We scored every task in 341 occupations against AI products shipping today. Click any occupation to see exactly which tasks are changing and which companies are behind it.
Last updated November 12, 2026.
1. Pull each occupation's core tasks from O*NET
O*NET defines 8–15 tasks per occupation, ranked by importance. We select the tasks that cover 80% of the role's importance weight.
2. Score each task against shipped AI products
Perplexity searches the web for commercially deployed AI systems — not research papers or demos — and scores each task on a 7-point scale. Every non-zero score must cite a named product, company, and source URL. This includes robotics and industrial automation alongside software AI — both are commercially deployed systems transforming how tasks are performed.
3. Classify into a tier based on depth of change
We count how many tasks are deeply changed by AI — meaning AI handles major parts, leads the work, or fundamentally transforms the workflow. The count determines the tier:
Replacing — AI is actively displacing workers. Evidence of headcount reduction or role elimination, combined with deep task displacement.
Reshaping — The job still exists, but tasks and workflows are fundamentally different. Four or more core tasks have been deeply changed by AI, whether through displacement or augmentation.
Augmenting — AI helps with some tasks, but the core work is the same. Workers use AI tools without their day-to-day fundamentally changing.
Minimal change — AI has not meaningfully affected this occupation's core tasks.
Inspired by Andrej Karpathy's framework.
This is a research tool, not a career prediction. Use it alongside your own judgment and local labor market data.
The methodology above is simplified. Here is the full pipeline for anyone who wants to reproduce or audit our results.
Task selection
For each occupation, we query the O*NET Web Services API for task statements, importance ratings (IM scale), and frequency ratings (FT scale). We select tasks covering 80% of cumulative importance weight, capped at 8–15 tasks. If importance/frequency data is unavailable, we fall back to all “Core” tasks.
AI research via Perplexity
Each occupation's task list is sent to Perplexity's sonar-pro model with pro search enabled and high search context. The prompt includes the 7-point rubric (−3 to +3), disambiguation hints (alternate job titles, workflow-specific search terms), and domain exclusions (O*NET, BLS, job boards). Perplexity returns a score, evidence quality rating, deployment scope, and core-output match for each task, plus a list of products and workforce signals.
Quality-adjusted scoring
Raw task scores are not used directly. Each task's contribution is adjusted by three factors: core output match (does the AI product perform the task's core output, or just adjacent admin?), deployment scope (pilot = 0.45× weight, industry standard = 1.0×), and evidence quality (weak = 0.55×, strong = 1.0×). This prevents pilot-stage or loosely-matched tools from inflating scores.
Tier classification
We count deep-change tasks: tasks scoring ±2 or higher, or labeled “AI handles major parts.” Four or more deep-change tasks triggers “Reshaping.” For “Replacement,” we also require evidence of actual workforce displacement (headcount reduction, role elimination) or a strongly negative overall score.
Validation
Research results that fail validation — hollow retrieval (no meaningful citations beyond O*NET/BLS), schema failures, or missing task coverage — are quarantined and excluded from the treemap. Occupations without valid research appear as “Score pending.”
The full source code for the scoring pipeline is available on request. If you find an error in a specific occupation's scoring, please let us know.
143.1M US jobs, scored by AI exposure
Based on 341 occupations mapped to real AI products
Key findings
11% of the US workforce currently shows replacement-level AI exposure
Looking specifically at health care? How AI Is Changing Health Care Jobs — an evidence-based analysis of AI exposure across 55 health care professions.