Brown student maps the hidden global labor powering AI

What it claims

Cunyan Ma, a Brown University senior, built an interactive map that links technology companies and their intermediaries to data workers around the world. The map is freely available online. Ma built it with Brown's Center for Technological Responsibility, Reimagination and Redesign as part of an ongoing project on the material underpinnings of the digital world. According to Ma, most knowledge of data workers comes from anecdotal reports by authors and journalists. The map is meant to show a larger picture of what Ma calls "AI colonialism."

How it works

The map draws on three sources:

• existing data worker databases
• Ma's own data collection
• news reports on the working conditions data workers face

The page does not name the databases or describe how Ma collected data.

The article describes the kinds of work the map is about:

• labeling objects in photos or parts of speech in sentences
• transcribing long passages of speech
• rating good or bad chatbot responses
• marking tumor boundaries in medical scans
• labeling disturbing images and text so AI systems can filter that content or avoid producing it, which requires workers to view it

The article says much of this labor is low-wage work in the Global South.

Limits

The fetched text ends partway through the article. It gives no link to the map. It also gives no count of the companies, intermediaries, workers or countries covered, and no description of the data schema. There is no date range for the sources and no licence or terms for reusing the data. Anyone planning to cite or build on the map would need to open it directly and check its coverage and sourcing.