Skip to main content

uncategorized

New GPT-5.4 clobbers humans on pro-level work in OpenAI's tests - by 83%

By David Gewirtz Source: ZDNet
New GPT-5.4 clobbers humans on pro-level work in OpenAI's tests - by 83%

OpenAI's new GPT-5.4 model has achieved an 83% performance score in tests, surpassing human professionals across nine industries and 44 different professional-level tasks. The results suggest that advanced AI systems are now competitive with expert human workers in specialized domains. OpenAI claims this represents a significant milestone in AI capability, though the specific nature of the tests and their real-world applicability remain unclear from the available information.

Read Full Article at ZDNet

Related Articles