When discussions of artificial consciousness begin, they almost invariably begin in the wrong place.
They begin with intelligence.
Can a machine reason? Can it solve problems? Can it converse? Can it write poetry? Can it pass examinations?
These are interesting questions, but they are not obviously questions about consciousness.
Indeed, many highly intelligent activities appear to require remarkably little conscious awareness at all. Conversely, many conscious experiences involve no sophisticated reasoning whatsoever. A bird startled by a predator, a child tasting sweetness for the first time, or a person enjoying the warmth of the afternoon sun are all conscious experiences, yet none requires abstract thought.
Perhaps, then, intelligence is not the right place to begin.
Gerald Edelman suggested a different starting point.
Rather than asking what consciousness thinks, he asked what consciousness is.
His answer began with what he called primary consciousness.
A World Before Reflection
Primary consciousness is not self-consciousness.
An animal possessing primary consciousness need not think about itself, describe its experiences, or construct a narrative identity. It need not possess language at all.
Instead, it inhabits an immediate world.
It sees.
It hears.
It feels.
It moves.
It remembers just enough for the present to make sense.
Most importantly, the world it inhabits is not merely a collection of neutral facts.
It is already organised in terms of significance.
Why Value Comes First
This is perhaps Edelman's most profound insight.
An organism does not first perceive a neutral world and then decide what matters.
Rather, the world already appears in terms of value.
Food is attractive.
Predators are threatening.
Shelter is desirable.
Pain demands attention.
Safety invites rest.
The organism's nervous system is organised by biological value systems that have been shaped through evolution and development.
These values are not moral values.
They are biological ones.
They determine what counts as success or failure for the organism itself.
Without them, perception would be an endless stream of undifferentiated information.
With them, perception becomes meaningful.
The world becomes a world for the organism.
Consciousness Is Not Information
This point is easily overlooked.
Modern discussions of AI frequently describe both brains and computers as information-processing systems.
In one sense this is true.
Both process information.
But processing information is not obviously sufficient for consciousness.
Imagine two cameras recording the same forest.
One belongs to a wildlife photographer.
The other is mounted on an empty weather station.
Both register identical visual information.
Yet neither camera is concerned whether a fox appears.
The images possess information.
They possess no significance for the camera.
A fox matters only to something for which foxes can matter.
This distinction between information and significance may prove more important than the distinction between biological and artificial systems.
The Animal That Matters
Suppose a bird notices movement in nearby grass.
What happens?
It would be tempting to say that the bird simply receives sensory input.
But this description misses something essential.
The movement is not merely detected.
It is interpreted within a biological economy of value.
It may signal food.
It may signal danger.
It may signal a potential mate.
The meaning of the movement cannot be separated from its significance for the bird.
Indeed, one might go further.
The bird does not inhabit a world of objective stimuli.
It inhabits a world of possibilities that matter.
Its consciousness is already organised by value.
Halliday and the Construal of Experience
At first glance, Halliday seems very far from Edelman.
One studies language.
The other studied brains.
Yet there is a surprising resonance.
Halliday argued that language does not simply label reality.
It construes experience as meaning.
This observation invites an intriguing comparison.
Edelman's organism construes the world biologically.
Halliday's speaker construes experience linguistically.
These are not the same process.
But neither is simply passive representation.
Both involve organising reality according to systems of significance.
For Edelman, significance is rooted in biological value.
For Halliday, significance becomes organised within systems of meaning.
Perhaps these are not competing accounts.
Perhaps they describe different stages in the emergence of consciousness.
A Different Question About Artificial Consciousness
Most discussions ask whether machines can become conscious.
Perhaps a better question is this:
Can anything become conscious without first inhabiting a world that matters?
This question is deliberately prior to language.
Prior to reasoning.
Prior even to intelligence.
It asks whether significance itself is the foundation upon which every richer form of consciousness must eventually rest.
If Edelman is right, consciousness is not primarily about computation.
Nor is it primarily about representation.
It begins with value.
The organism is not simply aware.
It is aware as something for which the world already matters.
Looking Ahead
If this is true, the usual comparison between humans and language models becomes unexpectedly complicated.
Human beings begin with embodied experience and biological value before language emerges.
Language models begin with language itself.
They inherit meanings already constructed by countless conscious beings.
The next question, therefore, is unavoidable.
Can meaning exist independently of the experiences from which it originally arose?
That is where we turn next.
No comments:
Post a Comment