Jobs That Share a Fate
We computed the task composition of 0 occupations and connected those with similar profiles. Jobs in the same cluster face the same AI displacement pattern — even if they look nothing alike on paper.
Key finding: Jobs cluster into distinct "fate groups" based on their mix of routine cognitive, analytical, interpersonal, and manual tasks. A bookkeeper and a data entry clerk look like different jobs — but they're 90%+ routine cognitive tasks, so when AI automates that category, both roles are disrupted simultaneously.
Methodology
Each occupation's tasks are categorized into five types: routine cognitive, non-routine analytical, non-routine interpersonal, routine manual, and non-routine manual. The percentage breakdown creates a 5-dimensional "composition vector."
Edges connect occupations with cosine similarity ≥ 92% in their composition vectors. Node size reflects media mentions from our news pipeline. Color reflects average task-level AI risk score. Force-directed layout via d3-force.