L
Intelligent Systems Interaction
Foundations of AI Interaction
48 cards- Mismatch Between Stochastic Output and Deterministic UI 8
- Expressing Capability Boundaries 3
- Visualizing Uncertainty 3
- Presenting Confidence 3
- Human in the Loop 3
- Graceful Degradation of AI Failure 8
- Confidence Displays and Their Misreadings 5
- Intervention Points in the Human Loop 5
- Reproducibility of Non-deterministic Output 5
- Latency and Streaming Experience 5
Prompting and Control
59 cards- Openness of Natural-Language Commands and Its Cost 8
- Discoverability of What Can Be Said 3
- Examples and Template Guidance 7
- Complementarity of Parameter Controls and Natural Language 8
- Iterative Revision 3
- Granularity of Control 3
- Prompt History and Reuse 3
- Discoverability of What I Can Say 5
- Iterative Edits and Local Regeneration 5
- Control Granularity and Editability 4
- Carrying and Clearing Context 5
- Ambiguity and Clarifying Follow-ups 5
Generated Content
53 cards- Generating Multiple Options 3
- Source Attribution 3
- Hallucination and the Fact-Checking Burden 8
- Content Labeling 3
- Copyright and Material Provenance 3
- Streaming Presentation 3
- Multi-option Generation and Side-by-side Comparison 5
- Source Attribution of Generated Results 5
- Labeling and Watermarking AI Content 5
- Streaming the Generation Process 5
- Editing and Taking Over Generated Content 5
- User Feedback Loops on Generation Quality 5
Automation and Agents
57 cards- Levels of Automation 3
- Automation Bias 3
- Automation Complacency 3
- Takeover and Handoff Design 8
- Interruptibility and Rollback 3
- Permission Boundaries of Agents 3
- Pre-action Confirmation 3
- Visibility of Task Progress 3
- Skill Degradation 3
- Pre-action Confirmation and Irreversible Operations 5
- Task Progress and Intermediate-state Visibility 5
- Agent Failure Reporting and Escalation 5
- Plan Visibility and Revision for Multi-step Tasks 5
- Accountability and Traceability 5
Explanation and Trust
48 cards- Types of Explainability 3
- Local and Global Explanations 8
- Trust Calibration 3
- Collapse of Trust 3
- The Moderation Principle of Transparency 8
- Risks of Anthropomorphism 3
- Counterfactual Explanations 5
- Overtrust and Trust Collapse 5
- Asymmetric Effect of First Failures on Trust 5
- Expectation Inflation from Anthropomorphism 5
Personalization and Recommendation
53 cards- Recommendation Reasons 3
- Filter Bubbles 3
- Cold Start 3
- Feedback Loops 3
- Turning Personalization Off 3
- Inferred Preferences and Their Limits 8
- Presenting Recommendation Reasons 5
- Filter Bubbles and Diversity 5
- Feedback Loops and Preference Entrenchment 5
- Turning Personalization Off and Resetting It 5
- Conflict Between Personalization and Predictability 5
- Negative Feedback Channels for Recommendations 5