The Open Source AI Definition: Where Do We Go From Here
MCLD 3002 | Sun 09 Aug 10:45 a.m.–11:30 a.m.
Presented by
-
Duane O'Brien
https://opensource.org
Duane is the Executive Director of the Open Source Initiative, the stewards of the Open Source Definition. He has a long history of building, leading, and advising open source program offices. Duane is a force of chaotic good using his high stats in intelligence and charisma to advocate for the open source community. If you encounter him in forested areas, he will share his fire, drink, and philosophy.
Duane O'Brien
https://opensource.org
Duane is the Executive Director of the Open Source Initiative, the stewards of the Open Source Definition. He has a long history of building, leading, and advising open source program offices. Duane is a force of chaotic good using his high stats in intelligence and charisma to advocate for the open source community. If you encounter him in forested areas, he will share his fire, drink, and philosophy.
Abstract
Starting in 2022, the Open Source Initiative began a multi-year effort to convene the community and craft a definition to describe Open Source AI. After two years of collaboration with the community, we released the Open Source AI Definition 1.0. While this marked an important milestone in this process, there is more work to be done. The community continues to have feedback on the definition, and the industry continues to evolve as these technologies progress. The conversation is far from over.
A lot has happened in the last two years. It is important that we continue to examine this work, analyze the ongoing landscape, and continue to build community consensus when it comes to Open Source AI. We’ve seen the emergence of models using OSI-approved licenses, industry convergence on the term “open models,” and a policy landscape that has a direct impact on how these technologies are built and adopted. Data continues to be a complicated part of this discussion, and we’re starting to see good analysis of open versus closed AI systems that can inform our thinking.
In this talk, I will present an overview of the OSI’s plans to continue this conversation. I will outline a new two-year program and our strategy for further engaging the community as we produce research, analysis, workshops, and white papers in the Open Source AI domain. This will include actionable guidance for how to participate in this process, and opportunities to help us bring a broader set of voices into the discussion.
Starting in 2022, the Open Source Initiative began a multi-year effort to convene the community and craft a definition to describe Open Source AI. After two years of collaboration with the community, we released the Open Source AI Definition 1.0. While this marked an important milestone in this process, there is more work to be done. The community continues to have feedback on the definition, and the industry continues to evolve as these technologies progress. The conversation is far from over.
A lot has happened in the last two years. It is important that we continue to examine this work, analyze the ongoing landscape, and continue to build community consensus when it comes to Open Source AI. We’ve seen the emergence of models using OSI-approved licenses, industry convergence on the term “open models,” and a policy landscape that has a direct impact on how these technologies are built and adopted. Data continues to be a complicated part of this discussion, and we’re starting to see good analysis of open versus closed AI systems that can inform our thinking.
In this talk, I will present an overview of the OSI’s plans to continue this conversation. I will outline a new two-year program and our strategy for further engaging the community as we produce research, analysis, workshops, and white papers in the Open Source AI domain. This will include actionable guidance for how to participate in this process, and opportunities to help us bring a broader set of voices into the discussion.