GPT-5.4's 83% score suggests AI rivals expert professionals. Tests span nine industries and 44 real-world occupations. New capabilities boost coding, tools, and computer control. It seems like only ...

Could an AI ever truly think like a human? For years, skeptics have pointed to abstract reasoning and adaptability as insurmountable barriers for machine intelligence. Yet, that line in the sand may ...

ARC AGI 3, the latest iteration of the Artificial Reasoning Challenge, introduces a new benchmark for evaluating artificial general intelligence (AGI). This version emphasizes unstructured ...

GPT just keeps getting better at mathematics, increasingly solving the trickiest of problems. In January, AI testing company Epoch AI found that a previous version of the AI model, GPT-5.2 Pro had ...

OpenAI has released GPT-5.2, claiming significant gains in the AI model’s ability to complete real-world business tasks to an “expert level” compared to GPT-5.1, released in November. The new model, ...