tool
Correction affordance for AI output
AI output should be easy to edit, refine, undo, or recover from when it is wrong.
正文
Correction affordance means the product assumes AI output may be partly wrong and gives the user a direct path to recover. The interface can support editing, regeneration, undo, switching classifications, or saving a corrected version for later learning.
For idea tools, the key point is not to force users to accept or discard a whole generated result. Let them preserve the useful part, fix the weak part, and feed that correction back into review.
来源引用
Guideline 9: Support efficient correction
Source: Guideline 9: Support efficient correction
Microsoft HAX Guideline 9 recommends making it easy to edit, refine, or recover when an AI system is wrong.
相关卡片
Context-level accessibility testing
Accessible components still need testing in the actual page or workflow context where users encounter them.
Server-mirrored validation
Client-side validation can improve usability, but server-side validation must enforce the actual rule.
Adaptive AI feedback loop
A human-centered AI product should collect useful feedback during interaction and use it to improve future behavior.
所在阅读路径
Human-centered AI feedback loop
A route for keeping AI tools useful after first generation: map risks, support correction, capture feedback, and review production readiness.