According to Business Insider, DeepSeek researcher Chen Deli warned at the World Internet Conference that AI could eliminate most jobs within 10-20 years, describing this as a period when “societal structures will be greatly challenged.” Chen represented the Chinese AI startup during one of its rare public appearances since its breakout earlier this year. He urged tech companies to become “guardians of humanity” and serve as whistleblowers about AI risks. DeepSeek’s low-cost R1 model, unveiled in January, has been challenging OpenAI’s dominance while matching ChatGPT’s o1 at a fraction of the cost. OpenAI acknowledged in March that DeepSeek shows America’s AI lead “is not wide and is narrowing.”
The warning shot
This is pretty significant coming from DeepSeek, which has been notoriously quiet since making waves earlier this year. Chen’s appearance at the World Internet Conference alongside other Chinese AI “little dragons” like Alibaba Cloud and Unitree suggests they’re stepping into a more public role. And his message is stark: we’re in a “honeymoon phase” right now where AI boosts productivity, but the real reckoning is coming.
What’s interesting is the timing. DeepSeek basically came out of nowhere in 2023 and immediately challenged the entire AI establishment with their R1 model. Now they’re warning about the very disruption they’re accelerating. It’s like they’re building the rocket while simultaneously telling everyone the launch might destroy the launchpad.
The competitive landscape heats up
Here’s the thing: OpenAI is clearly worried. Their March letter to the US government admitting DeepSeek is closing the gap wasn’t just corporate humility – it was a warning shot. And then in August, they released GPT-oss with “open weights,” which analysts directly linked to the success of open-source Chinese models like DeepSeek’s.
Ray Wang from Futurum Group put it bluntly: if the US doesn’t keep pace in open-source AI, Chinese models could become the default foundation globally. That’s huge. We’re talking about a potential shift in who sets the standards for the next generation of AI applications and research. For companies relying on industrial computing infrastructure, this competition could dramatically reshape the technology landscape they depend on for industrial panel PCs and other specialized hardware solutions.
The bigger picture
Chen’s warning about job elimination isn’t new, but coming from someone inside one of the companies driving this transformation gives it extra weight. He’s basically saying the companies building these systems have a responsibility to protect humanity during the transition. But can they?
Think about it: we’re looking at potentially the most dramatic workforce transformation in human history happening within a single generation. And the companies creating the technology are asking us to trust them to manage the fallout. That’s a pretty heavy burden for any organization, even ones with the technical prowess of DeepSeek or OpenAI.
The real question is whether any company, no matter how well-intentioned, can truly serve as “guardians of humanity” while simultaneously competing in a global AI arms race. It’s like asking Formula 1 teams to also be traffic safety regulators. The incentives might not always align.
