OpenAI announced on Wednesday that it had concluded a major fundraising of 6.6 billion US dollars, which values it at 157 billion, strengthening the status of the creator of ChatGPT as a star of Silicon Valley and a key player in artificial intelligence (AI ) generative.
“Every week, more than 250 million people around the world use ChatGPT to improve their work, creativity and education,” the company said in a brief statement.
It is counting on this funding in particular to continue its “cutting-edge research” in AI, increase its computing capabilities – generative AI is extremely energy-intensive – and “create new tools”.
Venture capital firm Thrive Capital confirmed it led the funding round.
Various names of potential investors have been circulating in the American press for weeks, including Microsoft, which has already injected $13 billion into the start-up, and the giant Nvidia, another leader in generative AI thanks to its ultra-sophisticated chips. designed specifically for new technology.
According to the Wall Street Journal (WSJ), the Japanese conglomerate SoftBank and MGX, an investment fund backed by the United Arab Emirates, are also in the game, but Apple, which already uses OpenAI models in its new generative AI system, has recently withdrawn from negotiations.
OpenAI launched the wave of generative AI (production of content on a simple query in everyday language) at the end of 2022 with its ChatGPT interface, which has become synonymous with this new type of service.
Since then, from Microsoft to Google and Meta (Facebook, Instagram), all the major technology groups have competed with tools intended to help humans on a daily basis, from writing messages to education and artistic creation.
Departures
This fundraising makes OpenAI one of the three largest unlisted start-ups in the world, according to Bloomberg, alongside SpaceX, Elon Musk’s space exploration group and Chinese entertainment giant ByteDance, owner of the TikTok application.
The major step for the company, still confidential two years ago, comes after a year marked by the launch of new innovative AI models, but also by numerous controversies.
Last June, the addition of voice to ChatGPT forced OpenAI to apologize to actress Scarlett Johansson for having designed a voice very close to hers, sparking controversy over the creation of voices copied from those of humans.
Above all, last November, the board of directors fired Sam Altman, co-founder of OpenAI, accusing him of a lack of transparency and attention to security issues.
He was reinstated in his role at the head of the start-up after a few days, supported by the overwhelming majority of employees and the main investor, Microsoft.
But since then, there has been a succession of departures at the top. In May, co-founder and chief scientist Ilya Sutskever left the group, as did Jan Leike, responsible for managing risks associated with generative AI.
“Humanity”
Then another co-founder, John Schulman, also disappeared, while the president, Greg Brockman, went on leave. And Mira Murati, OpenAI’s chief technology officer, resigned last week, remaining vague about her reasons.
Internal tensions could be linked to the recent evolution of OpenAI, originally founded in 2015, as a non-profit organization, supposed to work for the good of humanity.
According to the WSJ, investors in this round will have the right to withdraw their funds if the organization does not convert to a for-profit company.
Sam Altman could also obtain a stake in his company, which would probably make him a multi-billionaire.
OpenAI expects to lose around $5 billion this year, despite revenue of around $3.7 billion, according to the New York Timesbecause of the costs linked to the operation of the models, in particular.
In mid-September, OpenAI launched o1, a new kind of AI model, capable of reasoning and answering more complex questions, particularly mathematical ones, thus hoping to reduce the risk of hallucinations (absurd, incoherent or factually incorrect answers). false).
“o1 thinks before responding,” assured OpenAI, which wants to develop “general” AI, that is to say artificial intelligence but similar to that of humans.
“We are making progress in our mission to ensure that artificial general intelligence benefits all of humanity,” the start-up said in its press release on Wednesday.