It may be easy to ignore the work that companies that create data labels do. Edwin Chen, CEO of Surge AI, said it could be due to a misunderstanding of their activities.
“When most people think of data labeling, they think of something as simple as labeling a photo of a cat or drawing boundary marks around a car,” Chen told Lenny Rachitzky on the Lenny Podcast.
“I've always hated the term data label,” said Chen, who previously worked at Google, Twitter and Meta.
“Because if you think what we're doing is completely different, it's just painting a very simple picture,” he said.
Surge AI, which Chen founded in 2020, competes with companies such as Scale AI and Mercor in the AI data labeling space. Surge has also partnered with Anthropic, which also operates DataAnnotation.tech, where freelancers can sign up and receive compensation for training AI models. These remote workers are sometimes referred to as “ghost workers” because they perform the behind-the-scenes labor essential to the development of AI.
Chen said that labeling data is much more than a mechanical task, it is a creative endeavor. He likened what companies like Surge are doing to how parents instill lifelong values in their children.
“I think about it a lot because what we do is like raising a child,” he says. “We're not just giving our kids information; we're teaching them values, creativity, infinite subtleties about what is beautiful and what makes a person a good person.”
Surge AI website Surge AI
In this way, companies like Surge AI are “nurturing the children of humanity,” Chen said.
Mr. Chen's views can be viewed by visiting Surge's website. The website asks visitors the question, “What made Hemingway, Kahlo, and von Neumann so extraordinary?”
“Their life experiences: war, love, victory, loss. The people they met, the cities they explored, the thousand choices that shaped them,” the website says. “Data will do for AI what life did for them: transform data into intelligence that can one day prove the Riemann Hypothesis, imagine new philosophies, and send rockets to the stars.”
“You don't have to be someone you're not.”
Mr. Chen previously worked at Big Tech companies such as Twitter, Google, and Facebook. You may remember one of his Twitter projects.
Mr. Chen was at the company about 10 years before Elon Musk bought it, and became known for geo-tagging tweet data from around the United States to create a “pop vs. soda” map that showed users what words they used to refer to soft drinks.
Looking back on Surge AI's launch, Chen said he was surprised to realize that he never had to stop digging into data.
“I thought if I was going to start a company, I'd have to be a businessman who looked at finances all day long, sat in meetings all day long, and did all these things that sounded incredibly boring. And I always hated it,” he told Rachitsky. “So I think it's funny that it ended up not being true at all.”
Chen said he wished he had known he didn't have to “constantly tweet, advertise, and fundraise.”
“You don't have to be someone you're not,” he said. “In fact, you can build a successful company just by building something so good that it cuts through all the noise. And if I had known this was possible, I think I would have started sooner.”
Do you work in data labeling? Contact reporters from non-work emails and devices. bgriffiths@businessinsider.com.
