01
Search any occupation
'Paralegal', 'truck driver', 'radiologist' — instant scorecard.
civlab·civlab-ai-replacement-map.psyverse.fun·matrix
▣ AI Replacement Map
An honest, occupation-by-occupation map of which jobs LLM-era AI is replacing, augmenting, or leaving alone.
Most AI-job-loss content is either panic ('AI will take everything') or denial ('it just helps with email'). Both lose information. AI Replacement Map looks at occupations one at a time. For each, it scores: what fraction of tasks AI can already do, what fraction it augments, what fraction it cannot touch in the next 5 years, and the empirical wage / posting trend in the relevant labor market over the last 24 months. The result is a heatmap of where the actual displacement is happening, the augmentation is happening, and the protective moats remain.
The scores have explicit uncertainty bounds and explicit assumptions. Disagree? Fork the methodology and re-score.
Interactive demo
Use it — don't read about it.
occupation
automate
augment
moat
Translator
65%
25%
10%
Paralegal
55%
30%
15%
Tax preparer
55%
30%
15%
Copywriter
55%
30%
15%
Graphic designer
45%
40%
15%
Data analyst
40%
50%
10%
Accountant
40%
45%
15%
Front-end developer
30%
55%
15%
Back-end developer
25%
55%
20%
Radiologist
20%
55%
25%
Primary-care physician
10%
50%
40%
Elementary teacher
10%
30%
60%
Construction manager
10%
30%
60%
Commercial pilot
10%
20%
70%
Long-haul truck driver
5%
10%
85%
Plumber
5%
10%
85%
Psychiatrist
5%
30%
65%
Barber / hairdresser
0%
5%
95%
Note: automate + augment + moat = 100% per occupation. Each is the share of that occupation's task decomposition.
Features
What this platform promises.
02
Heatmap by industry
Where is the most exposure right now?
03
Personal exposure check
Paste your job description; get task-level breakdown.
Modules
Modules that compose this platform.
01 · occupation-db
Occupation database
~250 occupations from O*NET aligned to task-level scoring.
02 · task-scorer
Task-level scorer
Each occupation broken into ~20 tasks; LLM-era automatability per task.
03 · wage-tracker
Wage / posting tracker
Monthly wage and job-posting volume by BLS / LinkedIn signals.
04 · moat-finder
Moat finder
Tasks that resist automation: physical, social-trust, regulatory.
05 · transition
Transition advisor
If your job is heavily exposed, what adjacent roles compound your skills?
Data model
Occupation
field
type
note
id
soc-code
Standard Occupational Classification code
tasks
Task[]
Constituent tasks (O*NET)
automate_pct
0..1
Share of tasks LLM-era AI can already do
augment_pct
0..1
Share AI meaningfully accelerates
moat_pct
0..1
Share that resists automation