Release truthful AI exposure rankings

[kirso]

Summary

• replace probability-like job-change language with the human-readable, SEO-aligned AI Exposure Rank across current pages, metadata, structured data, downloads, and generated social images
• publish the clean V8 Singapore contract with separate substitution, augmentation, demand, adoption, confidence, sensitivity, and transition context
• withdraw unsupported US/global detail scoring, replace the misleading calculator path, and make the BLS cross-country result explicitly null and non-causal
• widen data-heavy page shells and make tables/cards responsive without page-level mobile overflow
• replace risk-weighted wage claims with descriptive wage context for occupations in the High and Very High exposure bands

Verification

• bun run build:release-data
• bun run test:methodology — 40 passed
• bun run verify — 222 validation checks passed; Svelte, ESLint, Prettier, and release checks passed
• bun run build
• rendered QA at 390px and 1440px for home, roles, data, wage report, methodology appendix, compare, and occupation detail

Important interpretation

An AI Exposure Rank of 72/100 means an occupation is more exposed to current AI capabilities than approximately 72% of the Singapore reference occupations. It is not a 72% probability of job loss, a task share, or an employment forecast.

The deduplicated V8-to-BLS cross-country check is null (Spearman rho 0.0138; non-significant), and the public artifacts disclose that result rather than presenting it as external validation.

[cloudflare-workers-and-pages]

Deploying aiworkindex with    Cloudflare Pages

Latest commit:	26ac727
Status:	✅  Deploy successful!
Preview URL:	https://431b2c42.aiworkindex.pages.dev
Branch Preview URL:	https://v8-confidence-ratings.aiworkindex.pages.dev

View logs