DeepSeek Is Not an Idealistic Company, It Simply Subordinates Reality to Ideals
DeepSeek
DeepSeek founder Liang Wenfeng explained the company's strategy at an investor meeting: focus on AGI with a phased plan (language models, chain-of-thought, agent, self-learning, embodied intelligence), reject trendy directions (video generation, 3D) to stay on the main track, maintain a flexible organization without KPIs, emphasize team stability, and adopt open-source as a business model. The company does not seek monopoly, is willing to help competitors, and aims for a reasonable profit margin.
At an investor meeting, DeepSeek founder Liang Wenfeng detailed the company's strategy, which combines the long-term goal of AGI with pragmatic steps. The roadmap includes language models, Chain-of-Thought, Agent, continuous learning, self-iteration, and embodied intelligence. The company consciously avoids directions not on the AGI main path, such as video generation and 3D, even if they are commercially profitable. The organization is built on vision rather than KPIs: employees spend half their time on their own research (the "lottery"), and the company does not encourage overtime. Team stability is considered a key asset, and turnover is low. DeepSeek adheres to an open-source strategy, believing it does not harm the business because the real barriers are deployment cost and efficiency. The company does not seek monopoly, is willing to cooperate with competitors (e.g., Alibaba, Zhipu, Moonshot AI), and hopes to carve out its niche in the larger AI ecosystem. Pricing aims for 6x profit with equipment payback in 10 months, rather than profit maximization—Liang Wenfeng believes that those who want a lot lose to those who settle for little.
Source: 36Kr —
original