OpenAI releases OpenAI o1, a far more sceptical AI model

OpenAI has just released its latest model, but it is not branded with the terms "ChatGPT" or "GPT". Instead, it is naming the model after the company: OpenAI o1.

OpenAI explained how the new models function. These features sound similar to those rumored under the code name "Strawberry."According to the business, the OpenAI o1 models are designed to "spend more time thinking before they respond."

The blog article also stated that the goal is to have OpenAI o1 work on considerably more complex jobs and projects:

In our tests, the next model update performs similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology. We also found that it excels in math and coding. In a qualifying exam for the International Mathematics Olympiad (IMO), GPT-4o correctly solved only 13% of problems, while the reasoning model scored 83%. Their coding abilities were evaluated in contests and reached the 89th percentile in Codeforces competitions.

OpenAI o1 is unable to perform certain tasks that its ChatGPT LLMs can, such as searching the internet or reviewing uploaded photographs and other files. Nonetheless, the company believes that this new model represents a significant advancement in generative AI, which is why it was termed OpenAI o1, which stands for "resetting the counter back to 1."

In addition to the flagship OpenAI o1 model, the business is introducing OpenAI o1-mini. It has been developed to be faster and less expensive to use, with a focus on coding chores.

Subscribers to both ChatGPT Plus and ChatGPT Team will get access to the new OpenAI o1 models beginning today. The models will be offered to ChatGPT Enterprise and ChatGPT Edu plan members beginning next week. In the near future, free ChatGPT users will have access to the OpenAI o1-mini model.