How the model works https://metamachine.sukanthoriginal.com/guide/methods.html Follow the calculation from output targets to input needs, supply constraints and the unexplained remainder. Choose the question before calculating the answer The total doubling goal and its sector mix are selected scenarios. Historical data informs the starting estimate, spending-response factors and comparisons, while applying those relationships to 2035 requires assumptions. The resulting gaps and shortages answer a conditional what-if question. See the role of each number (https://metamachine.sukanthoriginal.com/guide/what-the-numbers-mean.html) Start with the accounting The code imports the earlier requirements table and the repaired baseline. In each layer it reconciles today, on-course growth, tier 1 and tier 2 with the target. It then checks that the layers sum to twice the 2026 estimate. Turn output into input needs For a fixed ratio, input needed equals output multiplied by input per unit of output, summed across layers. Some inputs use additional structure. Labour is split among people, AI and robots. Electricity includes machine work. Copper, lithium and rare earths use an inherited substitution approximation. Input need = sum of (layer output × input needed per unit) Compute each constraint separately The code removes eligible missing output until the input requirement fits supply, or until it runs out of eligible output. It protects on-course output. The layer ordering is based on input intensity or a labour proxy and is fixed during each test. A bisection calculation finds the final partial reduction. Tier 2 is tested against precedent supply with the full target attempted. Tier 1 is tested against on-course supply with on-course output plus tier 1 attempted. The two recorded blocks are added for each shortage. Extra diagnostic fields record competition between tiers and any remaining input shortfall. Handle overlapping constraints The individual shortage totals overlap. For each layer and tier, the code credits the largest individual block and calls the remainder unexplained. This gives $141.4T explained and $41.5T unexplained. Taking the maximum assumes the smaller blocks are contained within the largest one. The code has no task-level overlap map to establish that containment. The explained amount is therefore conditional on this attribution rule. See a worked overlap example (https://metamachine.sukanthoriginal.com/guide/unexplained.html) Interpret the checks The committed checks test reconciliation, cell limits, labels, selected criteria and unchanged upstream files. The current code relaxes two brief requirements. It accepts all eligible SOLID S/E picks when fewer than three exist, and allows shortages above $1T to pass without two alternative tests. The ranking also keeps the first variant matching each scenario. Clinical approval and cropland each have two S6 alternatives, so one endpoint for each is absent from the scenario-ranking columns. Their full ranges include both endpoints. Inspect the calculation code Source files · code/model.py (https://metamachine.sukanthoriginal.com/data/code/model.py) · code/v4.py (https://metamachine.sukanthoriginal.com/data/code/v4.py) · code/check.py (https://metamachine.sukanthoriginal.com/data/code/check.py) · code/common.py (https://metamachine.sukanthoriginal.com/data/code/common.py) These files are the source snapshot used for the model outputs. The downloadable input tables provide the files they reference. Read the evidence labelsS, E, A and robustness → (https://metamachine.sukanthoriginal.com/guide/evidence.html) Check who would buy the outputIncome and prices → (https://metamachine.sukanthoriginal.com/guide/demand.html) Compare production alternativesWhat changes across tests → (https://metamachine.sukanthoriginal.com/guide/scenarios.html) Follow the original inputsSource registries and data → (https://metamachine.sukanthoriginal.com/guide/sources.html)