OpenAI announces a new model of their generative pretrained transformer (GPT) named GPT-4o, capable of visual and video speech recognition and translation. (Reuters) link embed save