Technical, cognitive, and intent debt
martinfowler.com
April 22, 2026
6 min read
🔥🔥🔥🔥🔥
56/100
Summary
Cognitive Debt is a metaphor for the loss of understanding within a team regarding a system's functionality. Margaret-Anne Storey identifies three layers of system health: technical debt in code, cognitive debt in people, and a third unspecified layer.
Key Takeaways
- Technical debt accumulates in code, limiting future changeability and system adaptability.
- Cognitive debt arises from eroded shared understanding among team members, hindering their ability to reason about changes.
- Intent debt develops when goals and constraints are poorly captured, affecting the system's alignment with original intentions and its evolution.
- Cognitive surrender occurs when users rely uncritically on AI-generated reasoning, contrasting with cognitive offloading, which involves strategic delegation during deliberation.
Community Sentiment
Positives
- LLMs can be guided to produce minimal code changes, aligning their output with developer intent and potentially improving code quality.
- AI tools like Claude can emulate the instincts of senior developers, suggesting that they can be effective in optimizing code without unnecessary alterations.
Concerns
- LLMs often generate 'lazy' code, leading to increased complexity and runtime issues, which can hinder the intended abstraction layers.
- There are concerns that LLMs may introduce errors by over-editing or making inappropriate changes, reflecting a lack of understanding of the code context.
