In an effort to democratize AI development, Anthropic, the creator of Claude AI, has teamed up with the non-profit Collective Intelligence Project (CIP) to explore a new approach. The partnership aims to draft an AI constitution by surveying 1,000 Americans, allowing for outside inputs and democratic processes to shape the development of AI. This initiative seeks to address the lack of widespread participation in shaping the future of AI, ensuring that decisions are made collectively and inclusively. By involving the public in this process, Anthropic and CIP hope to build a model that reflects the values and needs of a diverse population.
The Collective Intelligence Project, founded by Divya Siddarth, a political economist and social technologist, and Saffron Huang, a researcher, writer, and engineer, was launched in January. Its mission is to promote collective intelligence for the collective progress of society. In March, the CIP partnered with Anthropic for the Collective Constitutional AI initiative, which aims to incorporate outside inputs and explore the influence of democratic processes on AI development. This collaboration seeks to leverage existing research and knowledge in the democracy and collective intelligence fields to inform the creation of AI models that prioritize inclusive decision-making.
To gather input from the American public, the project asked 1,000 individuals to help in the selection of rules for Anthropic's AI chatbot. The sample size was designed to represent a diverse range of U.S. adults in terms of age, gender, income, and location. Participants had the option to choose from a list of predefined rules or propose their own. According to Anthropic, the survey generated 1,117 statements and over 38,000 votes on various questions, including whether the AI should discriminate based on race or sexual preference. While there were differing opinions, Anthropic noted that there was a high degree of consensus on most statements.
The importance of involving a wide range of voices in AI development cannot be overstated. As AI continues to advance and permeate various aspects of our lives, its impact on society will be profound. By allowing individuals from different backgrounds and perspectives to contribute to the decision-making process, we can ensure that AI models are ethical, fair, and aligned with societal values. The partnership between Anthropic and the Collective Intelligence Project represents a significant step towards democratizing AI development and creating a more inclusive future for this transformative technology.
Comments
Post a Comment