OpenAI unveils most advanced reasoning model, competes with Google

• Published On: 21 December 2024 at 18:43

OpenAI has disclosed its intention to launch a novel AI model that possesses sophisticated human-like reasoning capabilities, indicating a significant advancement in AI innovation.

According to Bloomberg, this action is expected to intensify competition with Google and other competitors as they both strive to expand the boundaries of artificial intelligence, which has the potential to be applied in a variety of sectors, including healthcare and education.

On Friday, OpenAI will present its latest o3 model during a livestream event. The model is intended to address intricate, multi-step problems by dedicating additional time to processing before generating responses. Additionally, the o3-mini will be introduced in a smaller size.

The event will emphasise the extent to which o3 surpasses its predecessor, o1, which was released in September, particularly in its ability to address complex coding enquiries.

OpenAI will extend an invitation to safety and security researchers to submit applications for the purpose of testing the models prior to the release of new software.

According to a spokesperson, the organisation intends to introduce the o3 models in the first quarter of the following year.

Two years ago, the San Francisco-based startup initiated the AI arms race with the release of ChatGPT, which is powered by the GPT-3.5 language model.

Following this, in 2023, it introduced GPT-4, which provided enhanced creativity and accuracy. More recently, it introduced o1, its inaugural reasoning model.

The spokesperson indicated that OpenAI refrained from naming the new model o2 "out of respect" for the British telecommunications brand of the same name. This decision is consistent with previous reports.