Engineering Design Document: Cognitive Architecture of
Akien MacIain
2026-02-29 Akien Maciain
1. Purpose
This document formalizes the cognitive architecture
developed through decades of self-optimization and iterative reflection. It is
intended as an engineering design specification of the user’s cognitive
substrate… A structured, modular model that can be analyzed, implemented, and
extended. It describes the “operating system” of Akien MacIain: the
foundational processes that give rise to perception, memory, emotion, habit,
and self-awareness.
2. Core Thesis
Narrative produces emergent self-awareness.
Consciousness, as experienced subjectively, requires
narrative… The sequencing of interpreted experience over time. This sequencing
depends on long-term memory and the ability to recursively re-encode experience
through the lens of evolving internal state.
Narrative + Sequenced Long-Term Memory = Self-Awareness
This architecture does not require metaphysical assumptions.
It treats consciousness as an emergent property of recursive narrative
processing over time.
3. System Components
3.1 Somatosensory System
- Captures
internal and external bodily sensations.
- Feeds
raw data into the interpretive pipeline.
- Observes
the current emotional milieu and contributes to its modulation.
3.2 Emotional Milieu
- A
dynamic, chemical field representing the current affective state.
- Composed
of peptides, hormones, and neurotransmitters.
- Influenced
by:
- Sensory
input
- Recalled
memory
- Internal
narrative
- Properties:
- Mixing:
Multiple emotional signals can coexist and interact.
- Conflict:
Competing signals can create friction or ambiguity.
- Persistence:
Some signals decay slowly, creating emotional inertia.
- The
milieu modulates:
- Memory retrieval
- Search
horizon (See Horizon of Awareness below)
- Habit
activation
- The milieu directly expresses as the Interpretation Bias for new information
3.3 Executive Function
- High-cost,
high-flexibility cognitive control system.
- Functions:
- Goal
selection
- Planning
- Inhibition
of default responses
- Rule
evaluation and override
- Characteristics:
- Compute-expensive: The system prefers to avoid invoking it.
- Delegation to habits: Repeated decisions are compiled into habits to conserve resources.
- Override logic: Rules are stored with their generative why. Overrides are permitted only when the why no longer applies.
3.4 Memory
Encoding
- Perceptual snapshot (from the preceptorium - the internal workspace where raw sensory data is first interpreted into coherent experience)
- Affective valence: move-toward / move-away vector
- Flavor: emotional tone (e.g., joy, disgust)
- Meaning: narrative fragment
- Links: associative connections to other memories
Associative Memory
- Memory
is accessed via other linked memories, not indexed retrieval.
- Search
is biased by:
- Current emotional milieu
- Motivational gradients (overall push/pull of the emotional milieu)
- Habitual interpretation patterns (how we select which memories to retrieve based on the current context). Collectively they are key to the interpretation lens.
Search Duration & Horizon
- The
depth and breadth of memory search is state-dependent.
- High
stress → shorter search horizon → fewer options retrieved.
- Relaxed
state → longer search → more creative or nuanced solutions.
Drift
- Memories
are encoded through the interpretation lens of the current emotional milieu.
- Upon
recall, they are reinterpreted through the new milieu.
- This
re-encoding process causes drift… Memories evolve over time.
- Drift enables:
- Learning
- Healing
- Identity evolution
- Narrative revision
3.5 Habits
- Compiled
executive function: pre-processed decision fragments.
- Activation:
- Triggered
at decision points.
- Preferred
over executive function due to lower cost.
- Reward
structure:
- Small
internal reward (dopamine) for avoiding executive load.
- Types:
- Action habits
- Interpretation
habits
- Modification:
- Requires
rehearsal and repetition.
- Internal
simulation precedes external behavior change.
- Interpretive
habits:
- Shape which memories are retrieved.
- Filter perception through narrative priors.
- Collectively, along with the milieu, set the tone of stored memories.
3.6 Motivational Gradients
- The
system operates on vector fields:
- Move-away:
friction, pain, misalignment
- Move-toward:
joy, clarity, contribution
- Meta-reward:
Bliss… The signal of alignment
- These
gradients bias:
- Attention
- Memory access
- Narrative construction
- Habit reinforcement
3.7 Bliss as Alignment Signal
- Bliss
is not pleasure; it is the felt sense of coherence.
- It
arises from:
- Narrative
alignment
- Contribution
to purpose
- Reduction
of friction
- Creation
of joy
- Functions
as:
- Reinforcement signal
- Compass for long-term optimization
- Integrator of affective and cognitive coherence
3.8 Narrative Engine
- The
brain is constantly asking itself “what does this mean?” And in the
process of the looping and reinterpreting mentioned above, creates the
self.
- It’s
survival job is to create useful meanings that will prompt emotional
gradients to take actions.
- When
we sleep, dreams come as a byproduct of memory consolidation being
observed by the narrative engine.
- It is inherently predictive, and gets rewarded for predicting correctly.
- The
process of producing narrative in the moment does not have to use words, that
shows up in my dogs
- If
the meaning making Narrative Engine faults… Which is to say arrives at a
cusp it has NO habits for (even the default habit of ignore), then it has
faulted, and will try and invoke Executive Function.
- The
Narrative Engine can invoke Executive Function when narrative coherence
breaks down… E.g., when conflicting interpretations arise or Bliss drops
sharply.
- Failure: As Bliss' antithesis, failure is simply a signal that a Further Advance In Learning has occurred. Failure with learning then becomes a signal of more coherence, not less.
- Why: At each cusp, finding a why that explains the habit as well as the habit itself increases coherence further
- Access
to long-range, high-fidelity memory is a prerequisite for the emergence of
narrative-based consciousness.
4. Example Workflow
- Sensory
Input enters the system.
- Initial
Interpretation via habits produces a narrative fragment.
- Memory
Lookup retrieves linked experiences, modulated by current milieu.
- Neurotransmitter/Peptide/Hormone Release updates the emotional milieu.
- Somatosensory
System observes the updated milieu.
- Recursive
Loop:
- Newly updated milieu triggers interpretation on the new state. 'Reinterpretation'.
- Reinterpretation triggers new memory access.
- New memories trigger new affective shifts.
- Loop continues until equilibrium or executive override.
This loop can amplify or dampen
emotional states depending on the alignment of retrieved memories and
interpretations… Producing positive or negative lubricity.
5. Comparative Cognition
5.1 Dogs
- Exhibit
short, two-step narratives (“walkies” → excitement).
- Anticipate
rewards and routines.
- Lack
extended sequencing or self-reflective narrative.
5.2 Crows
- Demonstrate
social memory and grudge transmission.
- Capable
of episodic-like memory and tool use.
- Limited
narrative depth and temporal extension.
5.3 Great Apes
- Show
signs of self-recognition and social learning.
- Working
memory for sequences decays rapidly.
- No
evidence of sustained, branching narrative arcs.
6. Proof Framework
6.1 Necessity Proof
Without narrative sequencing over time, self-awareness
remains fragmented and reactive.
6.2 Sufficiency Proof
Narrative sequencing alone… With memory and emotional
modulation… Generates:
- Continuity
- Agency
- Meaning
- Self-modeling
6.3 Equivalence Proof
- Dreams
= recombined narrative fragments
- Intuition
= compressed narrative inference
- Identity
= long-form narrative arc
- Agency
= narrative branching and counterfactual simulation
6.4 Compression Proof
Narrative explains conscious phenomena with fewer
assumptions than rival models (e.g., global workspace, higher-order thought,
integrated information theory).
7. Conclusion
This document specifies the cognitive architecture of Akien
MacIain as a narrative-driven, affect-modulated, habit-optimized system. It
integrates:
- Sensory
processing
- Emotional
milieu dynamics
- Executive
function and habit compilation
- Memory
encoding and drift
- Motivational
gradients and Bliss as alignment signal
This architecture is both descriptive (modeling
cognition) and prescriptive (providing tools for optimization).
This document’s intention is to provide a basis to be
mirrored in software. Whether it is ever implemented in software is beside the
point. It’s for debugging my own head. A framework for self-reflection and
cognitive engineering.
đ§Š Appendix: Authoring the
Self
While not part of the OS-level architecture, the Player/Character
distinction is a useful metaphor for understanding the system’s application
layer.
- Character:
The narrated self… The identity constructed through memory, habit, and
narrative.
- Player:
The meta-agent… The part of the system that evaluates whys, edits
narratives, and seeks Bliss.
This dual-layer model allows for:
- Meta-cognition
(eg intuition)
- Narrative
revision (Personal growth)
- Ethical
override (Am I acting in alignment with the person I’d most like to
become?)
- Self-authorship
(How can I optimize out something that’s irritating me)
Whether this belongs in this architecture spec depends on
whether you view the Player as part of the OS or as an emergent application
running atop it. I haven’t been able to decide yet. Check in in a week or
twelve, maybe there’ll be news.
This is why we don't need to drag data centers around with us to do what we do. This mode of processing, using habit formation as compiled executive function, is far more efficient than modern AI.
Articles like this are why designing the self is a good skill to have: https://geediting.com/gb-if-you-want-to-stop-being-so-negative-all-the-time-start-by-unlearning-these-8-behaviors/

No comments:
Post a Comment