You Can't Critique AI You Haven't Built With (karozieminski.substack.com)

Recent discussions in the AI community emphasize the importance of hands-on experience in critiquing AI systems effectively. Leading researchers argue that to provide meaningful feedback or improvements to AI technologies, one must have firsthand experience in building and training these systems. This perspective highlights a growing consensus that the complexities of AI and machine learning require a deep understanding of the underlying mechanics, algorithms, and data interactions. This insight is significant for AI and machine learning development, as it underscores the value of practical expertise in fostering meaningful discourse around ethical AI practices, model transparency, and accountability. By advocating for developers and critics alike to immerse themselves in the technical aspects of AI, the field moves toward a more robust evaluation process. This approach not only aims to enhance system design but also seeks to bridge the gap between creators and users, ultimately leading to more responsible and informed AI applications. Engaging more stakeholders in the development process could pave the way for innovative solutions that address not only technical challenges but also ethical and societal implications associated with AI deployment.