“Within the next 12 months, AI models are expected to see transformative improvements in their ability to use images and video as inputs and outputs for productivity and educational applications.”