Adobe AI Experiments: Interactive Characters and Transforming Images into 3D Worlds
During its annual Adobe Max 2026 conference, Adobe unveiled a series of experimental projects known as "Sneaks," which are technologies aimed at redefining how creators interact with design and editing tools using generative AI. These tools are not yet final products, but rather a showcase of future capabilities that may later be integrated into the Firefly platform or Creative Cloud applications.
Developing Characters and 3D Environments
One of the most notable experiments is Project Screen Test, which aims to give visual content creators greater control over AI-generated characters. Instead of relying solely on text prompts, this project allows the user to "talk" to the character and direct them in an environment similar to auditions. The creator can define character traits such as anxiety or shyness, then engage in an interactive conversation to determine their backstory and personality, resulting in characters that are more consistent and deep compared to traditional models.
In a different context, the company introduced Project Sight Span, which has the ability to transform flat images (whether photographs or artworks) into 3D environments that can be navigated. This technology allows for repositioning the camera and changing viewing angles, as the AI generates the parts not visible in the original image to complete the scene naturally, while maintaining the original artistic style.
Advanced Tools for Audio Processing and Automated Editing
In the field of audio engineering, Adobe showcased Project Sound Bender, an audio mixing tool based on text descriptions. A user can upload a video clip and describe the desired effects (such as "sci-fi boiling sounds"); the system then provides a "Sound Canvas" that allows the user to control the intensity and balance of different sounds by dragging a slider in real-time.
To improve workflow efficiency in Photoshop, the company introduced Project Rinse Repeat. This tool allows for copying edits and applying them automatically to other elements in the image; if a designer retouches a person's skin, the tool can apply the same adjustment to the rest of the people in the image, or add consistent shadows to a group of different products, while ensuring these edits remain manually adjustable using native Photoshop tools.
What Does This Mean for Creators?
The philosophy behind these experiments centers on bridging the gap between the randomness that can result from AI and the precision required by professionals. Instead of just "generating" content, Adobe seeks to give creators Precise Control over the results.
- Shifting from Generation to Direction: The new tools shift the designer's role from merely a writer of prompts to a director who guides characters and controls camera angles and audio details.
- Simplifying Complex Tasks: Technologies like Rinse Repeat aim to automate repetitive and tedious tasks, allowing professionals to focus on major creative aspects.
- Future Challenges: These technologies remain in the experimental stage, which raises questions about the extent of their integration into commercial products and how they will impact the stock library market and traditional talent.