In a rare four-hour investor talk, DeepSeek founder Liang Wenfeng laid out his vision for AI, emphasizing that AGI is too large to be monopolized by any single company. He framed AI as a historic force or “tide” that should be approached with restraint, openness, and a willingness to earn only fair profits rather than maximize short-term gains. Liang said DeepSeek’s original goal was not money, IPOs, or capital markets, but to build something useful for humanity.
He argued that company vision, not rigid organization or KPIs, is what holds DeepSeek together. Open source, he said, is both a moral choice and a commercial advantage in AI because the market is so large that closed, monopolistic strategies will eventually fail. In his view, companies that try to capture excessive profits will be outcompeted by those satisfied with moderate returns.
Liang also discussed the AI race between China and the U.S. He believes America’s lead comes mainly from greater compute resources, not a lasting talent advantage. He expects top U.S. model leaders such as OpenAI, Anthropic, and Google to remain in a cyclical competition rather than maintain permanent dominance. He said China’s advantages lie in cost efficiency, user experience, and the opportunity to build a new chip ecosystem as Nvidia’s CUDA moat weakens.
Looking ahead, Liang predicted a smaller set of base-model companies, with only three or four fully competitive players. DeepSeek will focus on core AGI paths such as language models, reasoning, agents, and continual learning, while leaving areas like video generation and world models to the broader ecosystem.
Article timeline
Build this article’s timeline
Extract dated events, key actors, and open questions using the same timeline analysis available in Signal.
In a rare four-hour investor talk, DeepSeek founder Liang Wenfeng laid out his vision for AI, emphasizing that AGI is too large to be monopolized by any single company.
🤖 AI prefill
📅WhenWhen did this happen, or what is the timeframe?
📍Where
China, United States, and America
🤖 AI prefill
💡Why
Centers on open-source models, AGI, and compute
🤖 AI prefill
Login to answer and earn XP for each gap you fill.
Comments
Login to view and post comments.