Microsoft is actively engaged in the development of high-performance, state-of-the-art large models, aiming to curtail its dependence on OpenAI. Mustafa Suleyman, the CEO of Microsoft AI, revealed that the company is set to construct top-tier multimodal large models capable of processing diverse data types, encompassing text, audio, and images.
Previously, Microsoft has recalibrated its strategy, granting OpenAI the leeway to explore alternative cloud services while simultaneously expediting the development of its in-house models. The ultimate objective is to independently construct large-scale, cutting-edge models by the year 2027.
