- Liu Shengyu compared Anthropic achieving AGI first to Nazi Germany developing nuclear weapons first
- He distrusts US AI firms Anthropic and OpenAI to create open and affordable frontier AI systems
- Concerns grow about rapid AI development outpacing safety measures in the industry
A DeepSeek engineer has triggered a fresh debate about the future of AI after saying that if Anthropic were to achieve artificial general intelligence (AGI) first, it would be akin to Adolf Hitler and Nazi Germany developing nuclear weapons before the Allies. Liu Shengyu, a machine-learning systems engineer at Chinese AI startup DeepSeek, made the comparison in a blog post written in Chinese. He said he did not trust US AI companies Anthropic or OpenAI to make frontier AI systems sufficiently open and affordable.
Liu contrasted their approach with DeepSeek's open-source AI models and suggested that allowing a small number of companies to control the most advanced forms of AI could carry serious consequences, Anadolu Agency, based in Turkey reported.
His comments come as concerns grow within the AI industry over how quickly increasingly powerful systems are being developed and whether safety measures are keeping pace with advances in the technology.
The debate intensified after 27-year-old researcher Jacob Coxon left Anthropic earlier this week. In a social media post that received more than 150 million views, Coxon said leading AI companies, including Anthropic and OpenAI, were taking enormous risks by racing towards increasingly autonomous and potentially self-improving AI systems.
"The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible - but I hear the same people express fear privately. No other human activity poses this level of danger," Coxon wrote on X.
Coxon's concerns were echoed by Anthropic safety executive Evan Hubinger, who estimated that there was a greater than 10% chance AI could cause human extinction within the next decade.
Anthropic CEO Dario Amodei has also warned about the risks associated with rapidly advancing AI.
He called for slowing down AI development and bringing it under closer oversight. In an online essay, Amodei said the question was not whether to build advanced AI, but how to manage its "serious" risks, adding that companies and governments need more time to address them. The heads of two rival firms -- OpenAI's Sam Altman and Elon Musk, said they agree with his assessment.
