Your AI economic outlook
Adrian Douglas’s Caribbean extension of the ILO occupation catalogue
See how much time AI could save in your job and what that could mean for your pay.

Where do you work?
Which industry do you work in?
What is your profession?
How long have you worked in your profession?
How much of your working day is hands-on?
Think about physical care, repairs, building, driving or handling goods.
How do you work?
Choose the arrangement that covers most of your work.
Your result
Who does each task?
Five-year risk matrix
What affects your result
Pay, prices and work value
View the pay and price comparison by year
Compare professions
Browse all occupations
Job loss and work value over five years
View yearly figures
Recommendations for your job
Find AI examples and prompts for your job ↗
Request a job title →How the estimate works · sources and limits
Occupations and tasks
Adrian Douglas’s Caribbean extension starts with the ILO–NASK 2025 exposure index. Caribbean titles map to related ISCO groups. The catalogue adds role examples, common clients and the industries where each job can be practised. The examples describe task ownership. Each task counts once in the diagram. ILO task data retain their CC BY 4.0 licence. The curated bamboo rafting roles draw on the Jamaica Tourist Board guide.
How the model ranks roles
The model combines the ILO group mean, variation and average amount by which task scores exceed the midpoint with curated digital, on-site and relationship factors. It reduces exposure as hands-on work increases and uses the selected industry’s task mix to estimate substitution and time savings. An experience multiplier of 1 ÷ (1 + 0.026 × years) reduces displacement as experience rises. Experience ranges use representative years of 1, 4, 8, 13, 18, 25 and 35. Formal employees are the reference case. The model assigns lower direct replacement factors to informal employees and self-employed workers and changes how much of the AI work gain they retain. The ILO’s Caribbean work on informality informs the categories. The multipliers are the author’s inputs.
Country, growth and prices
Country adoption uses internet access and sector data from the existing CAIRME model, with updated World Bank and ITU internet data where available. The industry demand factor uses the country’s median 2027–2031 real GDP growth and the average shortfall below 2% in earlier years (2017–2025, excluding 2020–2021). Each industry has its own response to this demand measure. The model also applies a bounded factor for inflation relative to the regional median. Growth and inflation come from the IMF October 2026 World Economic Outlook where available, otherwise the model uses newer World Bank historic GDP growth and historic inflation, then the regional median if local inputs are missing. GDP and inflation reflect economy-wide conditions.
Comparison and purchasing power
The table includes the full catalogue. It uses your industry for professions that fit it and each other role’s main industry for the rest. Country, experience, work setting and hands-on answers are held fixed in the comparison. Roles with equivalent inputs can share a score. Rounded figures may also tie.
Risk categories use the same cutoffs for both periods. Very low is below 1%. Low starts at 1%, moderate at 5%, high at 10%, and very high at 20%. The matrix uses five-year risk and groups task exposure at one-third and two-thirds. The estimate assumes a constant annual rate within the five-year period.
Work value combines the applicable job and industry task mix with country adoption and demand. The pay calculator spreads the five-year raise in equal steps against starting pay. It compounds annual IMF inflation rates, the available local or regional inflation baseline, or the rate you enter. It divides future pay by the price index. Your optional salary stays in your browser.
Model limits. Use these figures to compare scenarios. They cannot predict individual job loss, a pay rise or new jobs. Task examples and scores reflect technical capability and the author’s assumptions about adoption, industry demand, work arrangements and experience. For selected on-site trades, the model keeps five-year displacement below 0.5% and adds 2.5 to 5 points to five-year pay based on hands-on work. For self-employed people, displacement refers to losing paid work. Informal work can still lose customers and income. Economic figures measure work in index units before costs. The pay calculator holds exchange rates constant. Keep these estimates out of individual staffing decisions and seek qualified advice for career decisions.
