Adobe has shown several experimental creative tools that replace some prompt-only workflows with more direct forms of control. Project Sight Span turns a flat photo, painting or illustration into a generated 3D-like space that a user can navigate and reframe from a new viewpoint.
Project Screen Test lets creators build an AI character by speaking with it as if conducting an audition. They can define personality, voice, history and movement, then generate animated tests in described settings. Another experiment, Rinse Repeat, copies a detailed edit from one person or object and adapts it to similar targets in the same image. Project Sound Bender offers a visual canvas for mixing generated sound variations synchronized to video.
These projects were demonstrated as Adobe Max “sneaks,” not promised product releases. Sight Span necessarily generates content outside the original frame, and Screen Test remains a standalone exercise with no confirmed path into Adobe’s video products. The previews are useful as interface experiments, but availability, pricing and production reliability will depend on whether Adobe develops them further.