OpenAI’s DeepResearch can complete 26% of ‘Humanity’s Last Exam’ — a benchmark for the frontier of human knowledge
OpenAI’s o1 and DeepSeek’s R1 models, which previously sat atop the leaderboard, could only get through roughly 9% of the ...
OpenAI’s new ‘deep research’ agent is still just a fallible tool – not a human-level expert
OpenAI has rolled out a new feature on ChatGPT that can produce reports at the level of a ‘research analyst’. But it doesn’t ...