Adobe is Buying Photos and Videos to Train its New AI Model
In the past, Adobe has added generative AI features to its impressive tools like Photoshop and Illustrator.
Adobe has begun buying photos and videos from photographers and artists to train its new AI text-to-video generator as it plans to rival other competitors like OpenAI.
The software giant offers between 6 cents and 16 cents for each photo and, on average, about $2.62 per minute of submitted video, although it could be as much as $7.25 per minute. The company is requesting over 100 clips of photos and videos showcasing people going about their daily activities, from walking to dancing and expressing emotions such as joy and anger, according to a report from Bloomberg.
Adobe also wants simple shots of people's feet, hands, and eyes and is interested in videos of people “interacting with objects” such as smartphones or fitness equipment. But the company has warned photographers and artists against submitting copyrighted materials, nudity, or offensive content.
Emmanuel Oyedeji
In the past, Adobe has added generative AI features to its impressive tools like Photoshop and Illustrator and has released more tools that can perform various functions that have been used billions of times.
But, with companies like OpenAI launching its video-generation technology called Sora, the company fears the new technology could disrupt its position as a leader in the software space. For this reason, it has decided to launch its video-generation technology, with plans to release more updates later this year.
Emmanuel Oyedeji
