How One AI Startup Quietly Became Worth More Than Ford

It sounds ridiculous on paper. Ford Motor Company, which has been putting steel on American roads for more than a century, is now worth less than a five-year-old AI startup that primarily operates out of Manhattan and employs people all over the world.

Edwin Chen founded Surge AI in 2020, and it is reportedly valued in the private markets at $24 billion to $30 billion. Depending on the week, Ford’s public valuation has consistently lagged well below that. One constructs trucks. The other creates training data.

It almost seems unfair to compare them. The metallic sound of automated presses, forklifts humming on concrete floors, and workers adjusting torque wrenches can be heard outside Ford factories in Michigan. There isn’t any physical drama in Surge. Its “factory” is more subdued, with engineers honing prompts at midnight, gig workers stress-testing chatbots, and professors going over datasets.

The market might not be comparing products at all. Futures are being compared.

Unlike Anthropic or OpenAI, Surge does not create AI models. Rather, it provides the high-quality human-labeled data that those models rely on. According to reports, the company made about $1.2 billion in 2024 alone, supporting train systems like Anthropic’s Claude and Google’s Gemini. It’s difficult to overlook that scale, accomplished in less than five years.

It seems as though investors have subtly determined that this decade’s more strategic asset is data, not automobiles.

The conventional mythology surrounding venture capital does not apply to Chen’s story. Instead of using the typical Silicon Valley fundraising gimmicks, he used his savings from his time at Twitter, Google, and Facebook to bootstrap Surge. Chen was improving data workflows and employing annotators who could pick up on subtleties in scientific jargon or political speech, while other AI founders were pitching big ideas on Sand Hill Road.

The contrast is difficult to ignore. Ford spends billions retooling factories for electric vehicles while managing supply chains and negotiating union contracts. Surge employs about 250 core staff members in addition to a sizable network of remote contractors. One company requires a lot of capital. The other uses human intelligence and software.

AI infrastructure is expected to fetch higher multiples than industrial manufacturing, according to investors. The levers propelling the market’s imagination at the moment are computation, data, and model refinement. When a truck sells, money is made. As long as models require training, a data pipeline creates ongoing demand.

However, there is also skepticism.

The valuation of Surge is predicated on the AI boom continuing. It makes the assumption that businesses will continue to invest heavily in model improvement. It is predicated on the idea that improved data produces better models, which in turn spur greater enterprise adoption. Today, that loop feels solid. Whether margins will hold as competition heats up is still up in the air.

The cultural shift is another factor to take into account. Because they constructed observable structures like steel mills, automobiles, and railroads, industrial titans in the early 20th century became famous. Value creation in the modern era is invisible. The work done by Surge includes calibrating model responses, identifying toxic outputs, and improving prompts. Although it is necessary, it is not a good photographer.

As we observe this, it seems as though the concept of “industrial power” is being subtly redefined. Efficiency was once represented by the assembly line. Maybe it’s the annotation process now.

According to reports, Chen has kept a 75% stake in Surge. He is among the wealthiest self-made tech founders of his generation if the valuation is accurate. He has, however, maintained a low profile and frequently discusses literature and linguistics rather than quarterly growth.

That disposition might have been advantageous. During the AI hype cycle, Surge refrained from overhiring. Early on, it prioritized profitability. Surge concentrated on making chatbots a little less incorrect by fixing hallucinations, improving tone, and adding contextual accuracy while rivals rushed to make headlines.

The work is dull. Boring work has always been valuable.

Buried here is a lesson that seems almost archaic. New technology is frequently operationalized and deployed by businesses faster than its creators. Henry Ford scaled the automobile, not invented it. The graphical user interface was commercialized by Microsoft, not invented by them. Surge helped make generative AI usable, but it did not create it.

However, there are more serious issues when comparing it to Ford. Tens of thousands of people work for Ford. Whole regional economies are anchored by it. The distributed model of Surge represents a different labor paradigm that is more globally fragmented and less rooted in geography.

Can you sustain that? Data labor practices may come under increased regulatory scrutiny. AI models may become less dependent on human annotators as they become more capable of producing artificial training data. Investors are placing bets that those risks can still be controlled.

For the time being, the numbers are very clear. A 2020-founded business that specializes in data labeling is worth more than a 120-year-old automotive legend.

Standing in a Ford showroom, looking at a shiny F-150, and then looking at private-market valuations on a phone screen, it’s difficult not to feel a little shocked. Code versus steel. Datasets versus assembly lines.

The industrial age was created by one. The other is assisting in the construction of whatever follows.