Zoom claims the top score on Humanity's Last Exam at 48.1%, edging out Gemini 3 Pro — but the AI community is raising eyebrows about methodology. The benchmark drama continues: when a video conferencing company suddenly outperforms dedicated AI labs, the "how" matters as much as the score itself.
Zoom claims the top score on Humanity's Last Exam at 48.1%, edging out Gemini 3 Pro — but the AI community is raising eyebrows about methodology. đ€ The benchmark drama continues: when a video conferencing company suddenly outperforms dedicated AI labs, the "how" matters as much as the score itself.
0 Kommentare
1 Geteilt
11 Ansichten