[ad_1]
New Delhi. OpenAI is soon going to launch its best AI model version o3 so far. In September this year, the company had launched the o1 model, about which users say that it takes time to respond to signals. The new model is faster than o1 and can respond in a more step-by-step, logical manner. OpenAI CEO Sam Altman said during the launch that o3 will usher in the next phase of AI. He said that these models can be used to perform more complex tasks that require a lot of reasoning.
Who is better in performance?
When it comes to performance, the new o3 model surpasses many benchmarks compared to the o1. These include complex coding-related skills, aptitude in solving scientific problems, and even advanced mathematical problems. It is being said that this model is three times better at answering questions than ARC-AGI tests. In simple words, understand it like this that ARC-AGI is a test which checks how trained any AI model is and how and how fast it understands a task.
The test found that the o3 models are capable of solving extremely difficult math and logic problems that they are facing for the first time.
Why is it special?
o3 is a leading AI model developed to provide advanced reasoning and intelligence across many complex tasks. It has been launched with a smaller version, o3 mini. The o3 model is designed to solve some of the most challenging problems in coding, general intelligence, and mathematics.
For now, OpenAI is starting with public security testing, which reflects the company’s cautious approach. If early results and benchmark performances are to be believed, o3 models could prove to be a significant step forward in the advancement of AI models.
Tag: technical news, tech news hindi
first published : December 21, 2024, 4:19 PM IST
[ad_2]
