There is a new participant in the Chinese artificial intelligence chatbot market. It is competing with giants such as OpenAi, Gemini, Claudeai, etc. Interrupt US hegemony in generative chatbot models based on AI. Deepseek, the Chinese alternative to American manufacturing AI models, has taken the industry by assault.
Deepseek founders spent only $ 5.6 million to build the model. This is just a fraction of the budget used by giants such as Openai, Gemini of Google and Claudeai to build their models.
Now, an old video of the founder of Operai, Sam Altman is viral, where he called to build an AI model with a budget of $ 10 million “without hope.” Speaking in an event in 2023 in India, Mr. Altman was asked: “How is a small and intelligent team with a budget of $ 10 million could build something substantial within the AI?”
This is quite fun in retrospect.
In India in 2023, Altman was asked as if a small and intelligent team with a budget of $ 10 million could build something substantial within AI.
His answer: “It is totally desperate to compete with us in the foundation models” https://t.co/pdyihv2x1m
– Arnaud Bertrand (@rnaudberand) January 28, 2025
His answer: “It is totally desperate to compete with us in the foundation models.” The Deepseek chatbot has become the best -qualified free application in the American Apple application store and the American chip manufacturer Nvidia has lost almost $ 600 billion in valuation in a single day, the largest for a single action in The history of the market.
Deepseek won a traction last month when he declared in a document that training his Deepseek-V3 model required less than $ 6 million in computer energy using the NVIDIA H800 chips of lower capacity H800.
Deepseek-R1, launched last week, is 20 to 50 times more affordable to use than OPENAI’s O1 model, according to the task, according to a position in the official Wechat account of Deepseek.
Altman praised the RI model of his rival in X and said: “De Deepseek, R1 is an impressive model, particularly around what they can deliver for the price.”
“But above all we are excited to continue executing ourselves in our research route and believe that more computer science is more important now than ever to succeed in our mission,” Altman added.
The popularity and low cost of Deepseek have asked questions about the American narrative of promising billions of dollars to build generative models.