According to tech outlet The Verge, OpenAI is set to release its next AI model, Orion, in December. Orion is anticipated to be a significant leap forward, potentially offering 100 times the power of the current GPT-4 model.
Unlike previous launches, though, Orion won’t be accessible to the general public on ChatGPT immediately. Instead, OpenAI plans to give early access to select partners, allowing them to develop products and features with the model before it reaches a broader audience.
What is Orion, and why is it so special?
Orion, viewed as the successor to GPT-4, is set to introduce a new level of AI capabilities, although it remains unconfirmed whether it will be officially named GPT-5. Unlike OpenAI’s recent “O1” model, which was designed to enhance logic and reasoning skills, Orion aims to support a broader array of AI applications and cater to a wider audience.
Orion’s development has benefited from synthetic data generated by the o1 model, known internally as “Strawberry.” This technique has enabled OpenAI to refine and expand the model’s capabilities.
According to The Verge, Microsoft Azure engineers are preparing to launch Orion on their platform early in November.
This release aligns with OpenAI’s recent shift to a for-profit model, following a significant funding round that raised $6.6 billion and valued the company at $157 billion. It also follows the recent debut of OpenAI’s ChatGPT app for Windows.
While OpenAI hasn’t officially confirmed Orion’s launch or shared further details, the model is expected to significantly impact AI, enhancing the accuracy and responsiveness of current capabilities.
