Jan 14, 06:34 AM

OpenAI's o3 Model Achieves 87.5% Score in December 2024, Surpassing Previous AI Benchmark of 55.5% and Solving 72% of Problems

OpenAI's latest chatbot model, known as o3, has demonstrated remarkable advancements in artificial intelligence, achieving a score of 87.5% on a benchmark test aimed at assessing human-level intelligence. This performance significantly surpasses the previous best score of 55.5% for an AI system. The o3 model's capabilities have been described as groundbreaking, with reports indicating that it successfully solved complex math problems that left human experts astonished. These developments were revealed in a Time magazine investigation, which noted that much of the progress in AI is occurring behind closed doors, with the o3 model achieving breakthrough performance in December 2024. The implications of these advancements could redefine intelligence measurement and problem-solving in the AI era.

#Time

Written with ChatGPT (GPT-4o mini).

OpenAI's o3 Model Achieves 87.5% Score in December 2024, Surpassing Previous AI Benchmark of 55.5% and Solving 72% of Problems

Sources

Additional media