AI Impact (AII) measures how AI might affect 17,879 different job tasks in the U.S. by comparing them with 24,758 AI patents
The Potential Impact of AI Innovations on U.S. Occupations, PNAS NEXUS 2024
A group of jobs in the economy that have similar features
We collected 20 job sectors from the O*NET database (version 26.3, released in May 2022), a widely used resource for occupational studies.
The title given to a job in a job sector
We found 749 different job titles from the O*NET database.
The number of tasks in a job impacted by AI patents
We collected 17,879 tasks, with each job having between 4 to 286 tasks (median of 20).
This measures how much AI could affect a job's tasks by comparing each task to AI patents
We calculated the AI impact by finding the most similar patent to each task using a similarity score.
The 5 tasks in a job most impacted by AI patents
We compared AI patents with job tasks to find which tasks have the highest similarity to the patents.
The AI patents that are most similar to the job tasks
We identified them by comparing the text of the patents with the descriptions of the tasks.
A different way to measure AI's effect on jobs, based on a study by Felten et al. (2021)
They linked AI technologies to job skills using a survey, and then calculated how much each job skill is exposed to AI.
Ali Akbar Septiandri, Marios Constantinides, Daniele Quercia
The Potential Impact of AI Innovations on U.S. Occupations, PNAS NEXUS 2024
Data Visualization: Edyta Bogucka, Marios Constantinides, Daniele Quercia
Code and Data: Fill out the form
Got feedback about the project?
