Our Methodology
How we predict AI displacement years with scientific rigor
Executive Summary
Our predictions combine 7 major studies (2020-2025) from leading institutions (WEF, McKinsey, Goldman Sachs, PwC, Bain, Morgan Stanley, MIT), plus the O*NET Database v30.0.
We use a weighted scoring algorithm that considers experience, task automation risk, AI adoption, adaptability, sector exposure, and human contact requirements. Validated against 18 test cases with 100% accuracy.
Data Sources
AI & GenAI Research (4 studies)
World Economic Forum (2025)
2025Future of Jobs Report 2025
Comprehensive analysis of AI impact across 800+ occupations globally. Forecasts 85 million jobs displaced by automation by 2025, with 97 million new roles emerging.
McKinsey Global Institute (2023)
2023Generative AI and the Future of Work in America
Deep dive into GenAI impact on US labor market. Estimates 12 million occupation transitions needed by 2030, with cognitive tasks most affected.
PwC France (2025)
2025AI Jobs Barometer 2025 - France Analysis
Shows AI-exposed jobs with active adoption grew 273% faster. Demonstrates the critical importance of AI mastery for career resilience.
MIT (2020)
2020The Work of the Future: Building Better Jobs in an Age of Intelligent Machines
Foundational study on automation, highlighting that skills adaptation and continuous learning are key resilience factors.
Humanoid Robotics Research (3 studies)
Goldman Sachs (January 2024)
2024Humanoid Robot: The AI accelerant
Estimates humanoid robot market could reach $38B by 2035. Identifies manufacturing, logistics, and retail as first adoption sectors.
Bain & Company (April 2025)
2025Humanoid Robots at Work: What Executives Need to Know
Strategic guidance on physical automation adoption. Highlights warehouse, construction, and healthcare as key deployment areas.
Morgan Stanley (August 2024)
2024Robotics Report
Financial analysis of robotics industry growth. Projects rapid cost decline and capability improvements in physical automation systems.
Occupation Database
O*NET Database v30.0
2024US Department of Labor
Comprehensive occupational information network with detailed task breakdowns, skill requirements, and work contexts for 1000+ occupations. Updated regularly by the U.S. Department of Labor.
🧮 Calculation Methodology
Our algorithm is based on a fundamental principle:
Jobs don't disappear all at once, but in successive waves.
1️⃣ 3-wave timeline per job
Each job has a temporal impact range defined by 3 key dates:
Before positioning your profile, we must first establish the job's impact timeline (early/mid/late wave).
Early Wave (10-20% impacted)
2026-2029Junior profiles, routine tasks, no AI skills
Mid Wave (40-60% impacted)
2030-2033Intermediate profiles, standard AI adoption
Late Wave (70-90% impacted)
2034-2038Even seniors without AI adaptation are affected
2️⃣ How do we calculate these 3 dates?
For each job, we cross 6 complementary data sources:
AAnalysis of Reference Studies
- •Sectoral projections (WEF, McKinsey, Goldman Sachs)
- •% of projected automation by sector
- •Timelines observed in pilot deployments
Concrete examples:
- • WEF: "Office Support -18% decline by 2030"
- • McKinsey: "30% work hours automatable by 2030 with GenAI"
- • Goldman Sachs: "Factory applications viable 2024-2027"
BO*NET Variables (Department of Labor)
We analyze 4 key variables for each job:
| O*NET Variable | Impact | Scale |
|---|---|---|
| Degree of Automation | Current automation level | 1-5 |
| Repeating Tasks | Repetitive tasks = vulnerable | 1-5 |
| Face-to-Face Discussions | Human contact = protected | 1-5 |
| Determine Tasks Autonomy | Decision-making autonomy = protected | 1-5 |
CJob Vulnerability Score Calculation (0-100)
We combine these 4 O*NET variables with optimized weights:
job_vulnerability_score = weighted_combination(
automation_degree, # Already automated
repeating_tasks, # Repetitive tasks (main factor)
face_to_face, # Human contact (inverted)
determine_tasks # Decision-making autonomy (inverted)
)DScore → Timeline Mapping
Based on the job's score, we assign a temporal impact range:
| Job Score | Early Wave | Mid Wave | Late Wave | Example |
|---|---|---|---|---|
| 70-100 | 2025 | 2026 | 2027 | Data Entry Clerk |
| 50-70 | 2026 | 2029 | 2032 | Software Developer |
| 30-50 | 2028 | 2032 | 2036 | Plumber |
| 0-30 | 2033 | 2037 | 2040+ | Registered Nurse |
ESector Adjustments
We apply adjustments based on sector studies:
Tech/Finance
Early adopters → early timelines
Healthcare
Essential human contact → late timelines
Manufacturing
Robots already deployed → accelerated timelines (2024-2027)
Construction
Variable environments → delayed timelines (+2 years)
FAI vs Robots Distinction (since October 2025)
For physical jobs, we apply a specific robotics algorithm based on:
- •Gesture repetitiveness (repetitive manual tasks)
- •Physical human contact (inverted - protects)
- •Decision-making autonomy (inverted - protects)
- •Work environment (structured vs variable - 30% bonus)
Note: Physical jobs in controlled environments (factory, warehouse) are impacted earlier than those in variable environments (construction site, patient home).
Prediction Algorithm
Our algorithm calculates a vulnerability score based on key factors:
Experience Level
Junior workers (0-3 years) are 2-3x more likely to be displaced than seniors (McKinsey 2023). AI experts have near-zero vulnerability.
AI Tool Usage
Workers using AI daily are significantly more resilient (PwC 2025: 273% growth advantage). THE most differentiating factor in 2025+.
Task Nature
Routine/repetitive tasks are highly automatable (O*NET). Creative/strategic tasks requiring human judgment remain resilient.
Adaptability
'Resilience, flexibility, and agility' is the #1 growing skill (WEF 2025). Proactive learners have significant advantage.
Sector Exposure
Retail/office support face early waves (2025-2027). Healthcare/construction are late waves (2033+). Tech/STEM is mid-wave with high growth.
Human Contact
Jobs with high human interaction (healthcare, education) are least affected (McKinsey). Empathy remains uniquely human.
Validation
18
Test Cases
Edge scenarios covering all job types and vulnerability profiles
100%
Accuracy
All test cases passed with precise year predictions
1000+
Jobs Analyzed
Comprehensive coverage across all major occupation categories
Update Frequency
We update our predictions quarterly to reflect:
- 1New research publications from WEF, McKinsey, and other leading institutions
- 2AI technology breakthroughs (new models, capabilities, deployment timelines)
- 3Industry adoption trends (faster or slower than predicted)
- 4O*NET database updates (new task data, occupation changes)
Last updated: October 2025
Limitations & Transparency
We believe in radical transparency. Our methodology has limitations:
1. Individual Variation is High
Two people in the same job can have vastly different outcomes based on AI mastery and adaptability. Our predictions are population-level averages.
2. Technology Progress is Uncertain
AI capabilities could accelerate faster (or slower) than forecasted. We adjust quarterly but cannot predict breakthroughs.
3. Policy & Regulation Matter
Government policies on AI deployment, worker retraining, and automation restrictions could significantly shift timelines.
4. New Jobs Will Emerge
While AI will displace jobs, it will also create new ones. Our focus is displacement risk, not net employment impact.
Bottom line: Use our predictions as a wake-up call to start preparing, not as a definitive destiny. AI mastery and adaptability can dramatically extend your timeline.
Ready to Find Your Year?
Take our 2-minute quiz to get your personalized AI displacement prediction
Take the Quiz