Reports

V7 adds a task-concentration exposure buffer (Hampole et al.) and a demand-persistence proxy to the structural formula. Corrected 7 Jun 2026: the concentration term now buffers rather than amplifies exposure, matching the cited finding.

V6 introduced the two-axis structural formula with 4-source exposure ensemble, human bottleneck, and explicit demand resilience. Superseded by V7.

What the task-weighted shadow model would change, which promotion gates remain open, and why it stays separate from the live headline score.

Shadow score is published. Promotion into the headline model still depends on validation and anchor-review sign-off.

The next scientific release program after V4.3: augmentation heterogeneity, empirical mobility, posterior uncertainty, and realized-risk forecasting.

Planning surface for the V5 program. Sidecars are published and the first integrated experimental model is now available separately.

The first integrated V5 candidate now combines posterior uncertainty, augmentation heterogeneity, empirical mobility, and realized-risk calibration into one auditable model output.

Current validation snapshot: structural 2/2, realized 2/4. This remains separate from the live headline score.

Annual wage context for occupations in the two highest AI exposure bands, with sector breakdowns and explicit limits on what the totals mean.

Based on V8 scoring, updated 2026-06-11

Initial V3 scoring results: 562 occupations scored across exposure, bottleneck, and market resilience. Key findings on paradox roles, highest-risk occupations, and theory vs practice gaps.

Archived snapshot rebuilt under the current V7 pipeline, updated 2026-06-11