“…“My job is gone because of ChatGPT, and I was being invited to train the model to do the worst version of it imaginable,” she says. The idea depressed her. But her financial situation was increasingly dire, and she had to find a new place to live in a hurry, so she turned on her webcam and said “hello” to Melvin.
It was a strange, if largely pleasant, experience. Manifesting on Katya’s laptop as a disembodied male voice, Melvin seemed to have actually read her résumé and asked specific questions about it. A few weeks later, Katya, who like most workers in this story asked to use a pseudonym out of fear of retaliation, received an email from Mercor offering her a job. If she accepted, she should sign the contract, submit to a background check, and install monitoring software onto her computer. She signed immediately.
She was added to a Slack channel, where it was clear she was entering a project already underway. Hundreds of people were busy writing examples of prompts someone might ask a chatbot, writing the chatbot’s ideal response to those prompts, then creating a detailed checklist of criteria that defined that ideal response. Each task took several hours to complete before the data was sent to workers stationed somewhere down the digital assembly line for further review. Katya wasn’t told whose AI she was training — managers referred to it only as “the client” — or what purpose the project served. But she enjoyed the work. She was having fun playing with the models, and the pay was very good. “It was like having a real job,” she says.
Two days after Katya started, the project was abruptly paused. A few days after that, a supervisor popped into the room to let everyone know it had been canceled.
[…]
Early labor skirmishes are already happening, mostly in California, which has some of the most aggressive rules around classifying platform workers. Three class-action lawsuits have been filed against Mercor in the past six months. (Similar suits were previously filed against Surge AI and Scale AI, which is settling.) The lawsuits all accuse the companies of misclassifying workers as independent contractors given the “extraordinary control” they exert over them. This is “an entirely new kind of work,” one that the company trains people to do and that cannot be done except on the company’s platform. Workers have so little visibility into what they’re working on that one person, alleges a suit filed in December, accepted a Mercor project only to be tasked with recording himself reading sexually explicit scripts. Once he discovered this, the worker risked deactivation if he abandoned the project, forcing him to “choose between being paid and being humiliated.”
These companies are reminiscent of Uber and Lyft a decade ago, says Glenn Danas, a partner at the law firm Clarkson, which is suing Mercor and several other data platforms. Yet in some ways these workers are in a worse position, more replaceable despite their advanced degrees. Uber drivers have to be physically present in a city to work, and they can organize and push for regulation there. If the same were to happen with data workers, companies could just recruit from somewhere else where people will work for less. When Mercor cut pay for its Meta project to $16 per hour, it dropped below the minimum wage in California and other states, yet people there kept working because they needed the money. This was something at least one supervisor acknowledged, writing in Slack, “While we won’t actively hire from any states where the minimum wage is above the project’s rate, if you are already active on the project and would like to work at the $16/hr rate, we want to enable you to do so.”
Entire professions risk a similar race to the bottom, says Acemoglu, if companies are able to pit workers against one another, each selling their data before someone else can underbid them. “We may also need unionlike organizations that exercise some sort of collective ownership and prevent any kind of simple divide-and-rule strategies by large companies to drive down data prices,” he says. “If there isn’t the legal infrastructure for a data economy of this sort, many of the people who produce the data will be underpaid or, to use a more loaded term, exploited.”…”