OpenAI’s DeepResearch can complete 26% of ‘Humanity’s Last Exam’ — a benchmark for the frontier of human knowledge

OpenAIs DeepResearch can complete 26 of ‘Humanitys Last Exam —

OpenAI’s o1 and DeepSeek’s R1 models, which previously sat atop the leaderboard, could only get…

fortune.com

More posts

How Argentina took a chainsaw to government, a year before Elon Musk's DOGE

February 12, 2025
PBS shutters its DEI office to comply with Trump order as GOP senator renews calls to defund public broadcasting

February 12, 2025
Currys is the latest UK retailer to let you register interest in Nintendo Switch 2 news and potential pre-order information

February 12, 2025
Diagnostic dilemma: One Direction concert left teen with ‘crackling’ bubbles in her chest

February 12, 2025