Compound Fork — Robotaxi × AGI
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
| Robotaxi ↓ × AGI → | AGI_FAST_2027 prior 30% | AGI_MID_2029 prior 35% | AGI_SLOW_2031 prior 25% | AGI_WINTER_2036PLUS prior 10% |
|---|---|---|---|---|
ROBOTAXI_TESLA_2026 prior 40% | 121 claims · Σ|Δ| 15.84 | 117 claims · Σ|Δ| 15.32 | 124 claims · Σ|Δ| 16.27 | 125 claims · Σ|Δ| 16.29 |
ROBOTAXI_NATIONWIDE_2028 prior 45% | 123 claims · Σ|Δ| 16.00 | 119 claims · Σ|Δ| 15.48 | 126 claims · Σ|Δ| 16.43 | 127 claims · Σ|Δ| 16.45 |
ROBOTAXI_MASS_2030 prior 30% | 120 claims · Σ|Δ| 15.63 | 116 claims · Σ|Δ| 15.10 | 123 claims · Σ|Δ| 16.06 | 124 claims · Σ|Δ| 16.08 |
ROBOTAXI_DELAYED prior 20% | 123 claims · Σ|Δ| 16.01 | 119 claims · Σ|Δ| 15.49 | 126 claims · Σ|Δ| 16.44 | 127 claims · Σ|Δ| 16.46 |
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.