DeepSeek engineer compares Anthropic's control of AI to Hitler obtaining atomic technology before the Allies
In China, DeepSeek engineer Liu Shengyu compared the prospect of the US laboratory Anthropic controlling the most advanced artificial intelligence systems to a situation in which Adolf Hitler and Nazi Germany obtained atomic bomb technology before the Allies. This was reported by Anadolu Agency.
DeepSeek engineer's position
In a post in Chinese, Liu Shengyu, a machine learning systems engineer working on the DeepSeek V4.1 model, said he does not trust Anthropic and OpenAI to distribute advanced AI openly and accessibly. He contrasted their approach with DeepSeek's open-source models.
“I believe advanced intelligence should be available to everyone in an open and affordable way,” Liu wrote. He added that he does not trust Anthropic and OpenAI to implement such an approach and fears Anthropic could gain control over the most advanced AI or artificial general intelligence.
AGI, or artificial general intelligence, usually refers to systems capable of matching or surpassing human-level intelligence. Liu also said that as AI develops, social mobility will become increasingly difficult, and access to the most powerful systems will be needed to overcome class barriers.
More current news is available on the UA.News Telegram channel Telegram.
Debate over AI regulation
Liu's comments came amid warnings from representatives of the US AI industry and researchers about the pace of technological development and the possibility of systems emerging that act beyond human control.
Anthropic CEO Dario Amodei called over the weekend for AI development to be slowed and oversight to be strengthened. He also proposed placing independent third-party evaluators within leading companies in the sector, while noting that the measures should not “sacrifice commercial advantage.”
Last week, former Anthropic researcher Jacob Coxon resigned, accusing leading AI companies of irresponsible behavior. Anthropic's head of AI alignment research, Evan Hubinger, later said that, in his personal opinion, the risk of humanity's extinction due to AI over the next decade exceeds 10%.
China's independent Dongbuyaqiao Research Institute suggested that Amodei's proposals may be an attempt to strengthen control over industry rules before they are finalized. Chinese Foreign Ministry spokesperson Guo Jiakun said that intimidation, confrontation, and unfair competition would hinder proper global AI governance.
