Anthropic with best AI is like Hitler with nuclear bomb: DeepSeek engineer
Who will get to AGI first, and who should control and be responsible for the world’s most powerful AI if it eventually reaches AGI-level capabilities? As the US and China race to build increasingly capable AI, a DeepSeek engineer argues that the most powerful AI should definitely not end up in the hands of Anthropic and its CEO Dario Amodei.
The engineer, writing just when there is a debate on AI getting way to powerful for humans to effectively control it, says that if a company like Anthropic indeed reaches AGI — short for Artificial General Intelligence or human-like intelligence — and holds control of it, it will be like Hitler obtaining atomic-bomb technology before the Allies.
In other words, he argues that giving one company full control of AGI could concentrate enormous power in its hands, creating a deeply unequal society where other companies and people are left at its mercy.
In a five-page essay shared on WeChat titled “I Have to Buy My Talent in Yesterday” (translated from Chinese), the DeepSeek engineer reflects on how quickly AI is advancing and how that progress could change not just individual careers but society as a whole.
For much of the essay, he discusses what could happen if one company ends up controlling the world’s most advanced AI. He argues that if a company like Anthropic — which has been releasing increasingly powerful models in its Claude family — were to develop the world’s most advanced AI or AGI and retain control over it, the company could gain an enormous advantage over other companies and people.
In his essay, he compares AI to the atomic bomb. He draws on an example from the history of World War II, pointing to the fear that if Hitler had gained the bomb before the Allies, the consequences could have been devastating. He says that today, in the AI war, reaching AGI is like developing the atomic bomb, and if one company indeed controls AGI, it could be deeply concerning. In his view, the fear is not simply about who develops the technology first, but about what could happen if a single actor gains control over a capability that could fundamentally change the balance
“I still believe that the most cutting-edge intelligence should be supplied to everyone in an open and cheap manner,” he wrote, according to the translation of his post that is floating on the web. “I do not trust that Anthropic or OpenAI can do this, and especially do not hope that Anthropic masters the most advanced artificial intelligence or AGI—to exaggerate slightly, its seriousness is no less than letting Hitler master atomic bomb technology before the Allies.”
The DeepMind engineer argues that the most advanced AI should remain open and affordable. He warns that the rapid development of AI could lead to two very different futures: one where AI dramatically improves productivity and living standards, and another resembling Cyberpunk 2077, where a small group of technology companies controls most of the resources and the most powerful AI.
And to address this shift in power, he suggests communism as one possible future, where the government plays a bigger role in shaping the development and access to AI, while its benefits are broadly shared. He highlights DeepSeek, one of China’s biggest AI labs, which he says is already focused on developing powerful AI and making its models openly available.
“This is also why I chose and persist in staying at DeepSeek: we research powerful, fast, and inclusive artificial intelligence and open-source it, which perhaps can pull the world back a bit from the 2077 side,” he wrote.- EndsPublished By: Divya BhatiPublished On: Sep 15, 2026 14:07 ISTAlso Read | Elon Musk says to get smart you can play video games, or stay dumbAlso Read | OpenAI is reading ChatGPT chats, humans look at what people say to AIAlso Read | Not just OpenAI and Anthropic, even Zoho wooing developers with AI tools
