Adobe has just joined the rise of Generative Artificial Intelligence with the official launch of Firefly, thus contributing to the increase in the creative capabilities driven by Artificial Intelligence that it offers to its users.
Firefly starts its journey using its first generative AI model in beta, supporting two new web-based tools, although the company plans to establish Firefly integrations to its line of creative applications such as Photoshop, Premiere or Illustrator in the future, which will allow you to save hours of work, obtaining the results in a matter of seconds.
For example, it could allow vectorization of sketches to work directly with them in Illustrator, avoiding the need to use specific tools for vectorization of images.
Firefly bringing new tools to life
Firefly is behind both a newly created tool, which allows the creation of images from scratch based on text descriptions, including partial generations in different areas that will be selected in the same image, as well as another new tool creation that allows you to style the texts based on the descriptions that are entered, being able to use already integrated style options.
Respecting and collaborating with the community of creators
The company takes a creator-friendly approach, where unlike other companies, it uses images that are not copyrighted or licensed or use rights, including the Adobe Stock library, to conduct training for developers. Models.
Adobe is open to working closely with creators, planning for them to receive financial compensation for allowing their images to be used in training Firefly models, although this scenario will come shortly before Firefly is released. beta.
Developing new security mechanisms
In addition, creators will also be able to contribute to training with their personal styles through Photoshop in the future, although Adobe is already planning mechanisms to prevent the impersonation of personal creative styles by third parties, involving the use of Behance, its network. social, to establish the comparison of images, although at the moment there is no established dynamic.
And for those works in which you do not want to allow their use for training, Adobe is working on a mechanism that allows you to establish metadata in the images so that the training systems take them into account and do not use said works.
Easier to set descriptions
And taking into account creative concerns, the new tools include experience that facilitates the development of the descriptions to be introduced through various integrated options, trying to further differentiate themselves from the competition, although creators will continue to have full freedom to establish the descriptions that they consider most appropriate to the results they wish to obtain.
More Information/Image Credit: Adobe Firefly