On January 27th, Shaw, the founder of ai16z, pointed out in a post that competition between AI models is driving technological progress. However, the future trend is towards model open-sourcing and cost reduction, with hardware and consumer products (such as NVIDIA GPUs and Apple devices) continuing to benefit. AI agents, as a new application paradigm, can be embedded in social media, combined with intelligent models and platform connections, generate training data, and drive AI iteration.

He believes that agent development is more of an engineering problem than a mathematical problem. Currently, top companies are limited in their progress due to heavy research and light product development, while open-source ecosystems and community collaboration are rapidly advancing agent technology. Although top model training resources have huge demands, agent technology provides developers with opportunities to participate. Shaw is confident in the future of agents and believes that they will become the driving force behind practical AI applications, further popularizing them through open-source and productization.