“Google announced a new Omni model that can take any type of input, such as voice, and generate any type of output, such as video or images.”