Dario Amodei, CEO of Anthropic, the main rival of Openai, wrote on his personal website: “(China is) an authoritarian government which has committed human rights violations, behaved in a way Aggressive on the world scene and will be much more unrected in these actions if they are able to match the United States in AI. Export controls are one of our most powerful tools to warn this… ”
Sam Altman, the CEO of Openai, wrote in the Washington Post: “We are faced with a strategic choice on the type of world in which we are going to live: will he be the one in which the United States and the Nations Allies advance a world AI that disseminates the advantages of technology and open access, or an authoritarian, in which nations or movements that do not share our values use AI to cement and extend their power? »»
But of course, they would say that. The two main Deepseek AI platforms could potentially undermine their business model. They charge a fraction of the costs of using the more advanced versions of Claude AI of Anthropic and the Chatpt of Openai. The software of American companies is closed, which means that no one, except that their engineers can display, modify or rely on them.
Deepseek’s is open-open; Researchers and external programmers can modify and build on its software. This is why, as nature reported, the scientific journal, “scientists flock to Deepseek -R1, a” model of reasoning “of cheap and powerful AI” – to use it to help evidence Mathematics to view data and create a more efficient cognitive neuroscientist.
Silicon Valley techno-lords used to plead for universal benefits of being open and transparent. Now, their richest and most powerful have jumped on the Washington train on the hi-tech being a priority for national security against “autocratic” China. They have become an integral part of the U.S. Intelligence Industrial Intelligence Complex.