NATARAJAThe Stairway Effect
Why Intuition Requires Calibration
Core Thesis
Intuition does not appear suddenly.
It develops when predictions repeatedly encounter consequences.
Over time, explicit reasoning becomes less visible.
Anticipation becomes more reliable.
The governance objective is not the preservation of intuition itself.
Intuition is the observable manifestation.
The capability being preserved is pre-explanatory judgement: the ability to reach reliable judgement before complete explanation becomes available.
As organisations increase their reliance on AI and automation, a quiet but material risk emerges: the erosion of human calibration opportunities. When humans stop personally testing predictions against consequences, pre-explanatory judgement degrades, even as information access improves.
The Stairway Effect
Consider walking down a staircase.
The first few steps require conscious attention.
Soon afterwards, movement becomes automatic.
The body predicts:
- Where the next stair will be.
- How far the foot must travel.
- How balance must be adjusted. The prediction occurs before conscious reasoning.
When the pattern changes unexpectedly:
- Prediction fails.
- Movement becomes unstable.
- Correction occurs immediately. The stumble is not the failure.
The stumble is the calibration.
Most intuitive capabilities require years to develop and are difficult to observe while forming.
The staircase effect makes the mechanism visible in real time.
From Prediction to Intuition
Repeated encounters with reality gradually transform explicit reasoning into intuitive capability.
Compression occurs when reasoning, tested across many consequence-bearing encounters, becomes embedded into anticipatory behaviour.
Over time:
- Fewer calculations are consciously performed.
- Fewer rules are explicitly recalled.
- Anticipation replaces analysis. The individual no longer reasons through every step.
They simply know.
The reasoning has not disappeared.
It has been compressed.
Intuition becomes relevant to governance because it enables pre-explanatory judgement, reliable action that precedes full articulation of reasoning.
The Distinction from Craftsmanship
Craftsmanship explains how expertise is developed.
The Stairway Effect explains how expertise becomes intuitive.
Craftsmanship emerges through engagement with reality.
The Stairway Effect describes how repeated calibration compresses that engagement into anticipation.
The governance challenge is preserving the conditions under which pre-explanatory judgement can continue to form.
Flight 1549 vs Flight 447
In 2009, Captain Chesley Sullenberger successfully landed US Airways Flight 1549 on the Hudson River after both engines were disabled by a bird strike.
Observers often describe the event as intuition.
It was not created in the crisis.
Years of training and consequence-bearing decisions had already calibrated the capability.
The emergency revealed the intuition. It did not create it.
Contrast this with Air France Flight 447.
When the aircraft's speed sensors froze, the autopilot disengaged.
The pilots possessed all the manuals and explicit training.
But after years of relying on automated systems, their manual calibration had eroded.
Faced with contradictory data, their pre-explanatory judgement failed.
They stalled a perfectly healthy aircraft into the ocean.
They possessed the information.
They lacked the calibrated judgement.
The Intuition Erosion Risk
Most organisational learning assumes capability develops through information, instruction, procedures, and training.
These are necessary. They are not sufficient.
A simulator where an operator can press "Restart" carries no consequence.
Without consequence, there is no calibration.
As organisations increasingly rely on AI recommendations, autonomous automation, and algorithmic optimisation platforms, they unintentionally reduce opportunities for humans to test predictions against real consequences.
The organisation becomes more informed.
It may become less intuitive.
An organisation can accumulate information long after it stops accumulating judgement.
This failure remains invisible while the automation works.
It becomes visible only when conditions change and no calibrated human capability remains to recognise the change.
CEO & Board Mandate
The question is no longer:
"How much information do we possess?" It becomes:
"What still calibrates human judgement inside this organisation?" Capability formation requires exposure to consequences.
Organisations must preserve environments where:
- Predictions can be manually tested
- Consequences remain visible to human operators
- Judgement can be calibrated
- Pre-explanatory judgement can continue to form
Closing Insight
We learn to walk the stairs because reality continuously corrects us.
Over time, those corrections become compressed into intuition.
The same mechanism shapes leadership, judgement, negotiation, and decision-making.
Intuition is not knowledge.
It is anticipatory capability that emerges from repeated calibration by consequences.
Pre-explanatory judgement is the capability that intuition makes possible.
The question is not whether intelligence can predict.
The question is whether intelligence can develop reliable judgement before complete explanation becomes available.
Governance constraints. Guarantees endure.
ACTION: Develop The Algorithmic Air-Gap
Organisations must actively prevent the algorithmic de-skilling of their strategic workforce.
- Mandate Consequence-Bearing Manual Execution: Critical operators must be periodically required to execute complex, live-fire workflows without algorithmic assistance. Simulation is insufficient.
ACTION: Develop The Algorithmic Air-Gap
- Protect the Stairway: Identify which forms of pre-explanatory judgement will disappear if AI fully automates the feedback loop, and deliberately architect friction back into the process.
Do not allow the enterprise to become a passenger in its own operation.
Governance constraints. Guarantees endure.
Hadi Hendrawan
Advising CEOs on AI Risk, Authority & Accountability
June 2026
- X: @hhwan888
- LinkedIn: https://www.linkedin.com/in/hhwan888
Selected References
1. Carl Jung (1921), Psychological Types
2. Herbert Simon (1979), Models of Thought
- Francisco Varela, Evan Thompson and Eleanor Rosch (1991), The
Embodied Mind: Cognitive Science and Human Experience
Gary Klein (1998), Sources of Power: How People Make Decisions 5. Karl Friston (2010), The Free-Energy Principle: A Unified Brain Theory?
Ruben Eero Laukkonen and Shamil Chandaria (2024), A Beautiful Loop:
An Active Inference Theory of Consciousness
- Hadi Hendrawan (2024), The Stairway Effect: Chaos and Cooperation toward Sustainable Agency of Humans and AI
SCHEDULE A: Intuition, Calibration, and
Future Agency
Core Principle
Intuition refers to anticipatory capability that emerges from repeated calibration by consequences.
The Stairway Effect demonstrates one observable mechanism through which this capability is formed.
Prediction Is Not Intuition
Prediction estimates what may happen.
Intuition anticipates what is likely to happen without explicit reasoning.
Prediction can be transferred through information.
Intuition must be calibrated through consequences.
The Calibration Loop
Prediction → Action → Consequence → Calibration.
Repeated calibration from consequence gradually compresses explicit reasoning into anticipatory capability.
This capability is experienced as intuition.
Intuition Is Not Knowledge
Knowledge can be documented, transferred, and stored.
Intuition cannot be fully transferred. It must be formed.
Can AI Develop Intuition?
The Stairway Effect does not imply intuition is uniquely human.
It implies intuition requires calibration.
Future AI systems may develop intuition-like capabilities through continuous reinforcement learning across rich calibration environments.
High-Dimensional Calibration, a Governance
Challenge
When humans calibrate intuition, they do so in sequential, physical reality.
When an AI calibrates, it does so across thousands of mathematical dimensions simultaneously.
The resulting pre-explanatory judgement may be mathematically valid but structurally alien to human logic.
An AI may execute a strategic decision that "feels right" to its architecture yet defies human reasoning.
If the AI cannot explain its reasoning, and its high-dimensional intuition operates in dimensions humans cannot comprehend, governance becomes structurally difficult.
The Relevant Governance Question
The question is no longer whether the AI can predict accurately.
The question is:
How do you constrain a machine whose intuition you cannot comprehend?
Governance constraints. Guarantees endure.
Hadi Hendrawan
Advising CEOs on AI Risk, Authority & Accountability
June 2026
- X: @hhwan888
- LinkedIn: https://www.linkedin.com/in/hhwan888 These supplements translate the framework into enforceable governance actions. They must be adapted to organisational context and regulatory constraints.
SUPPLEMENT 1: Supplemental Scenario:
The Ghost Override and The Financial Cost of Atrophied Judgement Note: This supplement illustrates the consequences of eroded human calibration when confronted with advanced AI systems. It should be read as a dramatised scenario, not as a literal case study.
Core Principle
When an algorithm assumes the consequences, human intuition erodes.
The Scenario
A Tier-1 global energy trading firm deploys an autonomous predictive hedging agent.
For four years, the AI flawlessly navigated spot-market volatility.
It handles 99% of all execution.
The human trading desk transitions from market participants to system monitors.
They watch the dashboard. They approve outputs.
They no longer feel the market. They no longer walk the stairs.
Their calibration quietly halts.
The Event
A severe geopolitical shock disrupts three major LNG supply routes simultaneously.
The market violently fractures.
In response, the AI initiates a massive, highly counter-intuitive short position on a seemingly unrelated renewable energy index.
To the human monitors, this move defies all conventional economic logic.
The AI has accessed its high-dimensional calibration.
It sees multi-variable correlations across global shipping, weather derivatives, and grid storage that human sequential logic cannot comprehend.
It cannot explain its pre-explanatory judgement. It simply executes.
The Reaction
The Human Risk Committee panics.
They execute the Algorithmic Severance protocol and pull the manual override.
The humans take the wheel.
They revert to their "gut feeling" and execute a traditional legacy hedging strategy.
The Catastrophe
The human strategy fails instantly.
The firm loses $850 million in 72 hours.
Post-mortem analysis reveals:
The AI's counter-intuitive trade was mathematically perfect.
Had it been allowed to execute, the firm would have cleared a $1.2 billion profit.
The disaster was not caused by the machine failing.
The disaster was caused by the humans trying to fly the plane manually.
Their "gut feeling" was a ghost.
Because they had not made a consequence-bearing trade in four years, their intuition was calibrated to a market that no longer existed.
They possessed the information.
They lacked judgement.
The Boardroom Reality
The Board realises they have architected a dangerous configuration:
An AI that is too complex to explain itself, monitored by humans who are too de-skilled to override it safely.
When you pull the kill-switch, you are betting the survival of the enterprise on the intuition of your human operators.
If that intuition has not been continuously calibrated by real consequences, the manual override is not a safety mechanism.
It is a suicide switch.
Closing Insight
You cannot build an autonomous enterprise and expect your humans to retain their edge in a vacuum.
If you do not deliberately force your strategic workforce to walk the stairs, they will fall the moment the elevator breaks.
Governance constraints. Guarantees endure.
Hadi Hendrawan
Advising CEOs on AI Risk, Authority & Accountability
June 2026
- X: @hhwan888
- LinkedIn: https://www.linkedin.com/in/hhwan888