AI Incentives · Workflow

Model Distillation

Train a smaller model to imitate a larger one.

"Compress the wisdom. Keep the speed."

Quick answer

What is Model Distillation? Train a smaller model to imitate a larger one. Cost and latency optimization.

In the wild

Edge deployments of distilled LLMs.

Why it matters in the room

Cost and latency optimization.

Counter-move

Distillation often beats fine-tuning at scale.

Visual · Reward gradient

Model Distillation shows where an optimizer climbs vs where we want it to go.

Live example · Train around Model Distillation

Pick what to reward the model for.

Train a smaller model to imitate a larger one. In the wild: Edge deployments of distilled LLMs.

● Live

Pick a lever. There are no neutral ones — every incentive funds a behavior somewhere.

How does this land?

Pick a reaction to Model Distillation

One tap. We'll point you at the most useful next surface based on how this hits.

Human Behavior Element™ · HBE Spec

The full taxonomy entry

Every concept in the Atlas uses the same structure — so Model Distillation can be compared, recombined, and cited like an element on a periodic table.

About the standard →

HBT-A1838

Official name

Model Distillation

AI Incentives · Workflow

Identity

HBT ID

HBT-A1838

Symbol

Official name

Model Distillation

Synonyms

Workflow

Keywords

AI Incentives, Workflow, human behavior, incentive design

Version

v1.0

Last updated

Maintained by The Incentives Lab

Classification

Kingdom

Systems

Domain

Machine Behavior

Family

AI Alignment & Incentives

Class

Workflow

Element

Model Distillation

Definition

Scientific

Train a smaller model to imitate a larger one.

Plain-English

Train a smaller model to imitate a larger one.

Feynman

Compress the wisdom. Keep the speed.

Core principle

Train a smaller model to imitate a larger one.

One-sentence summary

Cost and latency optimization.

Mechanisms

Psychological

Train a smaller model to imitate a larger one.

Behavioral econ.

Cost and latency optimization.

Neurological

Reward, threat, and salience circuits bias attention toward the cue.

Evolutionary

Heuristics that paid off in ancestral environments now misfire in modern systems.

Sociological

Group norms and status incentives reinforce the pattern across a team.

Computational

Models trained on biased human signals will replicate and amplify the pattern.

Systems thinking

Feedback loops between metrics, incentives, and behavior lock the pattern in place.

Signals & signature

Inputs (activators)

Edge deployments of distilled LLMs.

Outputs (observable)

Cost and latency optimization.

Behavioral signature

You see Model Distillation when the explanation for a decision sounds reasonable but the outcome keeps repeating.

Behavioral molecules

Often combines with related Atlas entries — see the rail below.

Pathways · before

A goal, metric, or contract clause makes the behavior rational locally.

Pathways · after

Locally rational choices accumulate into a systemic distortion.

Domains where it shows up

Business
Leadership
Government
Healthcare
Education
Sales
Marketing
AI
Negotiation
Media
Public Policy
Relationships

Examples

Everyday

Edge deployments of distilled LLMs.

Modern

Cost and latency optimization.

Historical

A pattern repeatedly documented since the foundational behavioral science literature on ai alignment & incentives.

Famous experiments

See the References block — primary papers in the Atlas link out to the original studies.

Design principles

How to leverage

Cost and latency optimization.

How to reduce

Distillation often beats fine-tuning at scale.

How to redesign

Distillation often beats fine-tuning at scale.

The Perverse Incentive Lens™

How it's exploited

Organizations weaponize model distillation — sometimes deliberately, often by accident — when metrics reward the symptom rather than the outcome.

Common perverse incentives

Volume metrics, short review windows, bonus cliffs, and contracts that pay on activity rather than impact.

Failure modes

When Model Distillation dominates, teams optimize for the dashboard while the real outcome quietly degrades.

Incentive redesign

Distillation often beats fine-tuning at scale.

Ethical considerations

Don't engineer model distillation into customers, employees, or citizens as a manipulation tactic — design for informed choice instead.

Diagnostic questions

Where in our org would Model Distillation most often show up unnoticed?
Which metric, ritual, or contract clause quietly rewards Model Distillation?
If we removed every payoff for Model Distillation, what behavior would replace it?
Who benefits when Model Distillation persists — and who pays the cost?

Organizational warning signs

Metrics

A KPI is hit while the underlying outcome stalls or worsens.

Behaviors

People route around the rule rather than challenge it.

Language

'That's just how we do it here.' / 'The system requires it.'

Culture

Naming the pattern is treated as disloyalty.

Red flags

People defend the status quo using the language of model distillation.
Decisions cluster around the easiest narrative rather than the strongest evidence.
New data changes the slide deck but not the decision.
Anyone naming the pattern is treated as the problem.

Intervention playbook

Immediate

Make the perverse payoff visible to the people creating it.

30-day

Run a small pilot that pays for the outcome, not the proxy.

Long-term

Rewrite the comp plan, contract, or ritual so the right behavior becomes the easy behavior.

AI considerations

Detect

Audit training data and reward signals for the same pattern this element describes.

Avoid amplifying

Don't optimize models on metrics that already encode the perverse incentive.

Counteract

Use the model to surface where the pattern is most active, then redesign the incentive — not the model.

Measurement

Metrics

Outcome-to-proxy ratio over time.

Assessment

The Incentives Lab III Diagnostic.

Survey

Calibrated pulse questions on rules vs. outcomes.

Behavioral signals

Where people work around the system.

Observational

Where the dashboard and the lived experience disagree.

Scientific evidence

Evidence grade

Synthesized from the behavioral science literature; see Atlas references.

Replication

Tracked in the Atlas as primary, replicated, or contested.

Intervention confidence

Moderate — patterns generalize, mechanisms vary by context.

Research consensus

Broad agreement on the pattern; ongoing debate on boundary conditions.

Known limitations

Local context, culture, and incentive structure all change the strength of the effect.

Open questions

How does Model Distillation interact with AI-mediated decisions at scale?

References

Primary papers

See the Research Center →

Books

Curated in the Magazine →

Meta-analyses

Tracked in the Atlas registry.

Seminal authors

Kahneman, Tversky, Thaler, Ariely, Cialdini, Ostrom, Simon — and the field they built.

Cross references

Every Atlas entry is a node in a knowledge graph. See the related rail below to follow the connections.

Disciplinary layers

See Model Distillation through 4 lenses

Each layer of the Incentives OS reframes this concept with its own thinkers, vocabulary, and diagnostic question.

Test yourself · 60 seconds

Do you actually know Model Distillation?

Three quick questions. Result is saved into your review streak — come back when the term is due to lock it in.

Question 1 of 3Score: 0/3

Which best describes Model Distillation?

Go deeper

Worked example, counter-example & concept map

On-demand AI analysis grounded in the Lab's research. Cached on your device after first run.

Diagnose

Find this pattern in your org →

15-minute III Assessment.

Learn

Master incentive intelligence →

12-week Executive Course.

Engage

Bring the Lab to your team →

Speaking, retainers, training.

How this lands in you

Your nervous system has a region for this.

Primary region

Prefrontal Cortex →

When you encounter Model Distillation, your prefrontal cortex has to do extra work to override the automatic response — and that override budget is finite.

Executive control, planning, impulse override, working memory, System 2. First thing to go offline under stress, fatigue, or low blood sugar. Why your 4pm decisions are worse than your 9am ones.

See Prefrontal in the Brain Atlas →