OPENAI IS PREPARING SORA 2: AN AI MODEL WILL BE TAUGHT TO "SPEAK" IN VIDEOS

Summary from the TLDR Design newsletter

OpenAI is developing Sora 2, a significant upgrade to its text-to-video model that will integrate audio capabilities, including speech and environmental sounds. While the current Sora generates silent videos lasting over 20 seconds, Google's competing Veo 3 produces shorter clips with synchronized audio tracks. The audio-enabled competition between these AI video models could establish new creative standards while raising concerns about potential misinformation.