We are all AI free data workers

Chandhiramowuli tells me about a case where a data annotator in India had to distinguish between images of soda bottles and pick out the ones that looked like Dr. Pepper. But Dr. Pepper is not a product that is sold in India, and the onus was on the data logger to figure that out.

The expectation is that the annotators understand the values ​​that are important to the company, says Chandhiramowuli. “They’re not just learning about these distant, distant things that are absolutely meaningless to them, they’re also trying to figure out not just what other contexts are, but what are the priorities of the system they’re building,” he says.

Indeed, we are all data workers for big tech companies, whether we are aware of it or not, say researchers from the University of California, Berkeley, the University of California, Davis, the University of Minnesota and Northwestern University in a new card presented to the FAccT.

AI models of text and images are trained using huge datasets that have been pulled from the internet. This includes our personal data and the copyrighted works of artists, and that data we created is now forever part of an AI model built for a business to make money. We unknowingly contribute our work for free by uploading our photos to public sites, upvoting comments on Reddit, tagging images on reCAPTCHA, or searching online.

Right now, the balance of power is heavily weighted in favor of some of the biggest tech companies in the world.

To change that, we need a data revolution and regulation. The researchers argue that one way people can take control of their online existence is by supporting transparency around how data is used and by finding ways to give people the right to offer feedback and share revenues from using their data. data.

While this data work forms the backbone of modern AI, data work remains chronically undervalued and invisible around the world, and wages remain low for annotators.

“There is absolutely no recognition of what the contribution of data work is,” says Chandhiramowuli.