ChatGPT parent OpenAI is reportedly planning to launch a new AI model soon but not everyone will get it



  • OpenAI is planning a new AI model
  • It could be out in December
  • It might be called GPT-5

When OpenAI launched ChatGPT, powered by GPT 3.5, in November 2022, the world couldnt’ stop marvelling at the new chatbot in town. It took everyone by surprise with its ability to respond in a human-like manner, give life advice, joke around, compose music and poetry and whatnot. As ChatGPT gained immense popularity, OpenAI kept working on making the chatbot better and better and has since rolled out many updates and new language models. Now, yet another AI model might just be in store.

OpenAI prepares for Orion launch

According to a report in The Verge, OpenAI is planning to release Orion, its next-generation AI model, by December this year. According to the report, Orion will initially be available only to select companies collaborating closely with OpenAI, aiming to allow early partners to build tailored products with its advanced capabilities. Microsoft, a primary OpenAI partner, is also reportedly preparing to host Orion on Azure by November. Internally seen as a successor to GPT-4, Orion could represent a substantial step toward more capable and versatile AI.

This approach marks a shift from previous OpenAI releases like GPT-4 and GPT-3, which became widely accessible through ChatGPT soon after their launch. Instead, Orion’s limited rollout will enable OpenAI to gather data and refine the model’s responses with real-world feedback, especially in corporate applications.

GPT-4 successor?

Internally, the report adds that OpenAI views Orion as a successor to GPT-4, but it remains uncertain whether the model will officially be marketed as GPT-5. Altman has time and again spoken about GPT-5 but he hasn’t ever made a definitive statement. This hints at the potential of Orion as more than just another generational leap; instead, it may be the first in a line of hybrid models that gradually amalgamate various machine learning approaches within a unified framework.

The Verge report also says that Orion is expected to be up to 100 times more powerful than GPT-4, showing OpenAI’s intent to create a model capable of handling more complex tasks and possibly advancing toward artificial general intelligence (AGI), which is a form of AI that can understand, learn, and perform a wide range of tasks in a human-like manner. Earlier, OpenAI used its o1 model, also known as Strawberry, to generate synthetic data for Orion’s training, which could improve its understanding and adaptability in diverse contexts.



Source link

Leave a Reply