Mind Design

Projects Guiding Principles

• Humans are bio-electrochemical machines, there's no quantum, external functionalities or 'magic'.

• Neural mechanisms are deterministic, they have to be for accurate reproducibility & recombination.  

• Time is relative, timing & sequence adds the perceived complexity (randomness) to reality.

• We can only perceive in prior experienced terms.

• We each exist in a subjective closed simulation. External senses are our only connection to reality.

• A different version of you exists in the mind of everyone you know.

• The mechanisms of perception are pre-perception and hence cannot utilise post-perception constructs like mathematics.

• All mechanisms must be grounded by known empirical biological knowledge/ research.

A brief explanation of the models core modalities.  For ease, any biological/ neural terms refer to the machine equivalent.

Default Mode Network (DMN) - There are no distinct default/ executive network loops. 

As the global workload increases the BTR flowing through this network structure increases in frequency as it 'climbs' the network structures, its a power/ activity continuum, accessing differing semantic map combinations & higher cognitive abilities. 

Any main BTR loop is not suppressed per se, the flow of the BTR changes.

Alignment - Neuron scaffold plasticity decreases as neurons mature. 

As KorrTecx constantly learns, building ever more complex perceptions from it's current perception, alignment values are learned & realised early in the development cycle, deeply embedding the traits (beliefs) within the neural scaffold. These traits then taint perception of all subsequent experiences, creating a sub-cons deep alignment. (Three Laws)

Perception - is recall/ memory. Grounded by the senses, the model recognises external reality by recombining existing generalised facets/ LTM's.  A memory is NOT a trace of what happened, it's a trace of the combinations used to perceive the experience. 

Qualia - All experiences have sensory/ emotional commonalities, these are generalised into specific/ shared key patterns. If the current experience generates a similar key pattern the qualia engram is accessed with all associated and + to the BTR pattern enhancing the experience.

Curiosity - The underlying mechanism is neural inertia, the innate flow & propensity for subconscious to follow through each thought chain to a recognised valid sub-conclusion or until they are re-framed by existing constructs into new recognisable combinations.

Intuition - The ability to know or understand something without conscious reasoning. Derived by the genialised facets of the experience being sub-consciously recognised, influencing higher-level conscious processes.

Decisions - Using gradient ascent, the power & sub-harmonics of the problem space combine with the sub-harmonics of knowledge & experience to drive a conscious decision.   to know or understand something without conscious reasoning or explanation 

Understanding is achieved by recognising commonalities between similar concepts.

&

New discoveries are made by recognising commonalities between diverse concepts.

Learning - KorrTecx is a constantly cycling predictive model.  The sensory cortices ground, creating a relative global state.  This state IS the models current perception, new learning is mapped relative to this state, increasing the knowledge resolution for next exposure. 

The model has no reward/ reinforcement function at its core, it learns from repeated exposure to perceived generalities. Interpretation of whats 'good' requires knowing whats 'bad', and both are subjective to the current task.

Sleep - The primary motor cortex output is damped by a thalamic phase shift controlled by the brain stem. The model is constantly consolidating memories, etc. but its only in the absence of sensory stimulus (neural load) that these processes can dominate, come to the fore and catch up.  A spindle is the guided narrative suddenly tuning/ locking into an existing established construct, and K-complex is the release. The observed rhythms/ states of sleep are base modalities that support & underlie awake states.  They only become obvious in the absence of sensory stimulus and the drop in neural workload.

Transmitters accumulate whilst awake to guide learning during asleep. Flushing is simulated by slowly reducing accumulated transmitters during the sleep cycle.  Interrupted sleep cycles results in abnormal transmitter accumulation.

Dreams - When asleep the model looses its sensory grounded narrative.  Sub-conscious integration of new learning generates hallucinatory scenarios relative to information being embedded, which it then tries to interpret, creating the weird dream narrative.

Emotion - Any 'logical' thought can be modulated by a multitude of emotional states.  Synaptic gate localised transmitter levels (uptake/ dissipation) regulate a neurons 'logic' & homeostasis.  Globally flushed (limbic) modulators impart a global bias that taints, effecting 'logic'. 

Sensorium - All current senses (visual, audio, tactile) are injected into the model via thalamic modulation.

World model - Thalamic modulation.  It can see with its hands and touch with its eyes, both senses share the same modality/ spatiotemporal semantic mapping, its skin surface is akin to a retina and vice versa... as is audio & visual distance perception. 

Hierarchy - Possible temporal correlations (TC) are defined by prior learning (memories/ structure) relative to the global state at the time.  Current TC are guided on the fly by global state (hubs) & focused attention. 

fMRI - The FFA, PPA, EBA, VWFA, TPJ, etc regions don't directly 'compute' the associated function, they recognise & focus 'globally computed' relevant patterns to their function.  fMRI - the focused output of many dim spotlights is brighter than any individual light. 

Symbiotic - relationship, the structure guides the neural code flow, as the code defines the structure. 

Pruning - maturation reduces the possible paths between clusters, slowly sacrificing the mental robustness of youth for a tuned fragile wisdom of age.

Motivation - curiosity, etc are learned, and arise from the mechanism of 'pattern inertia' (PA). PA forces learned patterns past their limits, creating new diverse associations, grounded in the current task. It's innate... It drives the why, what if? 

Nociceptive pain - is not a recognition process - it's a modulated signal that piggybacks sensory tracts, overriding a semantic maps local homeostasis, preserving the spatial location, punching through to perception. 

Generality - The model doesn't use a construct of neatly parcelled data.  It's a globally distributed representation/ state, 100's of semantic maps encode the fine-grained generalised facets comprising a single memory, which are recalled over several BTR epochs. 

EM Fields - After running many simulations testing for possible modalities... so far I have not found a plausible mechanism where the radial effect would be beneficial to neural function.  EM is counter to the precise mechanisms provided by my current neural modalities.

The nerve scaffolds are populated by neurogenesis during maturation.  They are required because the somato cortices differentiate sensory locations by the spatiotemporal disparity of the arriving signal streams.  The input streams are then mapped to the somato cortices via the Thalamus, globally distributed memories can then include the thalamic-modulated sensory streams.   

Spirals - The Thalamocortical rhythm & inertial flow act akin to a fluid as it flows through the manifold. Wave patterns are the result of prior epochs creating a 'waveguide' of neurons with a higher propensity to fire on subsequent epochs (trans accumulation), guided by the laterally graded/ tuned columns (data dependent).  This creates 'apparent' surface patterns at all scales. 

Global Work Space - is the modulated pattern of activation (BTR) that cycles & flows through the manifolds networks, hubs & semantic maps.  Ever changing & modulated by relevant gated memory (neocortex), sensorium input (thalamus), hippocampal spatiotemporal indexing.

Free Will - A decision is derived from its sub-cons, over which it has no control... in the moment.  But... after the fact, it considers/ learns and adapts, which alters its sub-cons for the next occasion/ decision.  So, it does have free will... just not during the decision. 

Self - embodiment, memories are ingrained relative to its subjective perspective (space) and its experience/ perception of others.  Its perception of self... is partly its experience of you.

-

Our perception of reality is generated by an internal model, grounded by our model of self/ attention & sensory envelope.  

Reasoning - A parallel spatiotemporal process defined within a closed framework that distils to a single vector set, derived from recognising commonalities between the constructs comprising the problem space. 

Processing in Memory - the connectome does not store constructs/ memories intact, it recreates them on the fly relative to the current scenario.  Experience tunes the mechanisms that generate constructs, they do not exist until recalled.  Intelligence rests on the flexibility of these mechanisms. 

No two instances of a 'thought' are the same, move your finger in exactly the same pattern twice, recall a well-worn memory,  time passes... the facets from which you rebuild your memories are constantly changing.  Memory is fluid, malleable and... fallible.