Quick Buzz Feed
With a grant from OpenAI, a team led by Penn State researchers proposed a democratic decision-making system that uses decentralized voting and discussion tools to give people with different perspectives a voice in how artificial intelligence is used.
Artificial intelligence (AI) systems are pervasive in daily life, impacting healthcare, education, and public policy. However, the public has limited opportunities to participate in their development or governance, leading to AI outputs that may not align with the diverse needs and values of communities.
Led by Penn State researchers, with support from OpenAI and other universities, InclusiveAI is a proposed system inspired by democratic principles. It employs decentralized voting and discussion tools to empower individuals with varied perspectives to influence how AI is utilized, addressing the lack of public input in AI development.
Existing methods for aligning AI with human values or governing AI models are largely top-down, controlled by major tech corporations. While surveys collect opinions, they lack the interactive discussion, mutual learning, and shared decision-making necessary to make AI tools genuinely fairer and more inclusive across different cultures and backgrounds.
To foster transparent governance, the researchers propose using Decentralized Autonomous Organizations (DAOs). These blockchain-based groups utilize smart contracts to facilitate collective decision-making, gathering public preferences and guiding AI improvements. The DAO process promotes discussion, compromise, and a deeper understanding of diverse perspectives to build more inclusive and balanced AI systems.
In practice, a DAO system would offer a deliberation forum where users and developers can discuss alternative viewpoints and vote on proposed changes, ensuring transparency and immutability. This democratic approach has potential applications in critical sectors like medicine, law enforcement, and community safety, enabling experts, policymakers, and regulators to contribute to design and improvement discussions. The goal is to balance majority opinion with the protection of minority viewpoints, promoting a 'digital democracy' for AI.
Penn State is actively shaping the future of higher education with a focus on human-centered, ethical AI innovation. Through strategic planning, partnerships, and investments, the university aims to equip its students, staff, and faculty with the necessary AI knowledge, experience, and confidence to thrive in an AI-powered future, ensuring meaningful impacts for the community.