Fostering Responsible AI:Empowering Openness and Community Collaboration
summary
Generative AI Commons presents the Model Openness Framework (MOF) that inspects 16 artifacts of a model lifecycle and ranks models into class one/class two/class three based on whether code, checkpoints, metadata, evaluation data, training data, licenses and docs are openly available. The community of 120 active members from 60 organizations builds a scripted 16 components checklist that emits a JSON manifest, assigns a badge/scoreboard, and requires OSI/CDLA/CC licenses per component; ongoing evaluations of top models are manually labeled. Bias and fairness remain out of scope.
AI-written summary · about
chapters (7)
- 0:00Introduction to Generative AI Commons
- 4:20Need for Responsible AI
- 9:05Understanding Open Principles
- 13:58Open AI and Licensing Challenges
- 18:59Model Openness Framework
- 26:22Call to Action and Next Steps
- 27:29Q&A and Community Engagement
related talks
41mThe Generative AI Commons: Unlocking Potential Through Collaboration and Transpare...
22mThe Role of Openness in Responsible AI
40mThe Generative AI Commons: Unlocking Potential Through Collaboration and Transpare...
40mUnleashing Collective Genius: Building GenAI Through Open Collaboration
34mThe Model Openness Framework: Promoting Completeness and Openness for Reproducibil...
26mIntroducing the Model Openness Framework: Achieving...
42mThe Open Source AI Dilemma: Crafting a Clear Definition for Open Source AI
37mIs Your AI Truly Open ? Cutting Through the Hype