Meta challenges OpenAI and Google with open-source AI
Meta has launched an updated, open-source version of its Llama artificial intelligence model, marking a strategic shift from offering proprietary systems to providing accessible tools for developers. By making the model’s underlying code available for modification, Meta aims to foster widespread innovation, allowing the broader community to improve upon and customize the technology. This approach contrasts sharply with competitors who keep their models closed, prioritizing transparency and collaborative development over exclusive control. The release emphasizes enhanced safety and security, arguing that open code enables more experts to scrutinize the system for potential issues and vulnerabilities. This transparency addresses growing concerns regarding the reliability and safety of closed AI models, which have faced criticism for generating misinformation or behaving unpredictably. By inviting public review, Meta seeks to build trust and ensure the technology remains robust and dependable for diverse applications. This move is highly relevant to open_data as it challenges the dominance of proprietary AI ecosystems by promoting data and code accessibility. It demonstrates how open-source frameworks can drive competition and offer businesses greater control over their data, avoiding the limitations of black-box solutions. This supports the open_data principle that transparency and shared access to technology infrastructure lead to better innovation, security, and user empowerment.
Source: philstar.comPublished on 2023-07-20