Sam Altman has reportedly issued a "red alert" after Gemini outperformed ChatGPT in recent benchmarks. According to an internal memo, Altman emphasized the need for OpenAI to enhance the speed and reliability of its main model and expand its topic coverage.
Gemini 3 achieved the highest score ever on the Humanity’s Last Exam, scoring 37.4 compared to GPT-5 Pro's 31.64. Furthermore, Gemini excelled in user satisfaction evaluations like LMArena.
These insights come from Tom’s Hardware and TechCrunch.
