Apparent Intelligence
By Jessica Lake
Today I started by tidying up a paper titled Brain versus Mind and ended up in a different universe.
Long term, I am exploring how the human nervous system might work. But I approach it from a logical layer a few degrees removed from neuroscience.
My aim is not to muck with neurons. It is to identify a logical architecture that produces the neurobehavioral phenomena we observe. That allows me to apply a lifetime of electrical engineering and computer science to the understanding.
Moreover, my neurodivergence—ASD, SDAM, and Aphantasia—places me in a peculiar position. I experience life once and recall knowledge.
To quote what is oft said by me:
Experiences are the examples.
Knowledge is the rule you learn.
In my lifetime that distinction keeps appearing everywhere.
One thread I revisited today was spoken language. I love language. On and off throughout my life I have treated language as an organic telecommunications protocol. Not a grammar.
Let us look at a stateful exchange. The speaker initiates a handshake with the listener to establish the basis of transfer. Afterward, they synchronize shared state to effect an efficient exchange. Lastly come the highly compressed messages.
This led me back to Strategic Vision. That shared map of reality that expedites F2F (Face-to-Face) exchanges.
For months I have been badgering myself with the question:
Why would the brain integrate a panoramic model of reality only to rip it apart again?
In other words, if navigation is satisfied by Tactical Vision, why bother to integrate everything into a Strategic Vision merely to rip it asunder?
The traditional answer is the ascendancy of consciousness. This bugs me since it leads toward the Homunculus loop.
Instead, I entertained an alternate possibility:
Strategic Vision exists to compress communication through synchronized worlds.
There I have said it. Not language, but synchronization between involved parties.
Consider the coordination of a herd of wildebeest in the absence of spoken language.
One animal senses a predator.
One animal freezes.
The others freeze.
One animal runs.
The others run.
Here the alarm state propagates.
The synchronization of the herd state predates language.
Later, language is a mighty refinement plastered atop synchronization.
This led me to a proposed hierarchy:
Strategic Vision
→ Shared State
→ Synchronization
→ Communication
→ Language
→ Culture
At this juncture I said what sounded trivial:
The easier the state synchronizes between speaker and listeners, the more efficient the overall communication becomes.
That observation immediately connected language, culture, religion, customs, law, friendship, and civilization.
All of the above can be considered means that foster synchronization between individuals.
Then I stumbled upon a marvel. And I coyly confessed:
“I have a secret. Promise not to tell.”
“No. The secret is not about language!”
The secret is:
Evolution never asks questions.
I had been meditating on whether language en vase—language under glass, protected from the real world—was “weak.”
Dhua!
That is an engineer’s question.
Evolution questions nothing.
And that correction reorganized everything.
X and Y no longer needed to exhibit a goal-directed cause and effect. Culture did not evolve to support language. Nor did language evolve to support culture.
Instead, anything that improves the synchronization of speaker and listener improves corporal fitness. And natural selection favors fitness.
Intelligent Designer
Bing.
The anthropomorphism of “Intelligent Designer” popped into my head.
That led me directly into a reorganization of my understanding of Darwin’s theory of evolution.
Admittedly, evolution is stochastic.
Only in hindsight does it appear intelligent.
That said, it sure looks like it is intelligent.
1. Vision improves.
2. Skull shape changes.
3. Posture changes.
4. Head musculature changes.
Everything observed seems indisputably to be goal-directed.
What could give this apparent intelligence?
Bottlenecks
To begin with, selection acts holistically on the specimen.
In systems there is often a sticking point that slows everything else down. This is a bottleneck. Once you relieve it, performance skyrockets.
Until you find the next bottleneck.
Take dinosaurs and feathers. Suppose the ability to glide away from a predator confers a substantial selective advantage. This overwhelming benefit suddenly alters the worth of alleles that aid flight.
● lighter bones
● larger wings
● stronger flight muscles
● better balance
When the selective benefit is big, things happen fast. Evolution exhibits apparent intelligence.
Allele Patterns
A healthy population of a species possesses a sizable reservoir of alleles. Let us define affinity as a statistical index that rates the potential for natural selection. You can gauge each allele as beneficial, neutral, or detrimental to natural selection.
What I believed above is wrong.
What I discovered is that natural selection operates on the pattern of alleles in an organism. Taken individually, a pair of alleles might be detrimental. But together they may be judged a benefit. A novel allele might recast the affinity of any number of extant alleles.
A better understanding of evolution became one of my major insights of the day:
Alleles are declarative, not procedural.
Their significance does not depend upon the order in which they are presented. What matters is the overall effect of their combination.
Let us suppose a novel allele appears. Within any generation an individual possesses some combination of alleles. You can consider the benefit of the novelty only in combination with an individual’s pattern of alleles.
I had been imagining a sequence. You get this, and then you get that. But an allele can show up and remain in the population for generations. It can be dominant or recessive. Then an individual is born with a combination that makes a difference. The last piece might be a new allele. It might simply be that the pieces have come together in one individual.
Think back to the flight example:
● lighter bones
● larger wings
● stronger flight muscles
● better balance
If the combination confers a substantial advantage, selection can change the population quickly. Much of the variation was already there. From the outside, it looks as though evolution suddenly figured something out.
This is what interested me about the word declarative. The alleles can enter the population at different times and persist alongside one another. When they come together in an individual, what matters is what they do together.
I had already run into the same advantage in Constraint Convergence. Each constraint declares a condition. I don’t have to finish evaluating one before I can begin evaluating another. I can check a candidate against several constraints in parallel. What matters to the resolver is how the candidate fares against the constraints taken together.
I kept finding that same arrangement in other systems I worked on. The pieces can be developed or evaluated independently. Their combined effect is what matters.
Subsystem Arbitration
There is a flaw in my localized approach to the nervous system. Subsystems are decoupled. They predict their best response. What happens when they disagree? You cannot fight and flee.
How can local predictors resolve disagreements without intelligence?
A common solution in systems when instruments disagree is a “voting” mechanism.
Each subsystem contributes a ballot with properties of this nature:
● Action proposed
● Confidence in proposal
● Peril without resolution
● Urgency of resolution
This method of arbitration is heuristic. Resolution emerges from an arbitration algorithm.
And best of all, no homunculi were involved in the filming of this movie.
Coherence Emerges
All this fun got me thinking. The same pattern kept reappearing:
Evolution:
Variations → Selection → Organism
Reasoning:
Candidates → Constraints → Conclusion
Action:
Predictors → Arbitration → Behavior
Communication:
Individuals → Synchronization → Shared State
Again and again, coherence emerged from distributed processes.
Shotgun
That brought me back to the challenge of consciousness.
For years I have been whittling away responsibilities from consciousness:
● free will
● decision making
● action selection
● prediction
● memory reconstruction
● reasoning
Each time the system continued functioning. To my mind, the conscious mind increasingly resembles a passenger.
No offense.
Apparent Intelligence
Then I lighted on a fascinating possibility.
It is a common fallacy that evolution requires an intelligent designer. Some mouse got lucky seems a poor way to build a better mouse. Yet Darwin showed that evolution itself is not intelligent.
What if human cognition works the same way?
Not because cognition is evolution. But because evolution is mechanistic yet is judged by many as intelligent.
That became the key connection. Not that evolution is the answer. Just that it proves a mechanism can appear intelligent.
At this point I stopped thinking about ideas and started thinking about models. It is the validity of their predictions that is the goal.
How do we remove deprecated models?
One answer is TTL, or time-to-live. A model dies when its TTL expires. But each time you use a model, you restart its TTL. This I call Model Decay.
What if our models imitated evolution? We seemed to have all the pieces.
1. Knowledge is prediction.
2. Models compete.
3. Reality serves as the proving ground.
4. Accurate models prosper.
5. Inaccurate predictors generate anomalies.
6. Models decay.
This suggested a new architecture:
Proving Grounds
↔ Prototypes
↔ Mechanism
1. The mechanism builds prototypes.
2. The proving grounds evaluate them.
3. And the mechanism adapts.
Bottleneck Redux
A bottleneck in systems is a pain point. It is the worst offender in distributed architectures.
In semantic knowledge, a bottleneck would likely be a model that is over- or under-constrained.
When we break a bottleneck:
● Latent possibilities become favorable.
● Competing models are reevaluated.
● Previously neutral models become favorable.
Keynote: No directed acyclic graphs, ergo nothing propagates.
As a thought experiment I imagined accelerating evolution by, say, a millionfold.
To the observer, the process would look exactly like intelligence. Yet underneath there remains only variation, selection, retention, and adaptation.
This led to the fascinating possibility:
What if cognition operates through a similar mechanism?
Not because cognition resembles evolution. But because evolution demonstrates that intelligence-like behavior can emerge from mechanisms.
I am not trying to build intelligence.
My ambition is to identify a mechanistic architecture that, to the observer, appears intelligent. That I coin as Apparent Intelligence.
The exact underlying implementation is secondary. Neurons are one implementation. Silicon could be another. Quantum another.
It is the logical architecture that matters.
This leaves me with a set of provisions:
● No magic.
● No homunculus.
● Favor local operation.
● Environmental evaluation.
● Adaptive reorganization.
● Prediction as fitness.
● Retention of alternatives.
● Bottleneck breaking.
● Declarative models.
● Stability and plasticity.
The strongest constraint may be this:
Apparent Intelligence must emerge from the pure mechanism.
Intelligence cannot be used as an explanatory primitive.
That is where I ended the day.
Not with an answer.
But with a possibility.
If evolution can generate the appearance of design through mechanistic adaptation, perhaps cognition can generate the appearance of intelligence through mechanistic adaptation as well.
The frightening part is not that it might be true.
The frightening part is how natural it suddenly seems.