English
