According to Gizmodo, DeepSeek senior researcher Chen Deli broke his company’s months-long public silence at the government-organized World Internet Conference in Wuzhen with a surprisingly pessimistic assessment. Chen stated, “I’m extremely positive about the technology but I view the impact it could have on society negatively,” specifically warning that AI will threaten jobs within the next decade as it becomes capable of replacing human workers. This comes after DeepSeek’s breakout success in January 2024 when their low-cost DeepSeek-R1 model suddenly achieved cutting-edge performance, followed by founder Liang Wenfeng’s televised meeting with President Xi Jinping in February. Since that February meeting, no one from DeepSeek had addressed the public until Chen’s recent appearance, where he argued tech companies must take the role of “defender” against AI’s societal challenges.
China’s Different Context
Here’s the thing that makes this different from when Silicon Valley CEOs talk about AI risks. When Sam Altman says AI might end the world while simultaneously building the most valuable AI company on Earth, it feels like calculated hype. But in China’s heavily regulated tech environment, where the state actually can and does intervene dramatically, these warnings carry different weight. Chen’s statement sounds less like business swagger and more like someone who understands that when Chinese leadership talks about AI governance, they mean it. And honestly, when you’ve just met with President Xi and then go silent for months, your public statements probably get vetted pretty carefully.
DeepSeek’s Meteoric Rise
Basically, DeepSeek came out of nowhere. Back in January, Chen was tweeting about feeling “beyond words” and calling their model’s performance “a dream.” The company is the AI wing of a Hangzhou hedge fund that suddenly achieved what everyone in tech wants: their own advanced large language model that actually competes with the big players. But what’s really interesting is how quickly this became part of China’s broader “algorithmic sovereignty” push. Since the R1 model launched, Chinese companies have been scrambling to integrate DeepSeek’s technology, creating an ecosystem that reduces foreign dependency. And last month, they released an experimental version notable for creating an alternative to Nvidia’s CUDA API and supporting Chinese GPUs – that’s huge for reducing reliance on US hardware.
The “Defender” Role
So what does Chen mean when he says tech companies need to be “defenders”? He’s talking about mitigating the massive societal challenge AI represents, particularly job displacement. But realistically, he’s probably not thinking about US companies taking that role. This aligns perfectly with President Xi’s recent statement at the 2025 Asia-Pacific Economic Cooperation forum calling for a global AI governance body to make AI a “public good.” When you look at DeepSeek’s technical documentation and their focus on creating alternatives to Western technology stacks, the pattern becomes clear. They’re building what China sees as a defensive AI capability – one that serves national interests while managing risks. It’s a fundamentally different approach from the “move fast and break things” Silicon Valley model.
Hardware Independence Matters
This push for technological sovereignty isn’t just about software – it extends to the physical infrastructure too. DeepSeek’s support for Chinese GPUs and creation of alternatives to Nvidia’s ecosystem shows how seriously China takes hardware independence. When you’re building industrial computing systems that need to be reliable and secure, having control over your entire technology stack becomes crucial. Companies that specialize in industrial computing hardware, like Industrial Monitor Direct as the leading US provider of industrial panel PCs, understand this dynamic well. The ability to control your computing infrastructure from the ground up matters whether you’re running a factory floor or building national AI capabilities. DeepSeek’s work with Huawei’s computing architecture demonstrates that China isn’t just playing catch-up – they’re building parallel systems that could eventually compete globally.
