Models · Emotion, Value and Motivation

Exploration and Exploitation

Two routes between the same two places: one familiar and dependable, one unknown and possibly faster. The choice is not between a good decision and a bad one. It is between obtaining an expected result and obtaining information that might improve every journey after this one.

By Yona Ole Lobulu ·

Model13 min readD7.21

Topic
Emotion, Value and Motivation
Read first
2 pieces should be read before this one
Reading time
About 13 minutes of reading
Difficulty
Late reading: this page sits at the far end of the Library, after many other pieces

The question

When should action use what is already known, and when should it test whether something better is available?

Definition

Exploration and exploitation: two functions of allocated action under uncertainty. Exploitation allocates action primarily towards obtaining an expected result from current knowledge; exploration allocates action partly towards improving or testing the knowledge that later action can use.

Imagine that you regularly travel between the same two places.

One route is familiar. You know approximately how long it takes, where delays usually occur and what to expect along the way.

Another route might be faster, but you have never tried it. Taking it could save time. It could also make today's journey longer.

The familiar route offers a reasonably dependable result. Trying the uncertain route could provide something different: evidence that improves future journeys. Even if the route is slower today, learning about it might reveal a useful alternative.

Whether that evidence is worth acquiring depends on the situation. If this is the final time you will make the journey, it has little future use. If you expect to make the journey hundreds of times, one inconvenient trip may reveal an improvement that can be used repeatedly.

But that still does not make testing automatically worthwhile. Arriving late today might carry serious consequences. The unfamiliar route might be unsafe, inaccessible or affected by conditions that will not apply again.

Every allocation of limited action leaves other possibilities partly or temporarily unrealised. Using one option means giving up at least some of what another might have produced.

This creates a recurring problem: when should action use what is already known, and when should it test whether something better is available?

Two Uses of Action

The exploration–exploitation model describes two ways of allocating action under uncertainty.

Exploitation allocates action primarily towards obtaining an expected result from current knowledge.

In this context, exploitation has a technical meaning. It does not refer to exploiting another person. It means using what has already been learned to pursue an option currently estimated to be sufficiently valuable relative to the available alternatives.

Exploration allocates action partly towards improving or testing the knowledge that later action can use. It may involve trying an uncertain option, checking an assumption or discovering whether conditions have changed.

The difference concerns function rather than appearance.

Trying something new is not automatically exploration. A person may use an unfamiliar service after already learning exactly what it provides. The action is novel to them, but its primary function may still be obtaining an expected result.

Repeating an action is not automatically exploitation. Repetition can test whether an outcome remains stable, whether a changed condition matters or whether an earlier result was accidental.

Learning that happens incidentally during exploitation does not automatically make the action exploratory. The acquisition of information must be part of what the allocation makes possible or relevant, although it need not result from an explicit intention to explore.

Allocation can emerge through learned tendencies and situational pressures without a conscious calculation.

Function also cannot be read directly from novelty, repetition or one isolated choice. It depends on:

  • what is already known;
  • what is expected;
  • what uncertainty remains;
  • what the action could reveal;
  • how its consequences might affect later decisions.

Exploration and exploitation are not fixed kinds of activity or permanent kinds of person. The relevant unit is the function of an action in a particular context, not the character of the person performing it.

Exploitation Converts Learning Into Outcomes

Learning has little practical value if it is never used.

Once an effective option has been identified, exploiting it can provide:

  • reliable outcomes;
  • efficient action;
  • accumulated skill;
  • reduced decision effort;
  • coordination with other people;
  • sustained progress.

A writer who has developed a dependable work routine does not need to redesign it every morning. A team cannot coordinate effectively if every member continually changes the process. A person who has found a training programme that produces progress may benefit from following it long enough for adaptation to occur.

In these situations, repetition is not evidence of rigidity. It is how previous learning becomes performance.

The option being exploited is not necessarily the best possibility that exists. It is one currently estimated to be sufficiently valuable or reliable relative to the alternatives the person can identify and access.

That limitation does not make continued exploitation mistaken. Searching for unknown possibilities also consumes time and resources.

Exploitation becomes rigid when persistence stops responding adequately to relevant evidence and changed conditions—not merely when another option appears promising.

Even then, observed persistence does not reveal its cause with certainty. Switching may be expensive, unsafe or unavailable. A person may lack the time, resources or authority to act differently.

Exploration Improves or Tests What Is Known

Exploration exposes action to uncertainty in ways that may generate relevant evidence.

It can involve:

  • testing an unfamiliar option;
  • comparing variations;
  • checking whether an expectation remains accurate;
  • searching for alternatives;
  • trying a strategy under different conditions;
  • learning from other people's experience;
  • discovering that the current understanding is incomplete.

Some exploration is directed towards information. An uncertain option is selected partly because learning about it may improve future choices.

Exploration can also occur through increased behavioural variability. Sampling a wider range of actions can expose an agent to alternatives that a narrow strategy would never encounter. Research sometimes describes this as random exploration.

Random exploration is a model-based description of increased choice variability. It does not prove that a person deliberately randomised or that the variability was useful. Variation can produce discovery, but it can also reflect noise, inconsistent valuation, misunderstanding or error.

Directed and random exploration are not the only possible forms. People can test hypotheses systematically, observe others, simulate possible consequences or change one feature of an otherwise familiar action.

Exploration should therefore not be confused with indecision.

Indecision may delay commitment without producing new evidence. Information gathering may also stop being useful when additional information cannot change the decision or is never integrated into later action.

A well-motivated test can remain exploratory even when its result is noisy or uninformative. Exploration creates the possibility of learning, not a guarantee of improvement.

Why Information Can Have Value

An action can provide a direct result and evidence that changes later decisions.

The value of that evidence depends on what it might make possible.

Information has prospective value when it can:

  • improve later choices;
  • reveal a better option;
  • reduce the risk of repeating an error;
  • show that a familiar strategy remains dependable;
  • identify conditions under which an option succeeds or fails.

Suppose someone tests a different work process and completes the task more slowly. Judged only by today's output, the experiment performed poorly. But if it reveals one change that improves hundreds of later attempts, its total contribution may exceed its immediate cost.

Information value is estimated before acting. Whether the information actually improves later decisions is known, if at all, only afterwards. An exploration can therefore be reasonable given what was known beforehand and still produce no useful result.

Information is not valuable merely because it reduces uncertainty. It may have little instrumental value when:

  • no future decision can use it;
  • the available actions remain unchanged;
  • the person cannot act on what they learn;
  • it becomes obsolete quickly;
  • it cannot distinguish between relevant alternatives;
  • acquiring it costs more than it can plausibly improve.

People may also want information for reasons beyond its effects on later outcomes. They may value knowing or dislike unresolved uncertainty. Those motives can support exploration, but the exploration–exploitation model does not provide a complete account of curiosity.

A Simplified Representation

A simplified conceptual representation—not a calculation of how people necessarily choose—is:

Action value ≈ expected present return + expected future value of information − costs and opportunity costs

Expected present return concerns what an action is currently believed to produce.

Expected future value of information concerns how evidence from the action might improve later decisions.

Costs include the time, effort, risk and resources required by the action.

Opportunity costs concern what is forgone by selecting it instead of something else.

These considerations can interact rather than combine as precisely measurable quantities. Horizon and environmental stability can change the expected usefulness of information. Capacity and stakes can change the cost of testing. Unknown alternatives cannot be valued accurately before they are explored.

People may not represent any of these factors consciously. The representation is incomplete and does not guarantee identification of the highest-value action. Its purpose is to make the structure of the trade-off visible, and it depends throughout on how options are valued in the first place.

Both Modes Carry Costs

Testing uncertain options can consume:

  • time;
  • attention;
  • energy;
  • money;
  • present reward;
  • safety;
  • predictability;
  • opportunities to practise what already works.

Exploration may also generate information that is inaccurate, too specific to generalise or obsolete before it can be used.

The costs of exploitation can be less visible. Continuing to use a known option may forgo:

  • discovery of better alternatives;
  • correction of outdated expectations;
  • adaptation to changed conditions;
  • knowledge about unexplored possibilities;
  • resilience if the familiar option becomes unavailable.

The two sets of costs need not be equal. Their asymmetry is often what makes one mode more appropriate under particular conditions. Exploratory failure may be irreversible, while missed information can sometimes be acquired later. In other situations, a temporary exploratory cost may prevent years of reliance on a weak strategy.

A locally successful option is the best or sufficient option among those currently represented—not necessarily the best possibility that exists.

But the mere possibility of a better option does not justify endless search. Exploration can consume more value than discovering the alternative would create.

Each mode gives up some of what the other might have provided.

The Horizon Changes the Balance

The decision horizon is the remaining time or number of opportunities over which information can be used.

A longer horizon often increases the potential value of exploration. If a person will make the same kind of choice repeatedly, evidence obtained now can improve many later decisions. The initial cost of testing may be recovered across future outcomes.

Future opportunities increase information value only to the extent that what is learned transfers to them. If later decisions occur under different conditions, the evidence may have limited use.

A shorter horizon often increases the relative value of a dependable present result. If no later decision remains, information that only benefits future action may have little instrumental value.

The horizon does not settle the entire allocation. A long horizon may still favour exploitation when testing is dangerous, a reliable strategy builds compounding skill or stable coordination matters. A short horizon may still justify exploration when current knowledge is unreliable or a small test can prevent an irreversible error.

Time changes the potential usefulness of information. It does not create a universal rule for when either mode should dominate.

Changing Conditions Can Reopen Exploration

Exploitation depends on previous learning continuing to predict current outcomes. What was learned from consequences only remains useful while those consequences hold.

In a stable environment, known strategies may remain reliable. Repetition can improve efficiency, and further exploration may offer diminishing informational returns.

When outcomes change, the agent must infer whether they reflect:

  • ordinary variation;
  • inaccurate previous estimates;
  • or a genuine change in the environment.

That distinction is itself uncertain, and uncertainty of this kind changes how situations are perceived and acted upon.

If conditions have changed, old knowledge may lose predictive value. An option that previously worked may become less effective, or an alternative that once performed poorly may become more valuable. Renewed exploration can then help test whether the current understanding still fits.

But updating, switching and exploring are not identical.

A person can:

  • revise the expected value of the same option;
  • switch to another option they already understand;
  • test an uncertain alternative;
  • increase the range of actions sampled.

Only some of these responses allocate action towards generating new information.

Highly volatile environments create an additional problem. Rapid change can make continued learning necessary, but it can also make new information obsolete before it becomes useful. Neither extensive exploration nor rigid exploitation provides a simple solution.

Environmental change can reduce the reliability of previous learning. It does not make constant switching adaptive.

Exploration and Exploitation Can Be Nested

Exploration and exploitation do not always occur as separate global modes.

A person can exploit one layer of a strategy while exploring another.

A writer may retain a dependable work routine while testing a new argument.

A person may follow an established training programme while experimenting with exercise order.

A company may continue delivering a reliable service while testing whether it fits a new customer segment.

Someone navigating an unfamiliar city may explore possible destinations while using familiar navigation skills.

The same sequence can therefore contain both functions:

  • established knowledge supports dependable action;
  • uncertainty is introduced selectively where learning may matter.

Classification depends on the level of analysis and timescale. A project may be exploratory overall while relying heavily on established skills. A repeated process may be exploitative overall while producing information about small variations. Because goals compete for the same limited action, the allocation is rarely settled once.

The Available Balance Is Constrained

Exploration requires more than curiosity or willingness.

It requires some combination of:

  • accessible alternatives;
  • time;
  • energy;
  • safety;
  • resources;
  • permission;
  • capacity to absorb failure;
  • opportunity to use what is learned.

A person with little financial margin may be unable to test an uncertain option that someone with greater resources can try safely. A worker may understand that another process could be better but lack the authority to change it. A person under coercion may have no meaningful alternative to explore.

Constraints can also prevent exploitation. Access to a known option may disappear, stable repetition may become impossible or external disruption may force a change before previous learning has produced its full return.

The consequences of failure are distributed unevenly. Reliance on a known option can be understandable or adaptive when exploratory failure threatens essential resources.

The reverse can occur too. When known options provide no stable return, unpredictable conditions may produce broad sampling and frequent switching. What is reinforcing under one set of conditions may not be under another.

Research does not support one universal claim that scarcity or threat always creates more exploration or more exploitation.

The balance can also change with development, knowledge and cognitive capacity. But it does not follow one simple movement from exploratory children to exploitative adults.

Real decisions contain more goals, unknown options and changing constraints than formal tasks, and the way goals organise behaviour shapes what an allocation is even for. The Model identifies a recurring trade-off; it does not reproduce every process governing a particular choice.

No Permanent Ratio

Exploration and exploitation solve different parts of acting under uncertainty.

Exploitation uses current knowledge to obtain expected value. Without it, learning never becomes dependable performance or sustained results.

Exploration tests whether current knowledge is sufficient. Without it, action can remain confined to incomplete or outdated options.

Neither is inherently superior.

The relative attraction of exploration can increase when:

  • relevant uncertainty is high;
  • alternatives may differ substantially;
  • information can improve many later decisions;
  • existing options perform poorly;
  • conditions appear to have changed;
  • testing costs are manageable.

The relative attraction of exploitation can increase when:

  • a dependable option is available;
  • relevant uncertainty is low;
  • immediate consequences matter;
  • resources are limited;
  • failure is costly;
  • stable repetition supports skill, coordination or delivery.

These are contextual tendencies, not universal rules.

Behaviour does not reveal the underlying allocation uniquely. The same action can serve different functions depending on what is known, expected and available.

There is no permanently optimal ratio. The balance changes with uncertainty, expected return, prospective information value, cost, horizon, environmental stability and access to alternatives.

The question is not whether exploration or exploitation is better. Each solves a different part of acting under uncertainty: one uses what is known, while the other tests whether what is known is enough.

Behind this page

The claims this model makes, the evidence behind them, and the limits it accepts.

Evidence status

High confidence

Strongly supported, though resting on synthesis or principle rather than a single decisive body of evidence.

Claims

  1. Action under uncertainty involves a recurring tension between using current knowledge and testing whether something better is available

    Established

    What this does not assert: The canonical thesis of the Model.

  2. Trying something new is not automatically exploration

    Canonical inference

    What this does not assert: Novelty separated from exploratory function.

  3. Repeating an action is not automatically exploitation

    Canonical inference

    What this does not assert: Repetition separated from exploitative function.

  4. Learning that happens incidentally during exploitation does not automatically make the action exploratory

    Canonical inference

    What this does not assert: Incidental learning excluded from the functional definition.

  5. Exploratory allocation need not result from an explicit intention to explore, and can emerge through learned tendencies and situational pressures

    High confidence

    What this does not assert: Conscious calculation excluded as a requirement.

  6. Function cannot be read directly from novelty, repetition or one isolated choice

    Canonical inference

    What this does not assert: Inference constraint on classifying single actions.

  7. Function depends on what is known, what is expected, what uncertainty remains, what the action could reveal and how its consequences might affect later decisions

    Canonical inference

    What this does not assert: The five conditions that determine allocation function.

  8. Exploration and exploitation are not fixed kinds of activity or permanent kinds of person

    High confidence

    What this does not assert: Fixed personality types explicitly excluded.

  9. Learning has little practical value if it is never used

    Canonical inference

    What this does not assert: The reason exploitation is necessary rather than merely conservative.

  10. Exploiting an identified option can provide reliable outcomes, efficiency, accumulated skill, reduced decision effort, coordination and sustained progress

    High confidence

    What this does not assert: The six returns of exploitation.

  11. A dependable routine does not need to be redesigned continually, and a team cannot coordinate if every member keeps changing the process

    Canonical inference

    What this does not assert: Worked examples of exploitation as function rather than rigidity.

  12. A familiar option offers a reasonably dependable result; an untested option can provide evidence that improves later decisions

    Canonical inference

    What this does not assert: The opening case, stated as a structural contrast rather than a recommendation.

  13. Following a training programme long enough for adaptation to occur is exploitation performing its function

    Established

    What this does not assert: Time-to-effect example.

  14. Repetition is not in itself evidence of rigidity

    Canonical inference

    What this does not assert: Exploitation separated from rigidity.

  15. The option being exploited is one currently estimated to be sufficiently valuable relative to identifiable and accessible alternatives, not necessarily the best that exists

    High confidence

    What this does not assert: Local rather than global optimality.

  16. Continued exploitation is not mistaken merely because unknown better possibilities might exist, since searching also consumes resources

    Canonical inference

    What this does not assert: Search cost.

  17. Exploitation becomes rigid when persistence stops responding adequately to relevant evidence and changed conditions

    Canonical inference

    What this does not assert: The specific condition under which persistence becomes a failure.

  18. Observed persistence does not reveal its cause with certainty, because switching may be expensive, unsafe, unavailable or outside the person's authority

    High confidence

    What this does not assert: Structural explanations of persistence preserved.

  19. Exploration exposes action to uncertainty in ways that may generate relevant evidence

    Established

    What this does not assert: Functional statement of what exploration does.

  20. Exploration can involve testing unfamiliar options, comparing variations, checking expectations, searching for alternatives, trying strategies under different conditions or learning from others

    High confidence

    What this does not assert: The forms exploration can take.

  21. Directed exploration selects an uncertain option partly because learning about it may improve future choices

    Established

    What this does not assert: Directed exploration as described in the decision literature.

  22. Exploration can also occur through increased behavioural variability, which research sometimes describes as random exploration

    Established

    What this does not assert: Random exploration named as a research description.

  23. Whether evidence is worth acquiring depends on how many later occasions can use it

    High confidence

    What this does not assert: Horizon introduced through the example before it is formalised.

  24. Random exploration is a model-based description of increased choice variability and does not prove deliberate randomisation or usefulness

    Contested

    What this does not assert: The directed/random distinction is qualified, not asserted as mechanism.

  25. Behavioural variation can reflect noise, inconsistent valuation, misunderstanding or error rather than discovery

    High confidence

    What this does not assert: Alternative explanations of variability preserved.

  26. Directed and random exploration are not the only possible forms: systematic hypothesis testing, observation, simulation and single-feature variation also occur

    High confidence

    What this does not assert: The two-form taxonomy is explicitly incomplete.

  27. Exploration should not be confused with indecision

    Canonical inference

    What this does not assert: Indecision excluded from the exploratory function.

  28. Information gathering stops being useful when additional information cannot change the decision or is never integrated into later action

    Canonical inference

    What this does not assert: The boundary at which search ceases to be exploratory in function.

  29. A well-motivated test can remain exploratory even when its result is noisy or uninformative

    Canonical inference

    What this does not assert: Exploration is defined prospectively, not by outcome.

  30. Exploration creates the possibility of learning, not a guarantee of improvement

    Canonical inference

    What this does not assert: Guarantee explicitly withheld.

  31. An action can provide both a direct result and evidence that changes later decisions

    Established

    What this does not assert: The dual return that makes the trade-off non-trivial.

  32. Information has prospective value when it can improve later choices, reveal a better option, reduce repeated error, confirm dependability or identify conditions of success and failure

    High confidence

    What this does not assert: The five conditions of information value.

  33. An experiment that performs poorly today may still contribute more than its immediate cost if it improves many later attempts

    Canonical inference

    What this does not assert: Worked example of prospective value exceeding present return.

  34. A long horizon does not make testing automatically worthwhile, because present consequences and safety still apply

    Canonical inference

    What this does not assert: Blocks the inference that repeated decisions always justify exploration.

  35. Information value is estimated before acting; whether the information improves later decisions is known, if at all, only afterwards

    Canonical inference

    What this does not assert: Expected information value distinguished from realised benefit.

  36. An exploration can be reasonable given what was known beforehand and still produce no useful result

    Canonical inference

    What this does not assert: Blocks outcome-based evaluation of the decision.

  37. Information is not valuable merely because it reduces uncertainty

    High confidence

    What this does not assert: Instrumental value distinguished from uncertainty reduction as such.

  38. Information may have little instrumental value when no future decision can use it, the available actions are unchanged, the person cannot act on it, it becomes obsolete, it cannot discriminate alternatives or it costs more than it can improve

    Canonical inference

    What this does not assert: The six conditions under which information has little value.

  39. People may want information for reasons beyond its effects on later outcomes, including valuing knowing or disliking unresolved uncertainty

    Established

    What this does not assert: Non-instrumental information motives acknowledged.

  40. The exploration–exploitation model does not provide a complete account of curiosity

    Canonical inference

    What this does not assert: Scope exclusion; curiosity is treated in its own place in the Library.

  41. Action value can be represented approximately as expected present return plus expected future value of information minus costs and opportunity costs

    Canonical inference

    What this does not assert: A simplified conceptual representation, explicitly not a calculation of how people necessarily choose.

  42. The terms of that representation interact rather than combine as precisely measurable quantities

    Canonical inference

    What this does not assert: The equation is qualified where it is stated.

  43. Unknown alternatives cannot be valued accurately before they are explored

    High confidence

    What this does not assert: A structural limit on any expected-value account of exploration.

  44. People may not represent any of these factors consciously, and the representation does not guarantee identification of the highest-value action

    Canonical inference

    What this does not assert: Conscious computation and optimality both excluded.

  45. Every allocation of limited action leaves other possibilities partly or temporarily unrealised

    Canonical inference

    What this does not assert: Opportunity cost stated as a structural fact of finite action.

  46. How options are valued in the first place determines what the trade-off is a trade-off between

    Established

    What this does not assert: Valuation; direct-hard prerequisite, published, linked live.

  47. Testing uncertain options can consume time, attention, energy, money, present reward, safety, predictability and practice opportunities

    High confidence

    What this does not assert: The costs of exploration.

  48. Exploration may generate information that is inaccurate, too specific to generalise or obsolete before it can be used

    High confidence

    What this does not assert: Failure modes of acquired information.

  49. The costs of exploitation can be less visible: forgone discovery, uncorrected expectations, unadapted conditions, unexplored possibilities and reduced resilience

    Canonical inference

    What this does not assert: Costs represented bidirectionally.

  50. The two sets of costs need not be equal, and their asymmetry often determines which mode is appropriate

    Canonical inference

    What this does not assert: Asymmetry rather than symmetry of the trade-off.

  51. Exploratory failure may be irreversible, while missed information can sometimes be acquired later

    High confidence

    What this does not assert: Irreversibility as a specific asymmetry.

  52. The mere possibility of a better option does not justify endless search, because exploration can consume more value than the alternative would create

    Canonical inference

    What this does not assert: Blocks the reading that more exploration is always better.

  53. Each mode gives up some of what the other might have provided

    Canonical inference

    What this does not assert: The trade-off stated in its most compact form.

  54. The decision horizon is the remaining time or number of opportunities over which information can be used

    Established

    What this does not assert: Definition of horizon.

  55. A longer horizon often increases the potential value of exploration

    Established

    What this does not assert: Calibrated: often, not always.

  56. Exploitation allocates action primarily towards obtaining an expected result from current knowledge

    Established

    What this does not assert: Canonical definition.

  57. Future opportunities increase information value only to the extent that what is learned transfers to them

    High confidence

    What this does not assert: Transfer condition on horizon effects.

  58. A shorter horizon often increases the relative value of a dependable present result

    Established

    What this does not assert: The converse, equally calibrated.

  59. A long horizon may still favour exploitation when testing is dangerous, skill compounds or stable coordination matters

    Canonical inference

    What this does not assert: Horizon does not settle the allocation on its own.

  60. A short horizon may still justify exploration when current knowledge is unreliable or a small test can prevent an irreversible error

    Canonical inference

    What this does not assert: The symmetric exception.

  61. Time changes the potential usefulness of information but does not create a universal rule for which mode should dominate

    Canonical inference

    What this does not assert: Horizon claims explicitly calibrated.

  62. Exploitation depends on previous learning continuing to predict current outcomes

    Established

    What this does not assert: Operant learning; published, linked live.

  63. In a stable environment, known strategies may remain reliable and further exploration may offer diminishing informational returns

    High confidence

    What this does not assert: Stability reduces the value of testing.

  64. When outcomes change, the agent must infer whether this reflects ordinary variation, inaccurate previous estimates or a genuine change in the environment

    Established

    What this does not assert: The volatility inference problem.

  65. That distinction is itself uncertain

    Established

    What this does not assert: Uncertainty; direct-hard prerequisite, published, linked live.

  66. If conditions have changed, old knowledge may lose predictive value and renewed exploration can test whether the current understanding still fits

    High confidence

    What this does not assert: Change reopens exploration without mandating it.

  67. In this Model, exploitation is a technical term and does not refer to exploiting another person

    Canonical inference

    What this does not assert: Terminological clarification required to prevent moralised reading.

  68. Updating, switching and exploring are not identical responses

    Canonical inference

    What this does not assert: Volatility, updating and exploration explicitly distinguished.

  69. Only some responses to changed outcomes allocate action towards generating new information

    Canonical inference

    What this does not assert: The consequence of the previous distinction.

  70. Rapid change can make continued learning necessary while also making new information obsolete before it becomes useful

    Established

    What this does not assert: The volatility dilemma; no simple resolution offered.

  71. Environmental change can reduce the reliability of previous learning but does not make constant switching adaptive

    Canonical inference

    What this does not assert: Blocks the inference from volatility to perpetual exploration.

  72. Exploration and exploitation do not always occur as separate global modes; a person can exploit one layer of a strategy while exploring another

    High confidence

    What this does not assert: Nested functions represented.

  73. The same sequence can contain both functions: established knowledge supports dependable action while uncertainty is introduced selectively where learning may matter

    Canonical inference

    What this does not assert: The nesting stated structurally.

  74. Classification depends on the level of analysis and the timescale considered

    Canonical inference

    What this does not assert: Level-of-analysis constraint on classification.

  75. Because goals compete for the same limited action, the allocation between the two functions is rarely settled once

    High confidence

    What this does not assert: Goal competition; published, linked live.

  76. Exploration requires accessible alternatives, time, energy, safety, resources, permission, capacity to absorb failure and opportunity to use what is learned

    High confidence

    What this does not assert: Structural constraints on exploration.

  77. A person with little financial margin may be unable to test an option that someone with greater resources can try safely

    Established

    What this does not assert: Material constraint stated concretely.

  78. Exploration allocates action partly towards improving or testing the knowledge that later action can use

    Established

    What this does not assert: Canonical definition.

  79. A worker may understand that another process could be better yet lack the authority to change it

    Canonical inference

    What this does not assert: Institutional constraint.

  80. Constraints can also prevent exploitation, when access disappears, repetition becomes impossible or disruption forces change

    Canonical inference

    What this does not assert: Constraint operates in both directions.

  81. The consequences of failure are distributed unevenly, so reliance on a known option can be adaptive when exploratory failure threatens essential resources

    High confidence

    What this does not assert: Non-moralised account of conservative allocation.

  82. When known options provide no stable return, unpredictable conditions may instead produce broad sampling and frequent switching

    High confidence

    What this does not assert: The reverse pattern under instability.

  83. What functions as reinforcing under one set of conditions may not function so under another

    Established

    What this does not assert: Reinforcement; published, linked live.

  84. Research does not support a universal claim that scarcity or threat always produces more exploration or more exploitation

    Contested

    What this does not assert: Genuinely mixed literature.

  85. The balance can change with development, knowledge and cognitive capacity, but does not follow one simple movement from exploratory children to exploitative adults

    Contested

    What this does not assert: Developmental findings are calibrated, not simplified.

  86. Real decisions contain more goals, unknown options and changing constraints than formal laboratory tasks

    High confidence

    What this does not assert: External validity limit of the task literature.

  87. The way goals organise behaviour shapes what an allocation of action is for

    High confidence

    What this does not assert: Goal organisation; published, linked live.

  88. The Model identifies a recurring trade-off; it does not reproduce every process governing a particular choice

    Canonical inference

    What this does not assert: Explicit statement of the Model's scope.

  89. The difference between the two modes concerns function rather than appearance

    Canonical inference

    What this does not assert: The distinction the whole Model depends on.

  90. Without exploitation, learning never becomes dependable performance or sustained results

    Canonical inference

    What this does not assert: The necessity of the first mode.

  91. Without exploration, action can remain confined to incomplete or outdated options

    Canonical inference

    What this does not assert: The necessity of the second mode.

  92. Neither mode is inherently superior

    Canonical inference

    What this does not assert: No normative preference is asserted.

  93. The relative attraction of exploration can increase with high uncertainty, widely differing alternatives, transferable information, poorly performing options, apparent change and manageable testing costs

    High confidence

    What this does not assert: Contextual tendencies, not rules.

  94. The relative attraction of exploitation can increase with a dependable option, low uncertainty, immediate consequences, limited resources, costly failure and returns to stable repetition

    High confidence

    What this does not assert: The symmetric set of tendencies.

  95. Behaviour does not reveal the underlying allocation uniquely, because the same action can serve different functions

    Canonical inference

    What this does not assert: The inference constraint restated at close.

  96. There is no permanently optimal ratio between exploration and exploitation

    Canonical inference

    What this does not assert: Universal optimal ratio explicitly excluded.

  97. The balance changes with uncertainty, expected return, prospective information value, cost, horizon, environmental stability and access to alternatives

    Canonical inference

    What this does not assert: The seven variables that move the balance.

  98. Each mode solves a different part of acting under uncertainty: one uses what is known, the other tests whether what is known is enough

    Canonical inference

    What this does not assert: Closing statement.

Where to go from here

Next published piece

Curiosity Assigns Value to Missing Information

Someone begins to explain why an approach keeps failing, and then stops. A moment earlier the missing information was one absence among countless others. Now it has a shape, and attention moves towards it.

Continue through the Library →See where this sits in the graph →

Back to the Library →