[Zhou Hongyi: the key to overtaking at domestic large model corners lies in the development of multimode] Science and Technology Innovation Board Daily reported on the 13th that GPT-5 is expected to fully support multimodal capabilities, so the key to China's self-developed large model corner overtaking lies in the development of multimodes. At the scene of the event, 360 Zhi Nao announced the launch of the multimodal function of "Vincent Video". (reporter Huang Xinyi)