.jpg)
Microsoft is now sharing their latest research in the arena of artificial intelligence, where the latest model they developed can understand a picture, and content that is in visual form.
This at the same time shows the next development in the arena of artificial intelligence. Previously, users usually had to provide text-based input to ensure artificial intelligence understood something. Through this new development, artificial intelligence can understand and learn a context visually.
Microsoft named this development as Kosmos-1. This artificial intelligence can also read text in pictures, write captions for pictures, and even take a visual-based IQ test. For now, the accuracy rate is still low, but it is expected to be improved day by day.
With this step, Microsoft researchers believe that the development of multimodal artificial intelligence is something important before being able to achieve the development of AGI (Artificial General Intelligence) that can perform tasks at a human level.