Problem: Closed Weights and Guild Monopolies (abdullah417161.substack.com)

A recent commentary critiques the dominant corporate narrative surrounding large language models (LLMs), arguing that it employs misleading tactics, such as concept substitution and demagoguery, to justify restrictive practices in the AI space. The piece outlines three main fallacies: labeling algorithms as threats under the guise of safety, asserting intellectual property claims over knowledge derived from publicly available data, and misrepresenting liability to enforce moral surveillance. The author draws parallels to historical attempts in medieval Florence to restrict access to arithmetic, highlighting that these modern corporate moves are, in essence, a means of monopolizing information and technology. This analysis carries significant implications for the AI/ML community, emphasizing the need for equitable access to computational resources and a rejection of corporate monopolies over foundational knowledge. The text underscores that efficient algorithms, like those used in LLMs, should remain accessible to foster innovation and competition. By framing ethical concerns as mere tools for control, the piece calls for a critical reassessment of how AI technologies are deployed and governed, highlighting the potential risks of stifling individual creativity and progress in the field.