GenerativeCreativeNotUtility [principle] v1.0.0

AI-generated design artifacts should be evaluated on creative and aesthetic quality standards — not merely on whether they function correctly. A UI that works but is generic has failed the creative mandate even if it passes all functional tests.

Implications

• Design review criteria must include aesthetic quality, not only functional correctness.
• A/B test both functional metrics (conversion, error rate) AND subjective quality ratings (design quality score).
• When an AI agent generates 'acceptable but generic' output, treat that as a failure requiring iteration — not a success requiring ship.
• The Prime Wiki system exists specifically to raise the bar from 'acceptable utility' to 'genuine creative quality'.

GenerativeCreativeNotUtility [principle] v1.0.0

AI-generated design artifacts should be evaluated on creative and aesthetic quality standards — not merely on whether they function correctly. A UI that works but is generic has failed the creative mandate even if it passes all functional tests.

Implications

• Design review criteria must include aesthetic quality, not only functional correctness.
• A/B test both functional metrics (conversion, error rate) AND subjective quality ratings (design quality score).
• When an AI agent generates 'acceptable but generic' output, treat that as a failure requiring iteration — not a success requiring ship.
• The Prime Wiki system exists specifically to raise the bar from 'acceptable utility' to 'genuine creative quality'.

Sources

Rationale

Treating AI-generated design as a utility task (does it work?) rather than a creative task (is it good?) is why AI output converges on mediocrity. Functional correctness is the floor, not the ceiling. The ceiling is a design that provokes the response 'this is beautiful' or 'this is exactly right for this brand' — not merely 'this works'.

Source

• Prime Wiki A/B v4/v5 — design quality scoring rubric (0–10 scale)