aecifo

Sam Altman calls reports of OpenAI’s next frontier model launching in December ‘fake news’

Umito October 25, 2024

What you need to know

A new report suggests that OpenAI is preparing to deliver a next-generation frontier AI model named Orion in December.
The report further suggests that the ChatGPT maker would grant companies it works closely with priority access to the model, including Microsoft.
Sam Altman responded to the claims about X, calling them “fake news out of control.”

OpenAI has had a lot to offer in recent months, between the mass exodus of high-level executives, bankruptcy reportsand recent claims that it is violate U.S. copyright law. It’s been over a month since Chat maker GPT revealed its latest round of next-generation AI models, including OpenAI o1 and o1-mini who are equipped with advanced reasoning skills in the areas of coding, mathematics and science.

Strawberry AI models are already very promising, according to benchmarks shared by OpenAI. A study found that OpenAI o1 could pass OpenAI Research Engineer job interview for coding at 90-100% ratewhich could raise concerns among software developers on their job security in an AI-powered world.

The edge recently reported that the maker of ChatGPT is preparing to launch its next frontier model, named Orion, in December, on the occasion of ChatGPT’s second anniversary. The outlet further revealed that the next-generation AI model will not be available on a large scale through ChatGPT like previous versions. Instead, the company will prioritize access to the model to companies it works closely with, including Microsoft, which could host the model in Azure in early November.

According to The Verge, Orion is internally considered the successor to OpenAI’s GPT-4. It is unclear whether OpenAI will ship Orion as GPT-5, following an earlier report claiming the company could be move away from the traditional naming of its models. However, Sam Altman indicated that The successor to GPT-4 would be more intelligent and function as a “virtual brain”.

Was it a hoax?

OpenAI logo on 16x9 phone

OpenAI logo on the phone (Image credit: Getty Images | SOPA Images)

Following The Verge’s report, news of the next-generation AI model’s potential release spread like wildfire. But as it seems now, we may have to take this information with a grain of salt. It’s especially after OpenAI CEO Sam Altman responded to the report and called it “fake news out of control.”

This doesn’t entirely mean that The Verge’s report is false. It could be a matter of timing or something else. As you may know, this isn’t the first time we’ve expected OpenAI to deliver a particular product or service at a specific time, but it ultimately turned out to be something completely different .

Earlier this year, in May, rumors swirling around the windmill indicated that OpenAI was preparing to launch an AI and GPT-5-based Google search competitor. However, Sam Altman, CEO of OpenAI, published on the company was not shipping an AI-powered engine or GPT-5 during his special event. He suggested that the event would be magical, referring to the launch of GPT-4o.

Interestingly, it was later established that OpenAI had put its security team under immense pressure during security testing of the product, reducing the process to less than a week. At that time, several high-level executives had left the AI company, some citing security reasons, including prioritize shiny products over safety processes. While an OpenAI spokesperson admitted that the launch was apparently rushed, with RSVPs sent to guests before testing the modelhe was adamant that the company had not taken any shortcuts.