December 5, 2024 – OpenAI CEO Altman has announced on social media that the company will be hosting a series of 12 live streams, each occurring on a weekday, featuring a mix of product launches, demonstrations, major events, and even some holiday surprises.
According to sources familiar with OpenAI’s plans, as reported by The Verge, the lineup of new products includes the highly anticipated text-to-video tool, Sora, along with a new reasoning model. The inaugural live stream is set to kick off tomorrow, although the specific details of what will be unveiled have not yet been officially confirmed. Nevertheless, The Wall Street Journal previously suggested the possibility of Sora’s release before the end of 2024.
Sora made its initial debut earlier this year and was shared with a select group of testers. However, as reported by The Washington Post, approximately 20 artists, protesting against “uncompensated labor,” leaked the model to the public. OpenAI has faced widespread criticism regarding the data sources used to train its language models, and Sora is no exception. The company has not directly addressed whether Sora was trained using public YouTube videos, an action that YouTube’s CEO has stated would violate the platform’s terms of service.
OpenAI is not the only player in the text-to-video model arena. Google has also introduced its latest generative AI video model, Veo. Initially launched in May—three months after OpenAI announced Sora—Veo is currently available for private preview through Google’s Vertex AI platform.
