Back to Writing
Operating Question
"The question is not which brain is better. It is which thinking system is better designed for what the situation demands."
High Stakes
Low Stakes
← Preserve Thinking
Enhance Thinking →
01 · High Stakes · Preserve
Cognitive Integrity
Protect the judgment. Own the reasoning.
Thinking Type
Deliberate · Prefrontal · Explicit
AI Role
Mirror only — reflect your reasoning back. Do not synthesize. Do not smooth.
Routing Principle
This is where intuitive, socially-mediated cognition most frequently fails under complexity. Explicit reasoning is non-negotiable.
Board decisions
Executive narrative
Irreversible calls
Principled dissent
02 · High Stakes · Enhance
Cognitive Leverage
Expand the structure. Surface what you cannot see alone.
Thinking Type
Deliberate + Structured · Pattern Recognition
AI Role
Scaffolding — surface contradictions, map options, externalize the reasoning architecture.
Routing Principle
AI lowers the translation cost of complex reasoning. This is the quadrant where deep pattern-based thinking produces the highest return.
Scenario mapping
System diagnosis
Risk architecture
Strategic whiteboarding
03 · Low Stakes · Preserve
Cognitive Neutral
Low cognitive cost. Minimal routing required.
Thinking Type
Administrative · Minimal Load
AI Role
Optional assist. The cognitive cost of using AI here may exceed the cost of not using it.
Routing Principle
Neither preserve nor offload carries meaningful cost. The discipline here is recognizing it — and not letting it drift into Q1 territory.
Quick notes
Personal reminders
Informal check-ins
04 · Low Stakes · Enhance
Cognitive Offload
Delegate execution. Reclaim bandwidth for Q1 and Q2.
Thinking Type
Execution · Automatic · Repetitive
AI Role
Generate and complete. Full delegation is appropriate. Human review, not human creation.
Routing Principle
Social translation overhead is highest here. Offloading it is not laziness — it is the act of protecting deliberate cognition for where it is irreplaceable.
Email drafting
Summaries
Formatting
Repetitive reporting
Preserve
Keep thinking in human hands. AI functions as a mirror, not an author. The quality of the output depends on the quality of the reasoning — not the quality of the prompt.
Used in: Q1 (primary) · Q3 (optional)
Enhance
Augment thinking with AI scaffolding. Human direction remains central; AI structures, surfaces, and expands — but does not decide. The human remains accountable for the reasoning.
Used in: Q2 (primary)
Offload
Delegate to AI. Reserve human attention for higher-stakes cognitive work. The discipline is not in how well you offload — it is in how precisely you identify what qualifies.
Used in: Q4 (primary)
🧠
The Karp Reframe
What Palantir's CEO identified was a system failure — organizations that rely on implicit, socially-mediated cognition are fragile under complexity.
He found people who defaulted to explicit reasoning and called it a hiring advantage.
The correct conclusion: design a system that routes thinking correctly, and you stop filtering for the person — you start producing the outcome from anyone capable of that work.