Key points: Chen Deli of DeepSeek warned at a Wuzhen panel that AI could eliminate most jobs within 10–20 years and urged tech firms to act as "guardians of humanity." He described a short-term "honeymoon phase" when AI boosts productivity, but predicted rising capability in 5–10 years that could trigger significant workforce disruption. DeepSeek — known for its low-cost R1 model — has become a focal point in U.S.-China AI competition, a dynamic that influenced moves such as OpenAI's GPT-oss release.
DeepSeek Researcher Warns AI Could Eliminate Most Jobs in 10–20 Years — Urges Tech Firms to Be 'Guardians of Humanity'
Key points: Chen Deli of DeepSeek warned at a Wuzhen panel that AI could eliminate most jobs within 10–20 years and urged tech firms to act as "guardians of humanity." He described a short-term "honeymoon phase" when AI boosts productivity, but predicted rising capability in 5–10 years that could trigger significant workforce disruption. DeepSeek — known for its low-cost R1 model — has become a focal point in U.S.-China AI competition, a dynamic that influenced moves such as OpenAI's GPT-oss release.
09:55 PM, 11/10/2025Technology
DeepSeek researcher warns of long-term AI risks and calls for corporate responsibility
Chen Deli, a senior researcher at Chinese AI firm DeepSeek, made a rare public appearance at the World Internet Conference in Wuzhen and warned that artificial intelligence could replace a large share of jobs within the next 10 to 20 years.
Chen said the near term looks positive — a "honeymoon phase" in which AI amplifies human productivity because systems cannot yet perform tasks entirely on their own. But he cautioned that as AI capabilities grow, the balance between benefits and risks could flip.
"During this period, societal structures will also be greatly challenged," Chen said. He urged tech companies to "play the role of guardians of humanity, at the very least protecting human safety and helping to reshape social order."
Chen predicted that within five to 10 years AI systems will become substantially more capable, potentially forcing widespread workforce adjustments and layoffs. He argued that technology firms should sound the alarm — acting as whistleblowers to warn society about potential disruptions and to help manage the transition.
DeepSeek, founded in 2023 and based in Hangzhou, has kept a low public profile since its breakout earlier this year. Its CEO, Liang Wenfeng, has not been seen publicly since a televised meeting with China's leader Xi Jinping in February. Onstage in Wuzhen, Chen shared the panel with executives from five other firms collectively dubbed China's "six little dragons" of AI, including Alibaba Cloud and robotics company Unitree.
The company drew global attention in January when it unveiled R1, a low-cost reasoning model that DeepSeek says matches leading rivals — including OpenAI's o1 — at a fraction of the cost. That announcement prompted international discussion about the narrowing gap between Chinese and U.S. AI capabilities.
In March, OpenAI told the U.S. government that while America still leads in AI, "DeepSeek shows that our lead is not wide and is narrowing." In August, OpenAI CEO Sam Altman released GPT-oss, a family of models with "open weights" — a move analysts said was likely influenced by momentum among open-source Chinese models. Ray Wang of Futurum Group noted that GPT-oss has "narrowed the gap" with Chinese models on performance benchmarks and model sizes.
Why it matters: Chen's comments highlight the growing debate about how to manage rapid AI progress — balancing short-term productivity gains with longer-term social and economic risks. He called on tech companies to be transparent about those risks and to help guide policy and societal responses.