Feature Request: Text prompt-based object selection ("Select by Description")

Overview: Add the ability to select objects in a document using a natural language text prompt — for example: "blue shirt", "all yellow cars", "the person on the left". This would allow selecting specific objects by their visual description rather than manually using lasso, color range, or other existing tools.

Requested functionality:

• Select a single object by text description
• Select multiple matching objects simultaneously (e.g. "all red apples")

Why this is different from existing tools:

• Select Subject — selects the dominant subject automatically, no control over which object
• Color Range — selects by color only, not by object type or context
• Object Selection Tool — requires manual click/drag, no text input

None of the current tools allow describing what you want to select in natural language.

Use case example: A retoucher working on a product photo wants to isolate a specific clothing item — for example a brown jacket — without manually tracing it. Typing "brown jacket" should produce an accurate selection ready for masking or adjustment.

Suggested implementation: A text input field accessible from the Select menu or toolbar. The feature could leverage Adobe Firefly's existing vision AI capabilities already present in Photoshop.

Expected result: Photoshop generates an accurate selection based on the text description, which the user can then refine using existing selection tools.