United States: Industry leader Adobe has unveiled its plan to incorporate generative AI technology into its renowned Photoshop software. While addressing concerns about potential job displacement and the proliferation of manipulated content, Adobe aims to harness the power of AI to enhance and augment the capabilities of its popular program.
Adobe is set to integrate the cutting-edge generative AI product, Adobe Firefly, into its suite of tools. The recently launched beta version of Firefly will serve as a collaborative tool, working in tandem with graphic designers as a valuable “co-pilot” rather than a complete replacement for human creativity.
Introducing an innovative addition to Photoshop, users will now have access to the powerful “generative fill” feature. Leveraging text prompts akin to those used by renowned AI models like Dall-E and Midjourney, Photoshop users can seamlessly augment their images by adding, expanding, or removing unwanted elements. Whether it’s envisioning a “long-haired dachshund with long flowing rainbow hair” or any other imaginative concept, this groundbreaking feature empowers users to unlock a whole new level of creative possibilities within their image editing endeavors.
Users will get a taste of the highly anticipated generative fill feature as Adobe rolls out the desktop beta version. While initially available to a select group, a broader release is scheduled for later in 2023. This exciting update signifies Adobe’s commitment to refining and perfecting the generative fill functionality before making it accessible to a wider audience.
Adobe has been at the forefront of utilizing AI technology within its creative tools, exemplified by features like the background replacement tool in Photoshop. However, the integration of Firefly into Photoshop represents a significant expansion of capabilities. This integration opens up exciting possibilities, enabling users to create images and templates entirely from scratch, while also facilitating faster and more efficient edits to photographs.
The advent of image-generating AI has raised concerns among artists regarding the potential infringement of copyrighted artworks. To address these concerns, Adobe has taken a proactive approach with the training of Firefly AI. Instead of relying on copyrighted material, Firefly has been trained on an extensive dataset comprising 100 million images sourced primarily from Adobe’s stock library. Additionally, the training dataset includes images from the public domain, where copyright protections have expired.