Knowledge Failure Modes
Taxonomy of what can go wrong in knowledge systems
Knowledge failure modes are the ways that knowledge systems degrade, collapse, or mislead. Understanding these modes is prerequisite to designing defenses.
System-Level Failures
- Encyclopedia Meltdown — knowledge collapse when AI takes initiative without human intervention
- Model Collapse — degradation from training on AI-generated data
- Institutional Brain Rot — organizational verification capacity decay
Input-Level Failures
- AI Slop — low-quality AI content flooding the information environment
- Hallucination — plausible but fabricated AI outputs
Human-Level Failures
- Fluency Trap — mistaking smooth prose for accuracy
- AI-Induced Illusions of Competence — false mastery from AI assistance
- Deskilling Through AI Delegation — losing capacity by not practicing
Semantic Failures
- Semantic Drift — changes in meaning over time
- Definition Drift — meaning shifts over time
- Meaning Loss — claims becoming obsolete as terms change