Mistral AI co-founder does not believe in technological uniqueness and aims at a different threat

The “holy grail” of artificial intelligence companies is AGI, orgeneral artificial intelligencegeneral artificial intelligence. This is the AI ​​that will surpass human intelligence and become a starting pointa starting point technological singularity, a surge in the development of new technologies. However, not everyone is of the same opinion as Arthur Mensch, co-founder of Mistral AI.

MistralMistral AI is a French company founded by former employees GoogleGoogle DeepMind and Meta. In an interview with New York Times, he claims that for him the race for AGI is a religious hobby. He sees this as a desire to create God. Being an atheist, he does not believe in God, and therefore in the AGI. A view that stands in stark contrast to large AI companies for whom this is a major long-term goal.

Cultural hegemony imposed by AI?

In his opinion, the main threat from AI is the dominance of web giants, especially American ones, which will create global standards that are incompatible with the cultures and policies of other countries, which will find themselves in a situation of strategic dependence. For many AI experts and policymakers, Mistral represents the best hope for developing European AI that can compete with artificial intelligence developed in the US and China.

The startup has just published the latest version of its large 8x22b language model online. by using magnetic link BitTorrentBitTorrent, published in a simple tweet. Simply copy all the text and paste it into your BitTorrent program to download the full 262GB model. Unlike giants like OpenAI and Anthropic, which protect the source code of their AI, Mistral AI has decided to publish it under a free license so that anyone can use it.