# Deep Theory: Trident G

## Status

This document is a public theory note, not a training protocol and not an evidence claim by itself.

It summarises the deeper computational-neuroscience theory behind the Trident G name and explains how that theory motivates IQMindware's readiness, training, transfer, and validation protocols.

The source papers are preprints by Mark Ashton Smith:

- Trident G: A Computational-Level Theory of Intelligent Life: https://www.researchgate.net/publication/398500214_Trident_G_A_Computational-Level_Theory_of_Intelligent_Life
- Life on the Cusp: The Trident G Theory of General Intelligence: https://www.researchgate.net/publication/398153172_Life_on_the_Cusp_The_Trident_G_Theory_of_General_Intelligence

## Core idea

Trident G frames intelligence as regulation of a viable near-critical operating band.

The central claim is not that intelligent systems should minimise load, uncertainty, or free energy toward zero. A system that minimises too aggressively becomes rigid, over-safe, or poorly prepared for changing conditions.

Instead, Trident G treats intelligent life as the ability to stay within, leave, traverse, and re-enter a useful operating corridor where the system can think, learn, adapt, and recover.

## Main variables

The theory uses a small set of computational variables.

- `F`: realised free-energy / allostatic load.
- `F*`: defended allostatic set-point or operating-band centre.
- `x = F - F*`: current deviation from the defended band.
- `G`: expected or predicted load.
- `xi = G - F*`: expected challenge or engagement stance.
- `delta`: explore-exploit tilt, biasing the system toward exploratory or exploitative modes.

The system is intelligent to the extent that it can navigate a useful volume of this state-control space while remaining viable and able to re-enter its near-critical band after disruption.

## Why it is called Trident G

The "Trident" name comes from the geometry of the model:

- a lower-load autopilot shaft;
- a near-critical operating band;
- high-load fluid-intelligence branches that can tilt toward exploration or exploitation.

The "G" refers to general intelligence, but not as a single fixed trait alone. In this theory, general intelligence is closer to navigable volume: how much of the useful cognitive-control landscape an agent can reliably access and re-enter.

## How this informs IQMindware

IQMindware operationalises the theory in a practical training programme.

### Zone Pulse

Zone Pulse maps onto the state-regulation problem. It asks whether the user is currently in a good operating corridor for high-quality training.

The MFT-m bits/sec result and state classifier are not a full neural criticality measure. They are behavioural readiness signals used to decide whether the day should be a full route, reduced route, support route, or repeat check.

### Capacity Gym

Capacity Gym targets run-time control: attention, working memory updating, interference resistance, and flexible binding under load.

In Trident G terms, this is training the system's ability to hold controlled processing inside the operating band while challenge increases.

### Reasoning Gym

Reasoning Gym adds vertical demand. It asks the user to use capacity for relation matching, constraint solving, must-follow inference, and changed meaning/surface forms.

In Trident G terms, this tests whether the control gained in Capacity Gym can support higher-order reasoning rather than staying as one-game fluency.

### Tracker

Tracker supplies evidence beyond training scores. It separates in-app gains from pre/post reasoning checks, applied cognition trend checks, and later delayed re-checks.

In Trident G terms, this is re-entry and carry-over evidence: does the system show progress beyond one practised surface?

## Relationship to far transfer

The theory motivates the IQMindware transfer posture:

- train at the right state and dose;
- avoid thin automation inside one wrapper;
- use horizontal swaps to test portability;
- use vertical progression to move from capacity to reasoning;
- use delayed and external checks to avoid mistaking game fluency for transfer.

The theory does not prove that any individual will improve. It supplies the design rationale for why the programme uses readiness gating, wrapper swaps, vertical reasoning progression, and evidence checks.

## Relationship to H-AGI

Human-Artificial General Intelligence (H-AGI) extends the same idea to AI-assisted cognition.

The practical question is whether AI extends human cognition while the human keeps:

- problem framing;
- challenge selection;
- reasoning audit;
- stopping rules;
- final judgement;
- re-entry after disruption.

In Trident G terms, AI should increase useful navigable volume without narrowing the human into passive automation, overdependence, or brittle offloading.

## Claims boundary

Allowed:

- Trident G is the deeper theory origin for the IQMindware name and design logic.
- It motivates readiness gating, horizontal transfer, vertical progression, and delayed checks.
- It offers a computational frame for intelligence as viable-band regulation and navigable control volume.
- It provides testable directions for future EEG/HRV, behavioural, mission, and AI-use studies.

Not allowed:

- The theory alone proves the IQMindware programme works.
- The theory is a clinical model or diagnostic framework.
- A Zone Pulse state is a direct measurement of neural criticality.
- A single app score proves navigable volume or general intelligence.
- Preprint status should be treated as finished consensus.

## Source links

- Trident G: A Computational-Level Theory of Intelligent Life: https://www.researchgate.net/publication/398500214_Trident_G_A_Computational-Level_Theory_of_Intelligent_Life
- Life on the Cusp: The Trident G Theory of General Intelligence: https://www.researchgate.net/publication/398153172_Life_on_the_Cusp_The_Trident_G_Theory_of_General_Intelligence