“OpenAI's future model development is focused on multimodality, aiming to create robust world representations by enabling models to process text, images, and sounds.”