Overnight, the VC and industry circles were all sharing the four-hour investor meeting speech by DeepSeek's boss, Liang Wenfeng. It felt like a pilgrimage. I’ve compiled some of the most quoted remarks for everyone: • Humans can continuously learn, but for the same task, you have to provide AI with all the context, which is almost impossible. So, AI cannot replace employees. • For a normal person, their need isn’t a computer, but manpower. • Anthropic has surpassed OpenAI for now, but it’s not long-term—it’s just a phase. OpenAI and Google will likely alternate in leading the future. • There are too many companies in China working on models, and every single one is doing the same thing. Eventually, it will converge, but it will take time. • We generally don’t work overtime. First, research requires a relatively relaxed environment. Second, we are very focused. Many of our products are far from perfect, but we don’t rush to fix them. It’s a kind of restrained culture. • Open source has its benefits. AI is big enough that it might eventually account for 10% of human society’s GDP. If we try to monopolize this benefit, we will inevitably be abandoned by history. This is an objective law, a historical perspective. • Selling APIs isn’t that attractive. You just need a few people to maintain the API, and you don’t even need customer service—users will figure it out themselves. • I hope the next generation of models can help us with our own development. To put it simply, the primary goal of the models we’re building isn’t for everyone to find them useful, but for us to find them useful. This is the fastest path to achieving AGI.