OpenAI plans to launch its next-generation artificial intelligence model, Orion, by December. Unlike the GPT-4o and o1 models that OpenAI released in recent months, Orion will not initially be offered to a wide user base, but to companies with which OpenAI closely cooperates. The aim is for these companies to test Orion on their own products and features.
It is stated that OpenAI, which cooperates closely with Microsoft, aims to host Orion first on the Azure cloud platform, and this process may start in November. While it is said that engineers at Microsoft are preparing to support Orion on Azure, it is also stated that this model is described as the successor of GPT-4. However, it is not yet clear whether the company will introduce this model to the outside world as “GPT-5”. However, OpenAI and Microsoft officials refrain from making a statement about this new model.
According to the statements of one of OpenAI officials, it is predicted that Orion can be up to 100 times more powerful than GPT-4. This model is being developed independently of the o1 model, which was released last September and has advanced reasoning capabilities. In line with its future plans, OpenAI aims to build a more advanced artificial intelligence system by combining large language models (LLM), ultimately calling this system artificial general intelligence (AGI).
It was also previously reported that the o1 model, codenamed “Strawberry,” provided synthetic data for Orion’s training. In September, with a celebration organized by OpenAI researchers in honor of the completion of the training process of the Orion model, information was leaked that the end of the training of the model had been reached. This timing also coincides with OpenAI CEO Sam Altman’s mysterious post on social media platform X, “Looking forward to the winter star teams rising soon.” According to the comments made, Altman was pointing to the Orion constellation, which is most clearly seen during the winter months. A symbolic relationship is established between the name of the model and this star cluster.
Significant changes are taking place in OpenAI governance
The launch of Orion comes at a critical time in determining OpenAI’s future position in the field of artificial intelligence. The company recently closed a new investment round of $6.6 billion, providing the financing needed to transition to a profit-oriented structure. However, there is a remarkable staff change within the company. The departure of key figures such as CTO Mira Murati, head of research Bob McGrew, and vice president of post-model training Barret Zoph has raised questions about the future of OpenAI.
OpenAI is expected to take its artificial intelligence capabilities to a higher level with the Orion model and strengthen corporate collaborations accordingly. Introducing Orion as a powerful model that enhances OpenAI’s existing technologies has the potential to move the company closer to its AGI goal. However, what impact these developments will have will become clear once the model is fully put into use.