The reasoning is the future of artificial intelligence, and to promote this idea, Operai has just announced the launch of its O3 and O4-mini models for Chatgpt, now designed to “think more before answering”, the company announces in a press release.
The different tools offered with chatgpt, such as web navigation, understanding of the images sent or the generation of images, to solve the “most effectively” described “problems described as” complex “and which can be requested in several steps.
An almost autonomous, but also slower
It is not yet a completely autonomous AI, but it is a “progress”, according to the company, which sells the merits of O3 and O4-minini describing that the models that can “think with images.”
For example, if you download images such as sketches, whether vague or of poor quality, O3 as O4-mini you can adjust the image received when turning or zoom to understand what the user requires.
OpenAI ensures that these two models were tested as part of their security program not to “hallucinate” (present information as true when it is false, note of the editor) or, in any case, the most little possible, and to guarantee the best possible result. Keep in mind, however, that the result should be longer to generate. It is seen in the example given by the new company that reaches reasoning in more than a minute.
To use O3 and O4-mini, you must subscribe to Chatgpt Plus, Pro and Team.
This is a great advance for Openai, but others are still expected in the coming months. The objective, by the end of the year, is to offer Chatgpt 5, a review of 4.5, destined to the general public, and that should be economical for the generation of texts, images and videos.
Source: BFM TV
