OpenAI is planning to launch its new AI model called Orion by December. Unlike the last two models, GPT-4o and o1, Orion won’t be widely available on ChatGPT at first. Instead, it will be given to some partner companies to help them create new products and features.
Microsoft, a major partner of OpenAI, is already preparing to host Orion on its Azure platform, possibly as soon as November. Although Orion is considered a successor to GPT-4, it’s not certain if OpenAI will name it GPT-5. The plans could change, so the release date might be delayed.
An OpenAI executive hinted that Orion might be up to 100 times more powerful than GPT-4. This model is different from the o1 reasoning model released in September, and OpenAI’s goal is to combine its AI models to create a more advanced system, potentially achieving artificial general intelligence (AGI).
OpenAI had used its o1 model, code-named “Strawberry,” to generate data for training Orion. In September, the company celebrated the completion of the training process. Around the same time, OpenAI’s CEO, Sam Altman, hinted at the new model with a cryptic post mentioning “winter constellations,” referring to Orion, a constellation visible from November to February.
The launch of Orion comes at a crucial time for OpenAI, which recently secured $6.6 billion in funding and is undergoing changes to become a for-profit company. Additionally, the company is facing significant staff departures, including its CTO Mira Murati, chief research officer Bob McGrew, and VP of post-training Barret Zoph.
Source: theverge
