[ad_1]
Mistral AI, a Paris startup founded seven months ago by researchers from Meta and Google, has raised 385 million euros, or $415 million, in another sign of interest in a new type of the artificial intelligence that drives online chats.
The deal values the 22-person company at about $2 billion, two people familiar with the deal said. Investors include Silicon Valley venture capital firm Andreessen Horowitz and Lightspeed Venture Partners.
The starting price has increased more than seven times in six months. In the summer, it raised seed funding of €105 million (about $113 million) which valued the company at about $260 million.
Mistral creates technology that other businesses can use to build chatbots, search engines, online tutors and other AI-powered products. It’s among a small group of companies — including many tech companies and a handful of startups — that are building AI that can rival the technology under development at OpenAI, the San Francisco-based startup that started ate the AI boom last fall. the release of the ChatGPT chatbot.
Mistral is also among the companies that believe in sharing this technology as open-source software – software that can be freely copied, modified and reused – provided export everything they need to quickly build their own chatbots. Antitrust companies such as OpenAI and Google argue that the open-source approach is dangerous and that dangerous technology can be used to spread information and other harmful things.
The result of Mistral has been very important in France, where leaders like Bruno Le Maire, the finance minister, have said that the company gives the country an opportunity to challenge the big American companies. Europe did not produce many meaningful technology companies until the dot-com boom and saw artificial intelligence as a field that could gain ground.
Investors are pouring money into other startups that believe in the open-source approach. Perplexity, founded last year by another group of high-profile researchers, has raised a $70 million funding round that valued the company at $500 million, a person familiar with the deal said. Investors include IVP and Bessemer Venture Partners.
“We believe AI should be open,” said Anjney Midha, a general partner with Andreessen Horowitz who led the investment in Mistral. Many of the major technologies driving modern computing are open source, he added, including computer operating systems, programming languages and databases.
Mistral was founded by Timothée Lacroix and Guillaume Lample, who previously worked as researchers at the Paris AI lab of Meta, the owner of Facebook and Instagram, and Arthur Mensch, who was a researcher at DeepMind, who an AI lab acquired by Google in 2014 for $650 million.
The company’s employees like to joke that the first letters of the founders’ last names are “LLM” – also short for the big language, the AI technology that the company is building.
Companies like OpenAI, Microsoft and Google are leading the AI race, after spending hundreds of billions of dollars on this type of technology. By analyzing the vast amount of computer text collected from the internet, a multilingual model can learn to generate the text alone. This means it can answer questions, write around and generate computer code.
Companies like OpenAI and Google believe that this technology is so powerful, they can release it to the public in the form of an online chatbot after spending months using digital barriers to prevent it. from spreading misinformation, hate speech and other toxic things.
But many AI researchers, technologists and entrepreneurs believe that the AI race will be won by companies that build the same technology and then give it away for free – without barriers.
Meta, the first home of the two Mistral founders, was at the forefront of companies promoting this open approach. This year, the tech giant created a multilingual model called LLaMA and released it as a special app.
On Sunday, Mistral also revealed its new technology as open-source software, saying it is implemented at a level similar to Meta’s technology.
Public sharing is the bottom line for AI, said Mr. Midha, this is the safest way because many people can review the technology, find its flaws and work to eliminate or reduce them.
“No one engineer can find all the boxes,” he said. “Greater communities of people are better at building cheaper, faster, better, safer software.”
Adam Satarian contribution reports from London.