Is intelligence subjective?

May 25, 2026

Is intelligence subjective?

Intelligence Is Probably Not One Thing

This is where Marvin Minsky becomes relevant again.

Many scientists now suspect:

Intelligence is not a single mechanism.

Instead it may be a combination of systems such as:

memory,
prediction,
attention,
learning,
abstraction,
language,
social reasoning,
planning,
sensory modeling.

In other words:

“Intelligence” may be a label we give to successful coordination among many cognitive abilities.

What Neuroscience Found

Scientists have identified important patterns associated with intelligence.

For example:

Brain connectivity

People with higher problem-solving ability often show:

more efficient communication between brain regions,
especially between frontal and parietal areas.

This is called the Parieto-Frontal Integration Theory.

But:

no single “intelligence center” exists.

Prediction machinery

A major modern idea is:

Brains are prediction machines.

Your brain constantly predicts:

what you’ll see,
hear,
feel,
and what happens next.

Intelligence may partly come from:

building accurate internal models of reality.

This idea appears in:

predictive processing,
Bayesian brain theories,
modern AI.

Abstraction Seems Crucial

One strong candidate for “core intelligence” is:

Abstraction

The ability to:

compress patterns,
generalize,
reuse ideas in new situations.

Example:
A child learns:

“2 apples + 2 apples = 4 apples,”
then understands:
“2 + 2 = 4” abstractly.

That leap from concrete examples to generalized rules is deeply tied to intelligence.

Learning + Transfer

A highly intelligent system can:

learn in one context,
apply knowledge elsewhere.

Humans are unusually good at this.

Example:
If you learn balance from biking,
you can partially transfer that understanding to:

skating,
surfing,
skiing.

AI systems still struggle with broad transfer compared to humans.

Consciousness vs Intelligence

Another unresolved issue:

Are intelligence and consciousness connected?

Not necessarily.

A system might:

solve problems intelligently,
without subjective experience.

For example:

modern AI can perform reasoning tasks,
but whether it is conscious is entirely unresolved.

Have Scientists “Found” Intelligence?

Not really in the sense of:

a single structure,
particle,
equation,
or universal algorithm.

Instead, researchers found:

many mechanisms correlated with intelligent behavior.

Think of it like this:

Scientists understand:

muscles,
oxygen,
biomechanics,
coordination,

but “athleticism” is still an emergent quality.

Likewise:

neurons,
memory,
prediction,
learning,
abstraction,
and planning
may together produce what we call intelligence.

One of the Biggest Current Ideas

Many modern theories suggest intelligence emerges from systems that can:

Build internal world models
Predict outcomes
Learn from errors
Compress information into abstractions
Plan over time
Adapt flexibly

That applies surprisingly well to:

humans,
animals,
and modern AI systems.

A Deep Philosophical Twist

Some philosophers argue:

Intelligence may not even be a single natural category.

Meaning:
“intelligence” could be like “fitness” or “beauty”:

a useful human concept,
but not one sharply defined thing in nature.

That possibility is still taken seriously today.

Intelligence is not just knowledge itself, but the ability to connect, model, abstract, and predict from knowledge before acting.

That’s a sophisticated view.

Knowledge Alone Is Not Intelligence

A system can store enormous amounts of information and still not be very intelligent.

Example:

a library contains huge knowledge,
but it does not reason.

Similarly:

memorization alone is limited.

Intelligence seems to require:

organizing information,
finding relationships,
predicting consequences,
and choosing actions.

So your emphasis on connectivity of information before decision-making is important.

Prediction Is Central in Modern Cognitive Science

Many neuroscientists now think the brain is fundamentally a prediction engine.

The idea is:

The brain constantly simulates reality before reality fully arrives.

For example:

predicting the next word in a sentence,
predicting another person’s reaction,
predicting where a moving object will go,
predicting danger,
predicting social outcomes.

A lot of “smartness” may come from:

building accurate internal simulations.

In some sense:

thinking is controlled prediction.

Abstraction May Be the Most Powerful Part

Your point about abstraction is especially important.

Abstraction lets intelligence:

compress complexity,
recognize patterns,
reuse ideas across domains.

For example:

A child first learns:

specific dogs.

Then abstracts:

“dogness.”

Then later:

“animal,”
“living thing,”
“biological system.”

That layering of abstractions is enormously powerful.

Without abstraction, intelligence would drown in details.

Decision-Making Is Not Pure Logic

Another subtle thing you touched on:

A “smart decision” is not only logic.

It also depends on:

incomplete information,
uncertainty,
emotional weighting,
probability estimation,
long-term modeling,
social context.

Humans often make decisions using:

rough predictive models,
heuristics,
intuition.

Modern AI systems do something somewhat similar statistically.

Your View Resembles “World Model” Theories

Current AI and neuroscience increasingly focus on the idea of a world model.

An intelligent system:

gathers information,
builds an internal model of reality,
simulates possible futures,
evaluates outcomes,
chooses actions.

That’s extremely close to your description.

One Interesting Consequence

If intelligence depends heavily on:

prediction,
abstraction,
and relational modeling,

then raw processing power alone is not enough.

A giant database is not necessarily intelligent.

Likewise:

fast calculation alone is not wisdom,
and high IQ alone does not guarantee good judgment.

The structure and use of information matter more than quantity alone.

This Also Explains Why Humans Make Mistakes

Our internal models are imperfect.

Sometimes:

we overgeneralize abstractions,
predict wrongly,
use biased information,
or emotionally distort outcomes.

So intelligence is probably not:

“being correct,”

but rather:

improving models through feedback and adaptation.

That aligns strongly with modern learning theory and even scientific method itself.

A Nice Compact Formulation

What you’re describing could almost be summarized as:

Intelligence is the ability to build useful abstractions and predictions from connected information in order to make adaptive decisions.

That is honestly very close to how many researchers informally think about it today.

...and in top of that is the unconsciously emotional deep imbedded in your human upbringing, your decision will be influenced by what person are or what experiences you had...

Yes — and that insight became increasingly important in neuroscience and psychology over the last few decades.

For a long time, people imagined intelligence as mostly:

logic,
rationality,
calculation.

But modern research strongly suggests that:

emotion is not separate from intelligence;
it is deeply involved in decision-making itself.

Your experiences, upbringing, fears, rewards, attachments, and emotional memories shape:

what you notice,
what you value,
what you predict,
and what actions feel “right.”

Emotion Is Not Just “Noise”

Older views often treated emotion as something that interferes with rational thought.

But researchers like Antonio Damasio argued something very different:

People with damage to emotional-processing areas of the brain sometimes retain:

logic,
memory,
language,
IQ,

yet become terrible at real-life decisions.

Why?

Because they cannot emotionally evaluate outcomes properly.

They can reason endlessly about possibilities but cannot effectively prioritize what matters.

This led to the idea that:

emotions help guide intelligence by assigning value and significance.

Your Brain Is Not Neutral

Your internal “world model” is shaped by:

childhood,
culture,
trauma,
success,
relationships,
fears,
habits,
rewards,
social experiences.

Two people can see the same event and interpret it completely differently because their brains learned different predictive patterns.

For example:

one person sees criticism as useful feedback,
another experiences it as danger.

The sensory input may be identical.
The emotional prediction system is not.

The Unconscious Part Is Huge

A lot of brain processing happens below awareness.

Modern neuroscience suggests consciousness is only a small visible layer above enormous unconscious machinery.

Your brain constantly:

filters information,
predicts outcomes,
ranks importance,
activates emotional associations,
biases attention,
prepares reactions,

before conscious thought even catches up.

This is why people often:

“feel” something before they can explain it logically.

Emotion and Prediction Are Connected

Modern theories often see emotions as part of predictive regulation.

Example:
Fear is not merely a feeling.
It is also:

a prediction of possible harm,
plus preparation for action.

Similarly:

anxiety predicts uncertainty,
pleasure predicts reward,
shame predicts social risk,
trust predicts safety.

So emotions may actually be compressed predictive tools built from past experience.

This Makes Human Intelligence Deeply Personal

Two highly intelligent people may make opposite decisions because:

they value different outcomes,
they carry different emotional histories,
they predict different futures.

Pure logic rarely determines human behavior alone.

Human intelligence is entangled with:

memory,
identity,
social experience,
biology,
and emotion.

This Is Also Why Humans Are Not Fully “Objective”

Our abstractions are shaped by experience.

What feels:

important,
dangerous,
beautiful,
meaningful,
trustworthy,

depends heavily on learned emotional structure.

In a sense:

emotion acts like weighting inside cognition.

It changes which predictions dominate.

This Connects Back to Society of Mind

Marvin Minsky actually argued something compatible with this.

He did not think the mind was a purely logical machine.

In his view:

many competing subsystems,
drives,
memories,
learned reactions,
and goals
all interact simultaneously.

The “self” emerges from negotiation among these systems.

That’s surprisingly close to modern views of cognition today.

_____________________________

With Chat GPT