Recent reports have stirred interest around OpenAI’s upcoming advancement in AI technology, tentatively known as Orion. This model is poised to succeed the highly acclaimed GPT-4, although it remains uncertain if it will debut as GPT-5. Both OpenAI and Microsoft have remained tight-lipped, opting not to comment on these reports, adding an air of mystery around the impending release.
Back in September, OpenAI’s Shaun Ralston shared intriguing insights on X regarding the evolutionary leaps their models have taken since the introduction of GPT-3. He tantalizingly referred to a “GPT-Next” version, anticipated to launch before the year is out. According to Ralston, this model has undergone rigorous training using a “compact Strawberry (OpenAI o1) version” and boasts an astounding 100-fold increase in “computational volume” over its predecessor, GPT-4.
Further adding to the excitement, Ralston also hinted at the development of Orion, describing it as a distinct entity. This model has been developed utilizing the impressive horsepower of “10K (Nvidia) H100 GPUs.” With all these details emerging, it’s clear that OpenAI is paving the way for remarkable enhancements in AI capabilities, setting the stage for what could be a revolutionary shift in artificial intelligence technology. Enthusiasts and experts alike are eagerly waiting to see what groundbreaking features and improvements Orion may bring to the table. Stay tuned for what promises to be an exciting evolution in the world of AI!
