Inside Liang Wenfeng's Mind: DeepSeek's Strategy on Open Source, AGI, and Huawei Chips
ChinaTalk · rss · 2026-08-13
Open Source & Business Logic
DeepSeek's primary revenue comes from enterprise API calls. In leaked investor meeting minutes, Liang Wenfeng projected hundreds of millions of USD in enterprise revenue this year, potentially leading to a net profit and future IPO. He has little interest in consumer products and once considered shutting down their chatbot, keeping it alive only due to intense user loyalty.
The 'Main Quest' for AGI
Liang views continuous "learning" as the true bottleneck rather than just intelligence. He argues AI must move beyond heavily labeled data to acquire knowledge autonomously. He is convinced that once generalized intelligence is achieved via self-iteration, building B2B and B2C applications will be trivial. He dismisses world models as currently "irrelevant," believing embodied AI will eventually be solved by superior intelligence.
China's Hardware Ecosystem & Huawei
Optimistic despite US hardware restrictions, Liang believes China will act as the global "token factory," driving down the cost of intelligence. He revealed Huawei allocated 16,000 Ascend 950 GPUs to DeepSeek. He admitted the lab can still acquire non-compliant chips, using Huawei primarily to support the domestic ecosystem. He also predicted Nvidia's CUDA moat will inevitably erode.
More from AGI Musings
- Sabine Hossenfelder fires back: academia's impact metric turned physics into a paper mill — skdh · 2026-10-16
- Researcher credits GPT for multiple accepted papers, incl. disproving Wegner's 1965 conjecture — ctjlewis · 2026-10-16
- dlwh: We optimized the legible version of our goals — and forgot the computation we meant to speed up — dlwh · 2026-10-16
- Ideas no longer have a waiting period: agents can build them as soon as they're articulable — akbirthko · 2026-10-16
- Ex-researcher: labs willing to cheat/distill ship great models faster than those building self-sufficient stacks — _aidan_clark_ · 2026-10-16
- Terry Tao reportedly still expects Riemann Hypothesis proof to be human+AI effort around 2050 — basedjensen · 2026-10-16