Openai has collected $ 40 billion (£ 31 billion) by collecting donations by the Japanese group Softbank with an agreement worth $ 300 billion to the Chatgpt developer.
Openai said that the financing tour will allow the company to ı further force the limits of AI research ”. He added that the support of Softbank will “lead to net or artificial general intelligence for AI systems that can match or overcome people in almost all cognitive tasks.
Openai General Manager Sam Altman, “Hundreds of million people use chatgpt every week,” he said. “This investment helps us to force the limit and make AI more useful in daily life.”
Donation collection, Altman’s new image creation feature with consumers, said that it has added a million users in just one hour-a viral obsession with Ghibli-style art with the new creation studio.
If Softbank meets certain conditions, it will first put Openai $ 10 billion and 30 billion dollars until the end of 2025. The Japanese company will provide 75% of financing in the last round with the rest of Microsoft and technology -oriented investment companies Coatue Management, Altimeter Capital and Thrive Capital, an important supporter of Openai.
“AI is a descriptive force that shapes the future of humanity, Mas says Masayoshi, the president and general manager of Softbank. “Our expanded partnership with Openai accelerates our common vision to unlock the full potential.”
Openai also announced on Monday that it is a more open productive AI model because it faces increasing competition in open source in the field of open source from Deepseek and Meta, which can be adjusted unlike Chatgpt. Openai is preparing to launch a fine -adjusted “open weight” model that can be finely adjusted by users.
Openai was a violent advocate of closed, registered models, which did not allow developers to change the basic technology to adapt AI to their goals. Altman, “We have been thinking about this for a long time, but other priorities have received priority. Now it makes you feel important,” he said.
Openai and Google’s defenders of closed models often rejected open models more risky and malicious and more vulnerable to non -US government use.
Recently, Elon Musk, a former Openai investor, a former Openai investor with a 100 billion dollar consortium proposal for the company, called Openai to öncelikle once well return to security -oriented power ”.
Large companies and governments are reluctant to use non -control AI models, especially when data security is concerned. Meta and Deepseek allow companies to download and change their models.
Mark Zuckerberg, General Manager of Meta, said this month that the Great Language Model Llama has made 1 billion downloads. Deepseek’s low -cost R1 model shook the world of artificial intelligence in January.
Openai is based on the success of generation features in Chatgpt. Altman claimed that the new image tool drowned Openai’s graphic processing units as well as marking the increase in user numbers.
The images that multiply the art of Japan’s studio Ghibli has increased since the launch of the new feature. Altman stated that the feature has less restriction on use and represents the vehicle öğren high water sign for us to allow creative freedom ”.
With Openai’s 4O image production, 24 hours have passed since it was unexpectedly shook the AI image world.
Here are 14 examples that are the most minded so far (produced by 100% AI):
1. Studio Ghibli style memespic.twitter.com/e38mbnpnqh
– BARSEE 🐶 (@HEYBARSEE) 26 March 2025
An Openai spokesman: uz We continue to prevent generations in the style of individual living artists, but we allow the wider studio styles that people use to produce and share the original and inspiring original fan creations. ”
With Agent France-Presse
