Ex-Meta engineer says Anthropic's Scion closely mirrors Mononoke source control design

GoKawiil's interpretation of the reporting above, not reported fact.

The claim raises questions about how AI companies and their models attribute prior engineering work when designing new systems, since Anthropic reportedly framed Claude as having arrived at these design choices independently. The author suggests that because open-source projects like Mononoke sit in LLM training data, AI-assisted design work may increasingly echo existing architectures without clear acknowledgment, making proper credit practices more important going forward.

• A former Mononoke engineer alleges Anthropic's Scion closely replicates Mononoke's 'derived data' approach to efficient blame operations.
• Anthropic credited Meta's Sapling for 'hard-won lessons' but did not mention Mononoke, which the author calls close to plagiarism.
• The author argues that as LLMs assist system design, they may heavily reuse ideas from open-source projects in their training data without adequate attribution.