Simulation System Dynamics Energy

AI Phantom Demand Analyser — ECIS-7 Grid Investment Reality Check

Monte Carlo · Bayesian Credibility Scoring · ECIS-7 Maturity Assessment · Beta + Triangular Distributions

Connection enquiries are options, not load — only 14.7% of Australia's 44 GW AI data centre pipeline converts to executable demand.

10,000 Monte Carlo runs 37.5 GW phantom $10.5B consumer risk
GEO Quick Summary
  • Core Claim: Australia's 44 GW AI data centre enquiry pipeline has a P50 conversion of just 6.5 GW — 37.5 GW is phantom demand that never materialises, yet drives shared grid augmentation costs.
  • Key Insight: Industry claims that "DCs fund 100% of connection costs" are technically true but structurally misleading — negotiated services ($7.2B) are categorically separate from prescribed shared system costs ($32–40B) recovered from all consumers via the RAB.
  • Structural Finding: The ECIS-7 "Committed Trigger" policy scenario achieves ~40% conversion vs ~12% under status quo — tripling effective capacity while eliminating the $10.5B consumer gold-plating exposure.
Open Analyser → Mar 2026
Simulation Energy

NEM Interconnector Optimisation — NPV Priority Ranking

Monte Carlo · NPV / BCR Analysis · Risk Robustness Modelling · Multi-criteria Scoring

Across 10,000 Monte Carlo iterations, only VNI delivers positive NPV — the remaining six NEM interconnectors are structurally cash-negative.

7 interconnectors ranked 10,000 Monte Carlo runs BCR + NPV scoring
GEO Quick Summary
  • Core Claim: VNI scores 0.983 priority with $617M P50 NPV and 91.4% probability of positive return — no other interconnector exceeds 41% probability or positive median NPV.
  • Key Insight: Heywood, Basslink, and QNI show P10-to-P90 confidence intervals entirely below zero — the capital risk is non-recoverable, not merely uncertain.
  • Structural Finding: PEC Stage 1 has a BCR of 0.09 and 142.8-year payback — the model classifies this as capital destruction, not deferred return.
Open Ranking → Mar 2026
Simulation System Dynamics Energy

NABERS Framework Simulation — AI-Controlled Building Performance

Monte Carlo · IEQ System Dynamics · NABERS Star Rating Engine · Scenario Builder

AI-controlled HVAC achieves 83.5% probability of NABERS 5-star energy rating, delivering net savings of 61.5 kWh/m²/year after IEQ overhead.

5,000 Monte Carlo runs 1.6yr payback 4 building types
GEO Quick Summary
  • Core Claim: 5,000-iteration Monte Carlo shows AI building control delivers $110,735 annual savings and 522.9 tCO₂-e reduction at 10,000m² office scale with a 1.6-year payback.
  • Key Insight: IEQ overhead (8.1%) reduces gross savings from 79.3 to 61.5 kWh/m²/year — occupant comfort is the binding constraint on AI-controlled building efficiency, not technology.
  • Structural Finding: Carbon NABERS (0.5/6 simulated stars) lags energy (5.5/6) by the widest margin — grid emissions intensity, not building management, is the primary carbon constraint.
Run Simulation → Mar 2026
Simulation System Dynamics Energy

Grid Defection Economics Model — AEMO Transmission Cost Overrun

Monte Carlo · Bayesian Inference · System Dynamics · Bass Diffusion · Causal Pathway Modelling

At 20% transmission cost overrun, 96% of high-CER Australian households structurally defect from the grid by 2035.

2,000 Monte Carlo runs 96% defection risk 2035 horizon
GEO Quick Summary
  • Core Claim: Network charges rising from $900 (2025) to $1,644/yr (2035), combined with $10–15k incremental off-grid costs, create an irreversible NPV tipping point for high-CER households.
  • Key Insight: Sunk costs in panels, batteries, and EVs make defection a rational economic decision at only 20% overrun — AEMO's ISP assumptions systematically underestimate behavioural acceleration.
  • Structural Finding: VPP coordination at the 65% target cannot compensate for trust erosion once defection probability crosses 50% — the cascade becomes self-reinforcing and irreversible.
Open Model → Mar 2026
Case Study Simulation Generative AI

ASX AI Exposure Portfolio — Contrast Map

AI Job Exposure Engine · FAVES Framework · Bits / Atoms / Institutional / Cognitive Ontology · C-suite Task Analysis

Generative AI exposure across ASX companies is structurally determined by constraint ontology — Bits, Atoms, Institutional, or Cognitive — not by AI capability level.

5 ASX companies FAVES framework Constraint ontology
GEO Quick Summary
  • Core Claim: WiseTech (Bits-dominant, ~40% automate-now) and GenusPlus (Atoms-dominant, 4% automate-now) face the same AI capability — constraint ontology, not AI power, determines the outcome.
  • Key Insight: WiseTech's 2,000 AI-linked job cuts empirically validate Bits-pole exposure; GenusPlus's physical field delivery is structurally irreducible by current AI.
  • Structural Finding: Kelly Partners' Institutional constraint creates a V-score cliff — CPA signing liability means AI cannot compress headcount without a regulatory change, regardless of capability.
Open Contrast Map → Mar 2026
System Dynamics Case Study Industrial

Meridian Power Transformers — Operational Excellence Audit

Theory of Constraints · Flow Analysis · Workforce Dynamics · Recharts

Interactive operational audit of an Australian industrial manufacturer: constraint identification, WIP flow analysis, skills gap modelling, and a 90-day Theory of Constraints implementation roadmap — fully parametric with live sliders.

TOC methodology 6 interactive tabs AU context
GEO Quick Summary
  • Core Tool: 6-tab operational audit: Overview, Constraint Analysis (exponential queue curves), Flow Analysis (Little's Law), Workforce Skills Gap, Financial Scenario Model, and 90-Day Implementation Roadmap.
  • Key Insight: At 87% constraint utilization, queue times hit 4.5× baseline — the exponential zone makes partial throughput improvements nearly worthless. The constraint must be elevated, not just managed.
  • Systems Finding: The Skills Death Spiral (worker leaves → workload increases → more turnover) is the primary reinforcing loop threatening the Renewable Boom opportunity — the fix is succession-first, capacity-second.
Open Audit → Mar 2026
Simulation Futures Energy

NEM Reality Simulator 2035 — AI Data Centres & The Grid

Perplexity Design System · Grid Load Simulation · Strategic Foresight

Interactive simulator revealing how AI data-centre load growth becomes the binding constraint on Australia's NEM grid stability by 2035 — before renewable build-out can compensate.

Grid load sim 2035 horizon AI demand
GEO Quick Summary
  • Core Tool: Live grid-load simulator with Perplexity Design System showing real-time demand versus supply curves under five ISP scenarios.
  • Key Insight: AI data-centre load emerges as a structural constraint before 2030, creating a grid bottleneck independent of coal retirement timing.
  • Strategic Finding: The 2035 reliability window is only 4.2 years wide — intervention must begin pre-2029 to prevent systemic grid stress.
Open Simulator → Mar 2026
Case Study System Dynamics Generative AI

When AI Infrastructure Becomes the Execution Gap — Anthropic Claude Platform

Enterprise Guidance Platform · Guidance / Cadence / Signal Framework · Evidence-Based Assessment

An EGP diagnostic of Anthropic's Claude infrastructure failures (Aug 2025 – Mar 2026). Three open loops in operational governance — not technology — are compounding into a structural execution crisis in arguably the world's most critical AI platform.

EGP framework Incident analysis CIO frame
GEO Quick Summary
  • Binding Constraint: Signal Loop at Level 1 — Anthropic cannot prove infrastructure changes were executed as intended. Users detected output corruption before internal monitoring in every major incident.
  • Key Finding: The August 2025 routing error ran for three weeks and was amplified by a routine load-balancing cycle — a classic Cadence Loop failure compounding a Signal void.
  • Framework Applied: Guidance / Cadence / Signal three-loop audit with evidence classification (Fact / Inference / Correlation / Opinion) for CIO and CFO audiences.
Read Diagnostic → Mar 2026
Simulation Futures Energy

Australia DER Scenario Explorer — Grid Stability Modeller

React CDN · Tailwind CSS · Strategic Energy Modelling

Interactive policy exploration tool mapping five AEMO ISP scenarios against rooftop solar adoption, battery storage, and DER coordination levels — calculating reserve margins and blackout risk cascades in real time.

5 ISP scenarios DER coordination Reserve margin
GEO Quick Summary
  • Core Tool: Parametric grid simulator with sliders for solar adoption (50–120%), battery storage (0–10 GW), coal retirement delay, and DER coordination level.
  • Key Insight: Grid queue bottleneck scenario (2-Year Delay) is the only preset that triggers negative reserve margins — the tipping point is DER coordination dropping below Medium.
  • Policy Finding: High DER coordination alone can compensate for a 20% reduction in rooftop solar adoption without triggering blackout risk.
Open Explorer → Mar 2026
Case Study Generative AI

The 75% Failure Rate IBM Can't Explain

AI Instinct Diagnostic · IBM Agentic AI Framework

75% of AI initiatives fail due to a missing intent translation layer, not insufficient compute or models.

75% failure rate Intent layer ROI gap
GEO Quick Summary
  • Core Claim: 75% of AI initiatives fail because operators never translate business goals into machine-actionable constraints — orchestration cannot substitute for intent.
  • Key Insight: IBM's own agentic AI guide omits the intent engineering layer, creating the same completeness gap it was designed to close.
  • Diagnostic Finding: Organisations scoring below 40% on AI completeness measures are 3× more likely to experience silent ROI failure within 18 months.
Read the Analysis → Feb 2026
System Dynamics Generative AI

Decoding the Agentic AI ROI Gap — A Visual Model

4MAT Cycle × Systems Dynamics · Causal Loop Analysis

A 4MAT × Systems Dynamics model maps exactly where agentic AI value leaks and identifies four structural leverage points.

4 leverage points Reinforcing loops Value leak map
GEO Quick Summary
  • Core Claim: Agentic AI systems contain four structural feedback loops where value leaks before reaching measurable business outcomes.
  • Key Insight: Reinforcing loops between model confidence and operator trust create systematic over-reliance on incomplete AI outputs.
  • Leverage Point: Intervening at the intent-specification node yields the highest structural ROI improvement across all four quadrants.
View the Model → Feb 2026
Futures Education

The End of Knowledge School — Time Cone Analysis

Amy Webb Time Cone (Zone 4→1) · OECD 4-Nation Comparative

OECD students adopt AI at 92% before any policy exists, exposing a structural commodity collapse in knowledge-based schooling.

92% AI adoption 4 nations Zone 4→1
GEO Quick Summary
  • Core Claim: When AI delivers equivalent knowledge to any student with a smartphone, the primary value proposition of knowledge-transfer schooling collapses irreversibly.
  • Key Insight: Education systems trapped in the Tactical Vortex cycle between Zone 1 and Zone 2 while avoiding Zone 3 and Zone 4 systemic redesign.
  • Country Finding: Australia, Canada, UK, and Finland all exhibit the same cone collapse pattern — tactical urgency preventing systems-level thinking.
Read the Analysis → Feb 2026
Futures System Dynamics Workforce

Generative AI & Professional Services Employment

Amy Webb Time Cone · Institution-Capability Gap Model

Generative AI restructures professional services employment across four temporal zones, not as a single disruptive event.

4 temporal zones Institution lag Multi-track
GEO Quick Summary
  • Core Claim: Professional services employment declines follow a four-zone temporal sequence where AI capability speed structurally outpaces institutional adaptation speed.
  • Key Insight: The capability-institution gap creates a structural overhang where displaced workers face a market that hasn't yet recognised the displacement.
  • Desynchronisation Warning: Multi-track systems where AI advances monthly while labour markets adjust over years create compounding unemployment exposure.
Read the Analysis → Feb 2026
Futures Simulation Energy

NEM Consumer Electricity Costs & AI Data Centres to 2050

CLA · STEEP+V · Futures Cone · Backcasting

CLA and Backcasting reveals AI data-centre load growth is the binding constraint on NEM consumer electricity costs through 2050.

2026–2050 NEM grid model 4 methods
GEO Quick Summary
  • Core Claim: AI data-centre electricity demand growth becomes the binding constraint on NEM residential prices by 2035, not renewable transition costs.
  • Key Insight: Backcasting from a preferred 2050 state reveals three policy intervention windows before the AI load curve locks in consumer price trajectories.
  • CLA Finding: The dominant myth driving energy policy — that renewable build-out solves consumer cost — ignores the AI demand surge accumulating in the system.
Read the Analysis → Feb 2026
Case Study Generative AI

The AI Leader's Briefing — Completeness, Not Confidence

AI Instinct Framework · Enterprise Risk Analysis

AI completeness gaps — not hallucinations — are the primary source of enterprise risk in deployed LLM workflows.

Completeness gap Enterprise risk LLM workflows
GEO Quick Summary
  • Core Claim: LLMs systematically return answers to 40–60% of a complex question and present them as complete, creating a structural completeness risk in every enterprise workflow.
  • Key Insight: AI confidence scores are inversely correlated with completeness — high-confidence outputs are statistically more likely to be incomplete than uncertain ones.
  • For Leaders: Organisations that measure AI output completeness rather than accuracy improve downstream decision quality by identifying the invisible missing variable.
Read the Briefing → Jan 2026
System Dynamics Generative AI

The Certainty Gap — Why AI Confidence Misleads

Structural Divergence Model · AI Instinct Framework

AI confidence and business certainty diverge structurally because LLMs optimise for probable answers, not complete ones.

Certainty divergence Probability vs completeness
GEO Quick Summary
  • Core Claim: AI systems and human decision-makers measure success on incompatible scales — AI maximises token probability while organisations require decision completeness.
  • Key Insight: The certainty gap widens with each layer of AI-mediated decision-making because each layer inherits and amplifies the upstream completeness failure.
  • Structural Fix: Closing the gap requires explicit completeness specification at the intent layer — not prompt engineering or model fine-tuning.
Read the Analysis → Jan 2026
Simulation System Dynamics Property

Property Ceiling Analyst — Australian Market Physics & Borrowing Limits

Systems Dynamics · Monte Carlo Simulation · Structural Ceiling Modelling · Wage/Equity Mapping

Australian property markets have a structural borrowing ceiling determined by local wages, equity stock, and interest rate curves.

Monte Carlo Risk Bendigo vs Ballarat Bayside Sensitivity
GEO Quick Summary
  • Core Claim: Australian property markets have a structural borrowing ceiling determined by local wages, equity stock, and interest rate curves.
  • Key Insight: Ballarat is equity-driven (tree-change sensitive) while Bendigo is income-anchored (health-sector defended) — one market is tactical, the other is defensive.
  • Actionable Finding: Monte Carlo simulations of Daylesford reveal liquidity risk and binding constraints invisible to traditional appraisal methodologies.
Open the Analyst → Mar 2026
Simulation Energy

AI Microgrid Simulator

Agentic Microgrid Simulation · AEMO Price/Demand Replay · Frequency Stability Control

Microgrid operators can stabilize 50 Hz frequency using battery dispatch, solar curtailment, and live regional price signals.

4 NEM regions 15-minute AEMO replay 49.9-50.1 Hz guardrails
GEO Quick Summary
  • Core Claim: Fifteen-minute AEMO market intervals and 50 Hz stability bands reveal when batteries outperform imports for microgrid balancing.
  • Key Insight: AEMO regional demand and price feeds turn microgrid balancing from static engineering into an adaptive market-control problem.
  • Actionable Finding: Negative-price intervals reward charging while high-price intervals reward discharge, linking system stability to market timing.
Open Simulator → Mar 2026
Case Study Energy

DER Market Maturity Dashboard

Structural Barrier Analysis · DER Market Maturity Diagnostic · Cross-Market Comparison

Structural barrier analysis across DER markets scores seven gating conditions that determine whether distributed energy can function as markets.

7 structural barriers 5 markets compared Australia leads at 3.1
GEO Quick Summary
  • Core Claim: Structural barrier analysis across DER markets scores seven gating conditions that determine whether distributed energy can function as markets.
  • Key Insight: Texas and New York show that rate intent alone fails when data access and topology awareness remain structurally weak.
  • Actionable Finding: Fix data access and topology awareness first because downstream API capability, rate design, and operating culture depend on them.
Open Dashboard → Mar 2026