I tried GPT-5.4, and most answers were really good - but a few had me concerned

I tried GPT-5.4, and most answers were really good - but a few had me concerned ...

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

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

I ran 7 real-world prompts on Gemini 3 and Claude Sonnet 4.6 — the results surprised me

I tested Gemini 3 Flash and Claude Sonnet 4.6 with 7 real-world prompts to see which AI assistant performs better for ...
The New York Times

About New York Times Games

Since the launch of the Crossword in 1942, The Times has captivated solvers by providing engaging word and logic games. In 2014, we introduced the Mini Crossword — followed by Spelling Bee, Letter ...
The Debrief

Researchers Create ‘Humanity’s Last Exam’ to Test the Limits of Artificial Intelligence

Tests that once challenged advanced AI models are now being solved with ease, making it harder for researchers to pinpoint what current systems are actually capable of.