OpenAI is preparing to unveil its latest artificial intelligence model, named Orion, which is anticipated to significantly advance the capabilities of AI technology. Scheduled for launch in December, Orion is reported to be 100 times more powerful than its predecessor, GPT-4. This model is seen as a crucial step towards achieving artificial general intelligence (AGI).
The deployment strategy for Orion differs from previous models. Initially, access will be granted to select strategic partners, including Microsoft, allowing them to incorporate Orion into their products and services. According to sources, Microsoft engineers are set to host Orion on the Azure cloud platform as early as November, although this has not yet been officially confirmed by either OpenAI or Microsoft.
Orion has been developed using the o1 model, known as Strawberry, and completed its training process last September. OpenAI’s CEO, Sam Altman, has also hinted at the upcoming release through a recent post on social media platform X.
The introduction of Orion represents a significant milestone for OpenAI, particularly as the company undergoes a restructuring aimed at achieving profitability. This new model is expected to expand the possibilities for applications in artificial intelligence, spanning areas such as natural language processing and content generation.