Alexej Savreux, a 34-year-old man from Kansas City, says he has done all kinds of work over the years. He has made fast food sandwiches. He has been a custodian and transporter of scrap metal. And he has done technical sound work for live theater.

These days, however, his job is less hands-on: he’s an artificial intelligence trainer.

Savreux is part of a hidden army of contract workers who have been doing the behind-the-scenes work of teaching AI systems how to analyze data so they can generate the kinds of text and images that have captivated people using newly popular products. like ChatGPT. . To improve the accuracy of the AI, she has tagged photos and made predictions about what text apps should generate next.

Pay: $15 an hour and up, no benefits.

Out of the spotlight, Savreux and other contractors have spent countless hours over the past few years teaching OpenAI systems to be more responsive in ChatGPT. Their feedback fills a pressing and never-ending need for the company and its AI competitors: to provide sentence streams, labels, and other information that serves as training data.

“We are seasoned workers, but there would be no AI language systems without it,” said Savreux, who has worked for technology startups including OpenAI, the San Francisco company that launched ChatGPT in November and sparked a wave of enthusiasm around to generative AI.

“You can design all the neural networks you want, you can involve all the researchers you want, but without taggers, you don’t have ChatGPT. You have nothing,» Savreux said.

It’s not a job that will bring Savreux fame or wealth, but it is essential and often overlooked in the field of AI, where the seeming magic of a new technological frontier can dwarf the work of contract workers.

“A lot of the talk about AI is very congratulatory,” said Sonam Jindal, program leader for AI, jobs and the economy at the Partnership on AI, a San Francisco-based nonprofit that promotes research and education. around artificial intelligence.

“But we’re missing a big part of the story — that this is still hugely dependent on a large human workforce,” he said.

The tech industry has relied for decades on the labor of thousands of less-skilled and lower-paid workers to build its computing empires: from punch card operators to in the 1950s to the most recent Google contractors who have complained second class status, including yellow badges that distinguish them from full-time employees. Online work through sites like Amazon Mechanical Turk became even more popular early in the pandemic.

Now the burgeoning AI industry is following a similar playbook.

Work is defined by its unstable and in-demand nature, with people employed by contracts written directly by a company or through a third-party provider that specializes in temporary work or subcontracting. Benefits like health insurance are rare or non-existent, which translates to lower costs for tech companies, and the work is often anonymous, with all the credit going to executives and researchers at tech startups.

The Association on AI warned in a 2021 report that a surge in demand for what he called “data enrichment work” was coming. He recommended that the industry commit to fair compensation and other improved practices, and last year he published voluntary recommendations. guidelines for companies to follow.

DeepMind, an artificial intelligence subsidiary of Google, is so far the only technology company commit publicly to those guidelines.

“Many people have recognized that it is important to do this. The challenge now is to get companies to do it,” Jindal said.

“This is new work that AI is creating,” he added. “We have the potential for this to be high-quality work and for the workers who do this work to be respected and valued for their contributions in enabling this progress.”

A spike in demand has arrived, and some AI contract workers are asking for more. In Nairobi, Kenya, more than 150 people who have worked on AI for Facebook, TikTok and ChatGPT voted Monday to form a union, citing low pay and the mental drain of the job. Time magazine reported. Facebook and TikTok did not immediately respond to requests for comment on the vote. OpenAI declined to comment.

So far, AI work-for-hire has not inspired a similar movement in the US among Americans quietly building AI systems word for word.

Savreux, who works from home with a laptop, got on board with AI hiring after seeing a job posting online. He credits AI’s work, along with previous work at the Jimmy John’s sandwich chain, for helping him out of homelessness.

“People sometimes downplay these necessary and laborious jobs,” he said. «It’s the necessary entry-level area of ​​machine learning.» The $15 an hour is more than the minimum wage in Kansas City.

Job postings for AI contractors are about both the allure of working in a cutting-edge industry and the sometimes grueling nature of the job. A advertisement from Invisible Technologies, a temp agency, for an «Advanced Artificial Intelligence Data Trainer» notes that the job would be entry-level with pay starting at $15 an hour, but also that it could be «beneficial to the humanity».

“Think of it like being a language arts teacher or a personal tutor on some of the world’s most influential technology,” the job ad says. He does not name the Invisible client, but says the new hire would work «within protocols developed by the world’s leading artificial intelligence researchers.» Invisible did not immediately respond to a request for more information about its listings.

There is no definitive count of how many contractors work for AI companies, but it is an increasingly common way of working around the world. Time magazine reported in January that OpenAI relied on low-wage Kenyan workers to tag text that included hate speech or sexually abusive language so their apps could better recognize toxic content themselves.

OpenAI has hired some 1,000 remote contractors in places like Eastern Europe and Latin America to label data or train company software on computer engineering tasks, the online news outlet Semafor reported in January.

OpenAI is still a small company, with about 375 employees as of January, CEO Sam Altman said on Twitter, but that number does not include contractors and does not reflect the full scale of the operation or its ambitions. An OpenAI spokesperson said no one was available to answer questions about the use of AI contractors.

The job of creating data to train AI models is not always easy to do and is sometimes complex enough to attract potential AI entrepreneurs.

Jatin Kumar, a 22-year-old from Austin, Texas, said he’s been doing contract AI work for a year since graduating college with a degree in computer science, and said it gives him an idea. where is the generative artificial intelligence technology from. is directed in the short term.

“What it allows you to do is start thinking about ways to use this technology before it hits public markets,” Kumar said. He is also working on his own tech startup, Bonsai, which is creating software to help with hospital billing.

As a conversational coach, Kumar said his main job has been generating prompts — engaging in a back-and-forth conversation with chatbot technology that is part of the long process of training AI systems. The tasks have become more complex with experience, he said, but they started out very simple.

“Every 45 or 30 minutes, I would get a new task, generating new prompts,” he said. Prompts can be as simple as «What is the capital of France?» he said.

Kumar said he worked with 100 other contractors on tasks to generate training data, correct answers, and refine the model by providing feedback on the answers.

He said other workers handled «flagged» conversations: they read examples submitted by ChatGPT users who, for one reason or another, reported the chatbot’s response to the company for review. When a flagged conversation comes in, he said, it’s sorted according to the type of error involved and then used in further training of the AI ​​models.

“Initially, it started as a way to help out with OpenAI and learn about existing technologies,” Kumar said. «But right now, I can’t see myself walking away from this role.»