Visual Knowledge Library

From “What is this?” to “How could this help me?”

The High-Density Index tells people and AI what exists. The HDIMM shows how the parts connect. The Visual Knowledge Library turns those relationships into application pathways using registered infographics, audience guidance, starter prompts and reliance notes.

Choose an application pathwayBrowse all visuals
Published: 25 July 2026Updated: 25 July 2026Status: Research and Pilot ReleaseAuthor: Walter Shepherd

The visual application layer

A governed bridge between knowledge and use.

The library connects a visual explanation to a task, audience, method, instrument, next page and reliance boundary.

1

What is this?

Identify the registered concept or method.

→
2

What does it explain?

Read the visual and its underlying page.

→
3

How could it help me?

Choose a task-relevant application pathway.

→
4

What may I rely upon?

Apply the stated limits and human decision point.

Choose by need

Start with the question you are trying to answer.

Each path recommends the minimum visual starting point. The underlying page and professional boundaries should be loaded before substantive use.

ORIENTATION

What is GovAIaaS, and how could it help me think better?

Understand the human–AI partnership and select the minimum relevant specialist Skills.

SELECTION

Which AI capabilities should I use for my task?

Map the task to the minimum Skills and identify human decision points.

DISCOVERY

How could AI help discover and test possible mechanisms?

Generate and challenge candidate mechanisms using an explicit six-stage pathway.

AUDIT

How could I break a complex claim into auditable parts?

Construct a claim chain and identify necessary components, uncertainty and weak links.

AI

How could I compare independent AI audits fairly?

Compare audits using the same template and report convergence, divergence and repeated weak links.

Registered visual catalogue

Five visual pathways from orientation to governed application.

Open each entry to see use cases, related methods, reliance limits and an AI starter prompt.

AI as a Thinking Partner infographic.
orientationD1orientation only

AI as a Thinking Partner

Question answered: How can AI extend human thinking without replacing human judgement?

What it explains

  • common limits of human thinking alone
  • the distinct contribution of human judgement and AI capability
  • the six-stage governed workflow
  • intended outcomes such as broader view, deeper insight and stronger decisions

How it could help

  • orient a new visitor to the GovAIaaS proposition
  • frame an AI-assisted task as a human-and-AI partnership
  • identify where assumptions, blind spots or cognitive overload may affect a decision
  • explain why human control remains essential
Application notes and starter prompt

Use when

  • introducing GovAIaaS to a general audience
  • starting an AI-assisted decision or research task
  • clarifying the roles of the human and the AI

Related registered methods and instruments: METHOD-GOV-WORKFLOW-006, INSTRUMENT-HDI, INSTRUMENT-HDIMM

Reliance note: Use for orientation and task framing. It does not establish that a particular AI output is accurate, governed or safe to rely upon.

AI starter prompt

Using this visual, explain how human judgement and AI capability should divide responsibilities for my task. Label registered content and your own inference separately.
Open full size Open underlying page
GovAIaaS Skills Working Together infographic.
capability selectionD2method guidance

GovAIaaS Skills Working Together

Question answered: Which specialist AI capability should be used, and how can the Skills work together?

What it explains

  • the eight specialist GovAIaaS AI Skills
  • the role of each Skill in a governed workflow
  • how integrated Skills can produce deeper insight, lower risk and stronger solutions
  • illustrative use cases across policy, strategy, research, risk and assurance

How it could help

  • select the minimum Skills needed for a task
  • identify capability gaps in an analysis or project
  • sequence analysis, comparison, challenge, assurance and action
  • explain responsibilities to a team or client
Application notes and starter prompt

Use when

  • planning a governed AI workflow
  • choosing between Analyse, Compare, Improve, Defend, Discover, Approve, Assure and Decision & Action
  • designing an operating solution or pilot

Related registered methods and instruments: METHOD-GOV-WORKFLOW-006, INSTRUMENT-AI-ANALYSE, INSTRUMENT-AI-COMPARE, INSTRUMENT-AI-IMPROVE

Reliance note: Use to structure and allocate AI-assisted work. Skill selection does not validate evidence or grant permission to rely.

AI starter prompt

Given my task, choose the minimum GovAIaaS Skills shown in this visual, map them to the six-stage workflow and identify the human decision points.
Open full size Open underlying page
AI May Help Discover Candidate Mechanisms infographic.
discovery applicationD3advanced method

AI May Help Discover Candidate Mechanisms

Question answered: How can AI broaden causal exploration while governance prevents possibility from being mistaken for proof?

What it explains

  • the difference between collecting evidence and generating candidate mechanisms
  • five AI discovery capabilities
  • the six-stage SyncLogic and GovAIaaS mechanism-discovery flow
  • illustrative applications across climate, thermodynamics, materials, biology, medicine, energy and policy

How it could help

  • generate multiple causal explanations before converging on one
  • connect distant ideas and search larger possibility spaces
  • design alternative comparisons and discriminating tests
  • make uncertainty and permission-to-rely checkpoints visible
Application notes and starter prompt

Use when

  • a problem has competing causal explanations
  • an existing evidence base may be mechanism-poor
  • research, engineering or policy teams need a disciplined discovery pathway

Related registered methods and instruments: METHOD-MECHANISM-006, INSTRUMENT-AI-DISCOVER, INSTRUMENT-AI-COMPARE, INSTRUMENT-AI-DEFEND

Reliance note: Candidate mechanisms are provisional possibilities. Domain evidence, alternatives, discriminating tests and authorised human review are required before reliance.

AI starter prompt

Using the mechanism-discovery visual, generate several candidate mechanisms for my question, identify alternatives and discriminating tests, and stop before claiming confirmation.
Open full size Open underlying page
One Climate Claim, Many Claims infographic.
claim decomposition applicationD2worked application

One Climate Claim, Many Claims

Question answered: What necessary component claims are hidden inside a simple climate conclusion?

What it explains

  • the decomposition of “CO2 causes climate change” into eight linked claims
  • the movement from physical properties through attribution, impacts and policy
  • why a simple statement may be a complex bundle of dependencies
  • the need to audit components before accepting the conclusion

How it could help

  • turn a headline claim into an auditable claim chain
  • identify where evidence, mechanism, attribution and policy judgement enter
  • locate the weakest necessary component
  • separate scientific claims from later policy conclusions
Application notes and starter prompt

Use when

  • auditing the example climate claim
  • teaching claim decomposition and causal-chain analysis
  • designing a structured climate claim audit

Related registered methods and instruments: METHOD-CLIMATE-CLAIM-CHAIN-008, METHOD-CLIMATE-AUDIT-007, INSTRUMENT-AI-ANALYSE, INSTRUMENT-AI-DEFEND, INSTRUMENT-THREE-HATS

Reliance note: This is a worked climate-audit architecture, not a completed scientific adjudication. Each component requires current evidence and domain review.

AI starter prompt

Use this visual to decompose the climate claim into its eight components. For each component, list evidence needed, uncertainty and what failure would do to the overall conclusion.
Open full size Open underlying page
One Climate Claim. Four AIs. One Audit Template. infographic.
multi ai comparison applicationD3advanced method

One Climate Claim. Four AIs. One Audit Template.

Question answered: How can independent AI audits be compared using the same structured method?

What it explains

  • a common seven-part climate claim audit template
  • parallel use of ChatGPT, Grok, Gemini and Claude
  • comparison of agreement, disagreement and repeated weak links
  • why convergence should be treated as a process signal rather than proof

How it could help

  • standardise the unit and method of cross-AI comparison
  • identify repeated weak links across independent audits
  • make differences in assumptions, evidence use and uncertainty visible
  • reduce the temptation to choose an AI answer by brand preference alone
Application notes and starter prompt

Use when

  • comparing two or more AI audits of the same claim
  • testing process convergence and divergence
  • preparing an independent review or audit record

Related registered methods and instruments: METHOD-CLIMATE-AUDIT-007, METHOD-MULTI-AI-CONVERGENCE, INSTRUMENT-AI-COMPARE, INSTRUMENT-AI-DEFEND, INSTRUMENT-AI-ASSURE

Reliance note: AI convergence is not proof and may reflect shared training, prompting or assumptions. Independent evidence and human authority remain necessary.

AI starter prompt

Compare the independent AI audits using the same seven-part template. Report agreement, disagreement, repeated weak links and independence limitations without treating convergence as proof.
Open full size Open underlying page

AI use protocol

Use the visuals without turning explanation into authority.

  1. 1

    State the task

    Identify audience, purpose, consequence level and requested output.

  2. 2

    Select the minimum visual

    Choose the narrowest pathway that matches the task.

  3. 3

    Load the underlying page

    Use the source page and professional boundaries before making substantive claims.

  4. 4

    Separate knowledge status

    Keep REGISTERED content distinct from INFERENCE, EXTERNAL information, UNKNOWN matters and CONFLICTS.

  5. 5

    Stop at the reliance checkpoint

    State what the visual may support and what still requires evidence, expertise or human authority.

AI-readable knowledge surface

Load the entry, index, navigation map and visual application layer in order.

The machine-readable files allow an AI to establish current versions, registered entities, relationships and application pathways before answering.

1. AI-MANIFEST2. HDI-0023. HDIMM-0034. VKL-004

A visual application layer—not a substitute for evidence or judgement.

The library explains registered GovAIaaS concepts and intended applications. Task-specific recommendations created from these visuals remain AI inference unless separately registered and validated.

Recommended next page

Return to AI Knowledge Resources

See how the visual layer connects to the AI-MANIFEST, HDI and HDIMM.

Continue