Open-source AI must reveal its training data, per new OSI definition
The Open Source Initiative has defined "open" AI, requiring disclosure of training data and code, challenging companies like Meta, whose Llama model does not comply, amid ongoing debates on open-source values.
Read original article
The Open Source Initiative (OSI) has established a new definition for "open" artificial intelligence, which mandates that AI systems disclose their training data, code, and settings. This definition poses a challenge to companies like Meta, whose Llama model does not meet these criteria due to restrictions on commercial use and lack of access to training data. Meta has expressed disagreement with OSI's definition, arguing that it oversimplifies the complexities of modern AI. The OSI's definition aims to combat "open washing," where companies falsely claim their products are open source. The initiative took two years to develop, involving consultations with experts in various fields. Critics suggest that Meta's reluctance to share training data stems from a desire to protect its competitive edge and minimize legal risks associated with copyright issues. OSI's executive director noted parallels between Meta's current stance and Microsoft's historical resistance to open-source principles. The ongoing debate highlights the tension between traditional open-source values and the evolving landscape of AI technology.
- OSI's new definition requires AI systems to disclose training data, code, and settings.
- Meta's Llama model does not comply with OSI's open-source standards.
- The definition aims to prevent companies from misrepresenting their products as open source.
- Critics argue that Meta's data restrictions are motivated by competitive and legal concerns.
- The discussion reflects broader tensions between open-source principles and AI development.
Related
Begun, the open source AI wars have.. This is going to be ugly. Really ugly.
The Open Source Initiative is finalizing a definition for open source AI, facing criticism for potentially allowing proprietary systems to claim open source status, with ongoing debates expected for years.
The OSI lacks competence to define Open Source AI
The Open Source Initiative faces criticism for its handling of the Open Source AI Definition, with concerns over expertise, censorship, and transparency, as it approaches a deadline of October 28, 2024.
The open secret of open washing – why companies pretend to be open source
Open washing involves companies falsely claiming their products are open source, particularly in AI. This threatens collaboration, increases legal issues, and hinders innovation, highlighting the need for clear definitions.
Meta under fire for 'polluting' open-source
Meta's labeling of its Llama AI models as "open-source" has drawn criticism for being misleading, as they do not fulfill full open-source criteria, prompting calls for greater transparency in AI development.
Open-Access AI: Lessons from Open-Source Software
Open-access AI models, like Meta's Llama, impose usage restrictions, misleadingly labeled as "open-source." Access to training data is essential for innovation, raising concerns about monopolistic control in AI advancements.
4 commentsI guess I’m built different, but all this open-washing noise makes no sense to me. I don’t think “oh, open-free as in both” every time I see “open” in the wild from a bigcorp. Especially in the area that just emerged and is clearly nuanced in the “source” part. I mean, yes, it’s not fully correct, but also everyone to whom this matters figures it out sorta immediately. What is even washing here? These semantic arguments are the least problem we’ll have with all that power concentration, that is assuming the current tech is worth anything outside of its bubble.
In practice the key criteria seem to be:
1) Can I get the weights
2) Is commercial use permitted
More nuances sure but if those are met then many consider it open in the non Stallman sense
Related
Begun, the open source AI wars have.. This is going to be ugly. Really ugly.
The Open Source Initiative is finalizing a definition for open source AI, facing criticism for potentially allowing proprietary systems to claim open source status, with ongoing debates expected for years.
The OSI lacks competence to define Open Source AI
The Open Source Initiative faces criticism for its handling of the Open Source AI Definition, with concerns over expertise, censorship, and transparency, as it approaches a deadline of October 28, 2024.
The open secret of open washing – why companies pretend to be open source
Open washing involves companies falsely claiming their products are open source, particularly in AI. This threatens collaboration, increases legal issues, and hinders innovation, highlighting the need for clear definitions.
Meta under fire for 'polluting' open-source
Meta's labeling of its Llama AI models as "open-source" has drawn criticism for being misleading, as they do not fulfill full open-source criteria, prompting calls for greater transparency in AI development.
Open-Access AI: Lessons from Open-Source Software
Open-access AI models, like Meta's Llama, impose usage restrictions, misleadingly labeled as "open-source." Access to training data is essential for innovation, raising concerns about monopolistic control in AI advancements.