Sam Altman, OPENA CEO, during an interview with Bloomberg House on the day of the opening of the World Economic Forum (WeF) in Davos, Switzerland, on Tuesday, January 16, 2024.
Chris Ratcliffe | Bloomberg | Getty Images
Openai announced a research review of GPT-4.5, a large-language model of general purpose on Thursday. Initially, it will be open to software developers and people with CHETGPT PRO subscriptions.
The model represents incorrect information less frequently than its predecessor, Openii said in blog blog.
“Early testing shows that the interaction with GPT -4.5 feels more natural. Its wider knowledge base, improved ability to monitor users’ intention and greater ‘EQ’ makes it useful for tasks such as improving writing, programming and solving practical problems,” Openi wrote. GPT-4.5 showed a lower hallucination rate than the GPT-4O and O1 Openi model in one test, the company said in the company Report following the edition Thursday.
At the same time, the new model will not be the most important in reference values, said OPENAI co-founder and Sam Altman CEO on Thursday X postDescribing the model as “magic in it, I haven’t felt before.”
The edition, however, represents progress for one of the most respected startups in the world. Last month CNBC reported that Openii discussed a $ 40 billion funding circle on a Estimate of $ 340 billion.
With more 400 million Weekly active users, Openai intends to bring GPT-4.5 to those with Chatgpt Plus and subscriptions on that next week, said the company’s research leadership, Alex Pailo, a member of the company’s technical staff, said to A A Livove.
Chatgpt EDU and Enterprise subscribers will get access to next week, Pailo said.
GPT-4.5 is “Giant, AFRIGN MODEL,” Altman wrote on Thursday in X Post. “We really wanted to start it on a plus and pro at the same time, but we got a lot apart and we were left without a GPU. Next week we will add tens of thousands of GPU -a then they will take it to a plus.”
GPT-4.5 is available in review by Microsoft Azure Azure Foundry, CEO of Satya Nadella announced On X. Microsoft has invested over $ 13 billion In the Openi and has installed startup models into its products. Microsoft also helps to ensure computer resources Openii.
The edition of artificial intelligence models comes one month after computers with an effective approach used by the Chinese Deepseek Laboratory. Market capitalization Nvidiawhose chips are widely used among the developers of AI models, rejected Nearly $ 600 billion in one day.
Two weeks later, Altman wrote in X post That the company wants to improve the public division of the road.
Altman said GPT 4-5 would come out, followed by a GPT-5, which will be attracted to the wider number of Openi Technology. The company also offers so -called Riseling models who make many budgets at a time when people teach inquiries. GPT-4.5, for comparison, would be a “last model that is not thought out”.
