Imagine getting a video made just from a description. That's what Adobe's latest tool, Firefly Video Model, can do. This AI-powered text-to-video generator creates brand new videos based on written prompts. Other companies like OpenAI and Google have similar tools, but Adobe claims theirs is special. They trained Firefly Video Model using only licensed content, which could help avoid some of the tricky ethical and copyright issues other AI tools face.

    Adobe is calling Firefly Video Model "commercially safe" because of its licensed training data. This means it might be more suitable for professional use. Adobe hasn't said when the tool will be available to everyone, but they're letting people on a waiting list try it out during a beta test.

    The new model has been in development since at least April 2023 and builds on techniques Adobe used for their Firefly image synthesis models. Like their text-to-image generator, which is now part of Photoshop, Adobe hopes to make Firefly Video Model useful for video creators and editors. They say the model can create footage that looks like it was traditionally made.

    Isn't it fascinating how AI can blend with our creative tools? It's changing how we make things, from images to videos. But it also raises questions about how we train these tools and what kinds of content they should be able to create.

Adobe's New AI Tool Makes Videos from Text

Actions