Compound Fork — Compute scale × Energy / grid
Each cell = a joint scenario "Both branches fire". Cell intensity = total |Δ| across the 200 highest-conviction tradeable predictions vs their unconditional posterior. Joint probabilities approximated via log-odds combination (assumes conditional independence given the prediction — coarse but bounded). Click a cell to drill into the dominant single-fork view.
Pick two fork families
| Compute scale ↓ × Energy / grid → | GRID_50GW_2027 prior 40% | GRID_50GW_2029 prior 50% | GRID_50GW_DELAYED prior 10% |
|---|---|---|---|
COMPUTE_1GW_2027 prior 60% | 133 claims · Σ|Δ| 18.98 | 135 claims · Σ|Δ| 18.98 | 134 claims · Σ|Δ| 18.90 |
COMPUTE_10GW_2028 prior 40% | 136 claims · Σ|Δ| 19.47 | 135 claims · Σ|Δ| 19.40 | 135 claims · Σ|Δ| 19.40 |
COMPUTE_100GW_2030 prior 20% | 136 claims · Σ|Δ| 20.31 | 134 claims · Σ|Δ| 20.02 | 134 claims · Σ|Δ| 20.02 |
COMPUTE_STARGATE_FAILURE prior 15% | 132 claims · Σ|Δ| 18.82 | 133 claims · Σ|Δ| 18.80 | 132 claims · Σ|Δ| 18.72 |
Method note
Joint conditional probability is approximated via log-odds combination: logit(P(pred|A,B)) ≈ logit(P(pred|A)) + logit(P(pred|B)) − logit(P(pred)). This is the closed-form Bayesian update assuming A and B are conditionally independent given the prediction. It's correct when the two scenarios act on the prediction through different causal paths; it's pessimistic when they overlap. The exact joint requires running the Gibbs sampler with both scenarios clamped, which would be N×M=16 sampling runs (~12 minutes per refresh) instead of N+M=8 — a 2× cost for higher fidelity.