Open-Weight AI Could Reshape the AI Economy

Box co-founder and CEO Aaron Levie argued that open-weight AI could strengthen the U.S. artificial intelligence ecosystem rather than weaken it. In a discussion with MTS hosts Theo Jaffee and Sofia Puccini, Levie said open models can expand AI use cases, encourage faster innovation among closed-model developers and shift competition without fundamentally changing where AI profits accrue. The discussion covered model distillation, U.S.-China AI competition and the risk that restricting access could accelerate the growth of rival AI ecosystems. Levie suggested that U.S. AI labs could release open-weight versions of older models while keeping their newest systems proprietary. Levie also discussed AI’s impact on knowledge work and software engineering at Box. He argued that companies reducing engineering headcount may be limiting their ambitions rather than responding solely to productivity gains. He expects enterprises to use multiple AI models instead of relying on a single provider. As a result, infrastructure that routes between models, enterprise data and business workflows could become increasingly valuable. For traders, the key theme is the potential growth of the AI infrastructure and software sectors. The debate could influence sentiment toward model developers, cloud providers, data platforms and AI application companies, although the discussion itself contains no direct cryptocurrency or token catalyst.