Earlier this week, Adobe announced the Firefly Video Model, the company’s foray into the generative AI business. Adobe revealed earlier this year that it was looking to join the likes of OpenAI and Google by developing its own text-to-video program, with the possibility of inserting an AI video model into Premiere Pro. This week’s announcement confirms that intention, with Adobe exploring generative AI’s capabilities across its video production platforms.
Interestingly, Adobe is calling Firefly Video Model “the first publicly available video model designed to be commercially safe,” due to its implementation of watermarks, along with being trained exclusively on public domain or licensed content. The Firefly Video Model will have a beta testing phase later this year, and interested parties can sign up on Adobe’s waitlist now.
The following was originally published April 16, 2024:
It had to happen eventually. Adobe is looking into adding generative AI tools to Premiere Pro, powered by Adobe Firefly. Interestingly, the company is also considering integrating third-party generative AI models like OpenAI’s Sora or Pika Labs into Premiere Pro to give users “more choice” when it comes to dreaming up new videos via text prompts. Concept capabilities showcased in Adobe’s video on the subject includes adding and removing objects from videos, as well as extending the length of shots. Adobe is quick to point out that any footage shown in their video is in the early concept stages, but it opens up the possibilities of what might be possible within the Adobe ecosystem, especially if it decides to move forward with third-party integration. As far as AI tools from Firefly being added to Premiere Pro, Adobe says they will be released sometime this year.
See also: WATCH: Adobe’s project allows anyone to create and tweak music using AI
