Co-Building the Future: AI Alignment for Open Source Communities
MCLD 3002 | Sat 08 Aug 10:45 a.m.–11:30 a.m.
Presented by
-
Emma Irwin
@sunnydeveloper.mastodon.social
@NA
https://sunnydeveloper.com
Emma Irwin is founder and principal consultant at Open Practice Consulting and co-chairs the CHAOSS AI Aligment for Open Source Working Group. She has worked at the intersection of openness, source, education, science and data for over 20 years, including roles with Mozilla's Open Innovation Team, Microsoft's Open Source Programs Office, and Royal Roads University. Emma co-chairs the CHAOSS AI Alignment for Open Source Working Group.
Emma Irwin
@sunnydeveloper.mastodon.social
@NA
https://sunnydeveloper.com
Emma Irwin is founder and principal consultant at Open Practice Consulting and co-chairs the CHAOSS AI Aligment for Open Source Working Group. She has worked at the intersection of openness, source, education, science and data for over 20 years, including roles with Mozilla's Open Innovation Team, Microsoft's Open Source Programs Office, and Royal Roads University. Emma co-chairs the CHAOSS AI Alignment for Open Source Working Group.
Abstract
AI is possible because open communities have collaborated and shared openly licensed code and content. AI models are dependent on thousands of open source projects to function, and have relied on countless open resources to obtain the level of intelligence we see today. One might even say that without access to openly licensed code and content, AI would lack the context of humanity altogether. Communities are on the defensive: reactively creating policies around how AI may, if at all, interact with humans or the output of their labour. We've also seen (among other reactive approaches) projects change licenses, communities change platforms, bug bounty programs shut down, policies ban AI contributions, and writers add subscription barriers to combat the unknown.
"In the field of artificial intelligence (AI), alignment aims to steer AI systems toward a person's or group's intended goals, preferences, or ethical principles. An AI system is considered aligned if it advances the intended objectives. A misaligned AI system pursues unintended objectives." - AI Alignment, Wikipedia
The CHAOSS AI Alignment for Open Source Working Group (under the Linux Foundation) is a cross-community collaboration of practitioners working to define and build solutions that give power back to communities by providing tools and metrics to help communities move from objects of consumption to owners and facilitators of their future. In this talk, we will share efforts to define "what success looks like" for AI alignment with open communities around use, consent, embedded feedback, and detection. Specifically, we will cover:
- Metric: AI Use + Consent (declaration). A metric focused on evaluating AI use and adherence to community policies.
- Metric: AI Community Feedback (measurement). A metric model that focuses on whether feedback to providers produces observable change, at what cost, and whether the exchange is proportional to the effort.
- Tool: CHAOSS AI Detection Action (inform/enable enforcement). Communities verify what is happening. The tool surfaces AI-authored or AI-assisted contributions from commit-level signals, including AI tool type and declared use.
As Shannon Vallor writes in The AI Mirror (which focuses on ethics + AI), these systems "show only where the data say that we have already been, never where we might venture together for the first time." AI reflects our creative and imperfect pasts. We hope this work can help us forge a more optimistic future with communities in the driver's seat.
AI is possible because open communities have collaborated and shared openly licensed code and content. AI models are dependent on thousands of open source projects to function, and have relied on countless open resources to obtain the level of intelligence we see today. One might even say that without access to openly licensed code and content, AI would lack the context of humanity altogether. Communities are on the defensive: reactively creating policies around how AI may, if at all, interact with humans or the output of their labour. We've also seen (among other reactive approaches) projects change licenses, communities change platforms, bug bounty programs shut down, policies ban AI contributions, and writers add subscription barriers to combat the unknown.
"In the field of artificial intelligence (AI), alignment aims to steer AI systems toward a person's or group's intended goals, preferences, or ethical principles. An AI system is considered aligned if it advances the intended objectives. A misaligned AI system pursues unintended objectives." - AI Alignment, Wikipedia
The CHAOSS AI Alignment for Open Source Working Group (under the Linux Foundation) is a cross-community collaboration of practitioners working to define and build solutions that give power back to communities by providing tools and metrics to help communities move from objects of consumption to owners and facilitators of their future. In this talk, we will share efforts to define "what success looks like" for AI alignment with open communities around use, consent, embedded feedback, and detection. Specifically, we will cover:
- Metric: AI Use + Consent (declaration). A metric focused on evaluating AI use and adherence to community policies.
- Metric: AI Community Feedback (measurement). A metric model that focuses on whether feedback to providers produces observable change, at what cost, and whether the exchange is proportional to the effort.
- Tool: CHAOSS AI Detection Action (inform/enable enforcement). Communities verify what is happening. The tool surfaces AI-authored or AI-assisted contributions from commit-level signals, including AI tool type and declared use.
As Shannon Vallor writes in The AI Mirror (which focuses on ethics + AI), these systems "show only where the data say that we have already been, never where we might venture together for the first time." AI reflects our creative and imperfect pasts. We hope this work can help us forge a more optimistic future with communities in the driver's seat.