Problem
The AI system produced an incorrect or partially incorrect result/output and the user needs to edit, correct, refine, or recover the system’s behavior.
Solution
Enable the user to modify the AI system’s output by editing, correcting, or refining it.
Enable the user to edit all parts of the AI system’s output.
Use when
- The AI system is fully or partially wrong.
- The AI system produces outcomes the user might need to refine.
How
Enable editing and correction through two methods:
- Using existing user input elements/controls: Treat the AI output as if it were created by the user; allow the user to edit as if it were their own content.
- Using interactions designed specifically for correcting the AI system.
User benefits
- Increases efficiency by enabling the user to build upon the system’s output.
- It can help the user feel in control.
- It avoids the user settling for suboptimal outcomes.
- Corrections indicate the system may not be perfect and give the user insight to how well the system can do what it can do (G2)
- When corrections are used to help the system learn (Guideline 13), a better user experience can be provided over time.
Common pitfalls
- It’s unclear to the user that they can correct the AI system.
- It’s unclear to the user how to correct the AI system.
- In some contexts, it might be inefficient, dangerous, or disruptive for the user to correct the AI system.
Keep in mind that repeated correction of the AI system can be costlier to users than performing the task themselves. If the AI needs frequent correction, consider disabling the feature or making it possible for users to disable the feature (see Guideline 17). Also, the AI system should learn from this behavior so it can improve over time (see Guideline 13).
