Tim O’Reilly believes big AI labs are focusing on the wrong targets. In his view, companies like Microsoft in the 1990s are repeating history by trying to lock users into their products rather than embracing open-source alternatives.
O'Reilly sees AI as a new creative medium and is promoting efforts to open-source not just neural-net weights but the entire stack of an AI system. He fears that big AI labs are optimizing for specific use cases, which may not align with what ordinary people need.
He warns that while frontier models might be better for solving hard problems, they could overshadow more accessible tools. His nonprofit, the AI Disclosures Project, is working on open-memory consortiums to empower users and enable them to switch between different providers and models freely.
O'Reilly’s stance isn’t just about technology; it's also a critique of venture capital and the way big tech companies are shaping strategy. He envisions an open-source future where innovation thrives, painting outside the lines rather than within the confines of proprietary systems.
