Model 2 — Visual Alignment Engine

Spin the Machine

A visual alignment engine — where geometry, colour, and meaning propagate through the reels.

The Spin Machine is the visual front-end of DMI: a dimensional engine that aligns human tags, machine tags, and vocabulary tags through geometric selectors and chromatic propagation. Every spin recolours the frame, recalculates curvature, and regenerates the Insight Stories — turning alignment into a visual, playable, conceptual experience.

DMI · METRIC & CHROMATIC ALIGNMENT COCKPIT

Multi‑Scope & Vocabulary Compass

Load any registered vocabulary, switch between segmented (M3) and dynamic (M4) scopes, and watch palette, domain and drift alignment update in place.

  1. 1 · Source
    Anchor the submission identity
  2. 2 · Vocabulary
    Extract terms from the source
  3. 3 · Align
    Metric + chromatic alignment
  4. 4 · Edit → re-align
    Iterate integrity and consistency

1 · Load source

The source is the origin of the vocabulary — the only truth. Everything else is derivative. No vocabulary extraction may occur before a source is loaded.

Neutral pre-pipeline state — no source is anchored yet, so vocabularies are read without source attribution.anchoredRegister a source (M2) →
Paste website
Pasted sources are anchors only — extraction still runs in step 2.
Source id
any
challenge
Vocabulary
empty
not extracted
Alignment
empty
not computed
Drift
empty
no reading

2 · Load vocabulary

Vocabulary is extracted from the source — never invented, guessed, inferred, combined or edited. Extraction classifies terms into centre, source ring, palette ring and domain ring (including SOV and SOV-AI).

0 challenge vocabularies selectable · 0 anchored to ALLRegister more in the Library →

3 · Align (metric + chromatic)

Alignment is computed from the extracted vocabulary — not from the source directly, not from metadata, not from previous runs.

Load a vocabulary first.

4 · Edit → re-align loop

phase: idle · iteration 0

The diagnostic phase is done. Improve Information Integrity and Contextual Consistency by editing the source, then re-extract, re-compare the rods, re-align and update the galleries. Repeat until drift is minimal.

Complete LOAD SOURCE → LOAD VOCABULARY → ALIGN to activate the loop.

Source legend

Challenge Vocabulary

MISAICCLMDCE

Segment legend

M3 — Scopes
  • narrow
  • medium
  • wide
  • filewide
M4 — Modes
  • propagated
  • refined
  • multilingual
  • industrial
Colour 1 — Shared / propagatedColour 2 — High evidenceColour 3 — Local / single sourceColour 4 — DormantClimateChangeClimate Change — no terms in this scopeAICredibilityAI Credibility — no terms in this scopeMisinformation &InformationIntegrityMisinformation & Information Integrity — no terms in this scopeData Centres &EnergyEfficiencyData Centres & Energy Efficiency — no terms in this scopeAISovereigntyAI Sovereignty — no terms in this scopeNarrowM3 scope — One segment carries the term. The tightest scope: meaning is local, evidence is a single passage. · reading Challenge VocabularyMediumM3 scope — Two or three segments inside a single challenge — a term that holds across a section but not across files. · reading Challenge VocabularyWideM3 scope — Four or more segments, or more than one challenge — the term travels beyond its origin passage. · reading Challenge VocabularyFilewideM3 scope — Every extracted term, scoped to the whole corpus. The unfiltered baseline every other scope is cut from. · reading Challenge VocabularyPropagatedM4 mode — Terms that already crossed a domain boundary. Propagation is what makes a term reusable in another challenge. · reading Challenge VocabularyRefinedM4 mode — The high-evidence core: terms carried by at least three occurrences, with the long tail cut away. · reading Challenge VocabularyMultilingualM4 mode — Terms that carry a German, French and Spanish surface form, so alignment survives translation. · reading Challenge VocabularyIndustrialM4 mode — Plant-level vocabulary: energy, cooling, grid, capacity and the machinery words a facility actually reports on. · reading Challenge VocabularyCHALLENGERing 1 — centre: loaded vocabulary identity (never coloured).
All sources (unanchored)
source loaded — extract vocabulary
built-in
Polarity pointer — dominant colour cluster on the palette ring.Domain pointer — dominant domain on the domain ring.Scope pointer — active scope or mode (M3/M4) on the source ring.Identity pointer — highlights the loaded vocabulary at the centre.

Palette ring · Domain ring · Source ring · Identity centre

Drift legend

  • polarity drift — shift in shared vs. local term balance
  • colour drift — shift in polarity cluster distribution
  • domain drift — shift in domain term distribution

Palette legend

  • Colour 1Polarity AShared / propagated
  • Colour 2Polarity BHigh evidence
  • Colour 3Polarity CLocal / single source
  • Colour 4Polarity DDormant

Domain legend

  • CLMClimate Change0%
  • AICAI Credibility0%
  • MISMisinformation & Information Integrity0%
  • DCEData Centres & Energy Efficiency0%
  • SOV-AIAI Sovereignty0%

SOVEU Sovereignty

Global Challenge

EU Sovereignty is the global challenge. AI Sovereignty (SOV-AI) is the domain inside it that the compass measures: vocabulary terms classify into SOV-AI layers, and those readings feed the domain ring. The challenge frames the question; the domain carries the metrics.

Read the full EU Sovereignty challenge summary →

SOV-AIAI Sovereignty

Domain · compass ring

Authoritative sources agree that EU AI sovereignty is the ability to develop, deploy, govern and trust AI systems without strategic dependence on non-EU actors. It is not merely autonomy: it is strategic capability, regulatory power, infrastructure control and trustworthiness — five simultaneous layers, not alternatives.

  • European Commission — AI Act + AI Innovation Package
  • European Parliament — AI governance + legal sovereignty
  • European Council / Member States — strategic autonomy
  • EU AI Office (2024–2026) — capability + compliance + safety
  • Think tanks — Bruegel, CEPS, EPRS, ECFR, Chatham House

Capability sovereignty

0 terms · 0%

European Commission · AI Act + AI Innovation Package

Frontier models, compute, chips, data and talent. Europe builds its own AI capability, supports open-source and European foundation models, and reduces dependency on non-EU cloud and compute providers so it can shape AI according to its values and strategic interests.

Regulatory sovereignty

0 terms · 0%

European Parliament · AI governance + legal sovereignty

Legal primacy of the AI Act, governance through the AI Office and national authorities, data sovereignty via GDPR, the Data Act and data spaces, and algorithmic sovereignty through transparency, explainability and auditability. Europe sets the rules rather than following them.

Infrastructure sovereignty

0 terms · 0%

European Council / Member States · strategic autonomy

Cloud, HPC and supercomputing capacity, European data spaces, cybersecurity, joint procurement of AI infrastructure and industrial resilience — the physical and institutional substrate of AI independence.

Trust sovereignty

0 terms · 0%

EU AI Office (2024–2026) · capability + compliance + safety

European capacity to evaluate and govern frontier AI: validation and testing centres, risk classification, evaluations, red-teaming, monitoring and enforcement. This is credibility sovereignty.

Strategic sovereignty

0 terms · 0%

Bruegel · CEPS · EPRS · ECFR · Chatham House

Autonomy from US and Chinese AI ecosystems: strategic capability to act in AI without external constraints, control over critical AI infrastructure and standards, defence autonomy through AI, and collective European capacity to compete at scale.

MISMisinformation & Information Integrity

Global Challenge · domain ring

Authoritative sources agree that misinformation is not a single failure of truth but a layered failure of integrity. Provenance, evidence, manipulation, amplification and resilience are simultaneous dimensions — a claim can be sourced yet unevidenced, evidenced yet manipulated, accurate yet amplified out of context.

  • European Commission — Code of Practice on Disinformation
  • Digital Services Act — systemic risk assessment (Art. 34–35)
  • EDMO — European Digital Media Observatory
  • IFCN / fact-checking standards
  • Reuters Institute · EBU · C2PA content credentials

Provenance integrity

0 terms · 0%

C2PA · EBU · Reuters Institute

Where a claim came from, who published it, and whether the chain from origin to publication survives inspection: authorship, citation, content credentials, edit history and attribution of images, quotes and data.

Evidence integrity

0 terms · 0%

EDMO · IFCN fact-checking standards

Whether the claim is supported by verifiable evidence: data, documents, primary reporting, reproducible measurement, and the distance between what is asserted and what is demonstrated.

Manipulation detection

0 terms · 0%

EU Code of Practice on Disinformation · DSA Art. 34–35

Deliberate distortion: fabricated content, synthetic media, deceptive framing, decontextualised material, coordinated inauthentic behaviour and impersonation of trusted institutions.

Amplification dynamics

0 terms · 0%

DSA systemic-risk reporting · platform transparency data

How a claim travels: recommender exposure, virality, echo-chamber concentration, cross-platform propagation and the speed at which a correction can catch up with the original.

Societal resilience

0 terms · 0%

European Commission · media literacy and democratic resilience

Capacity to withstand distortion: media literacy, plural and funded journalism, institutional trust, transparent correction practice and democratic oversight of information infrastructure.

Domain readings feed the ring: 0 SOV-AI terms and 0 MIS terms of 0 in Filewide.Misinformation challenge summary →

Active scope — Filewide

Every extracted term, scoped to the whole corpus. The unfiltered baseline every other scope is cut from.

Terms
Domains
Segments
Shared
M3 — SegmentedOrigins: MIS · AIC · CLM · DCECycle: v0 · s0 · m0

Domain alignment

  • Climate Change0%
  • AI Credibility0%
  • Misinformation & Information Integrity0%
  • Data Centres & Energy Efficiency0%
  • AI Sovereignty0%

Drift preview

Switch scopes to compare polarity, colour and domain distribution.

View in Vocabulary Viewer →Open Alignment Gallery →Open Shared Vocabulary →Back to Library →

Hexagon — Human Direction

Six thinking directions. When crisp, it activates Input + Vocabulary reels and propagates semantic tags.

Octagon — Agentic Scope

Three scope partitions. When crisp, it segments meaning and reveals scope-driven machine tags.

Circle — Chromatic & Domain

Colour triads and domain identity. When crisp, it activates chromatic propagation, domain filtering, and the category tree.

Input Reel

Your tags — the human reference. Loaded from text, files, URLs, or challenge presets.

Scope Reel

Machine-attributed tags — segmented by the Octagon. Shows consistency, cohesion, variability, and cross-segment alignment.

Vocabulary Reel

Reference terms — aligned with the Input on spin. Carries semantic, chromatic, and domain meaning.

Chromatic Propagation

Colour flashes show 100% alignment between human and machine tags.

Scope Cohesion Graph

Fill = consistency · Ring = cohesion · Pulse = variability · Links = cross-segment alignment.

Insight Stories

Human-facing and machine-facing interpretations generated from the alignment scores.

Top of the Pops

The strongest alignment combination across scopes.

01

Choose a Challenge

MIS · AIC · NFA · CLM · EDC

02

Load Metric Rod-Sets

Tiny, authoritative, machine-readable calibration units.

03

Build the Vocabulary

Semantic, chromatic, domain, scope, drift-risk, stability tags.

04

Spin the Machine

Reels activate. Selectors propagate. Colours flash. Scores update.

05

Read the Outputs

Alignment scores · drift warnings · stability forecasts · insight stories · vocabulary maps · rod-set calibration reports.

Run M1 Alignment →← Return to Challenge Earth Demo