Adobe Ventures Into AI Video Generation to Compete with OpenAI's Sora

Software giant Adobe is taking steps to enhance its video capabilities and catch up with Sora, the text generation video technology developed by OpenAI. In an effort to develop artificial intelligence video generation tools, Adobe has initiated a program where it purchases video materials from its community of photographers and artists.

Under this program, Adobe is offering $120 to photographers and artists who submit videos showcasing everyday activities such as walking and various emotions like joy or anger. These videos will be used to train artificial intelligence models. Over the past year, Adobe has been integrating generative AI technology into its Photoshop and Illustrator software, and their image and illustration generation tools have already been widely used.

The recent demonstration of OpenAI's video generation model, Sora, has raised concerns among investors who worry that Adobe's dominant position in the creative software industry could be threatened by new technologies. In response, Adobe has stated that it is actively developing its own video generation technology and will share more details later this year.

To collect a diverse range of data, Adobe has requested photographers and artists to submit more than 100 short videos showcasing people in motion and expressing emotions. Additionally, they are seeking close-up shots of body parts like feet, hands, and eyes, as well as videos depicting people interacting with items such as smartphones or fitness equipment. The company has emphasized that submitted material must not contain copyrighted content, nudity, or any offensive material.

Compensation for submitted videos varies based on their length, ranging from an average of approximately $2.62 per minute to a maximum of $7.25 per minute.

An Adobe spokesperson has confirmed that the company's executives are committed to developing video generation capabilities. This move underscores the importance of substantial amounts of data in training AI models. However, there is ongoing controversy regarding the origin of such data. Mira Murati, OpenAI's Chief Technology Officer, mentioned in a recent interview that she is uncertain whether Sora was trained using user-generated videos from platforms like Google's YouTube and Meta's Facebook and Instagram.