©2026 Business Next Media Corp. All Rights Reserved. No.102, Guangfu S. Rd., Da'an Dist., Taipei City 106, Taiwan
Taiwanese Founder Tops 2025 Forbes Young Billionaires List, Surpassing Zuckerberg
Forbes released its 2025 Young Billionaires List in September, featuring familiar names like Airbnb co-founder Nathan Blecharczyk and Meta founder Mark Zuckerberg. Yet this year, the number one spot goes to a Taiwanese-American entrepreneur who may not be as widely known: Edwin Chen, the founder of data startup Surge AI.
At just 37 years old, Edwin Chen has built a fortune of around $18 billion, making him not only the youngest person on the 2025 Forbes 400 list of America’s richest, but also the wealthiest newcomer on this year’s ranking.
So who is Edwin Chen, and how did he rise to the top?
A Math Prodigy From Small-Town Florida
Edwin Chen grew up in Crystal River, a small Florida town of just 3,400 people, famous more for its manatees and retirees than for any connection to the tech world. His parents, immigrants from Taiwan, ran a Chinese-Thai restaurant called Peking Garden. Like many children of immigrant families, Edwin helped out at the restaurant from a young age, but his real passion was always mathematics and language.
According to Forbes, Edwin was a true math enthusiast who showed remarkable talent for both language structures and mathematical patterns even as a child. Thanks to a full scholarship, he attended a prestigious preparatory school before enrolling at the Massachusetts Institute of Technology (MIT), where he majored in mathematics, computer science, and linguistics.
While at MIT, Edwin experimented with polyphasic sleep—sleeping just 30 minutes every six hours instead of getting a full eight hours at once. This drive for efficiency hinted at his later entrepreneurial style. To this day, he walks 20,000 steps daily and prefers to clear his mind by taking long nighttime walks through Times Square, drawing inspiration from the city’s neon lights and bustling crowds.
Edwin’s gifts in math and linguistics led him to algorithm and content moderation roles at Silicon Valley giants like Google, Facebook, and Twitter (now X). It was here he encountered one of AI’s dirtiest secrets: the challenge of data quality.
Despite their massive data resources, these companies struggled to find the high-quality datasets needed to train new AI models. “It became the biggest bottleneck in development,” Edwin recalls. “If building even a simple classifier gets stuck because of poor data, how are we supposed to realize the dream of advanced AI agents?” He likens the situation to sending a child prodigy to school with textbooks full of errors—the potential never gets realized.
In 2020, Edwin Chen decided to leave Twitter and used his savings to found Surge AI.
What Is Surge AI? No Outside Funding and $1.2 Billion in Annual Revenue
So what does Surge AI actually do? In short, they act as private tutors for artificial intelligence.
Training large language models like ChatGPT requires vast amounts of data, but at the time, most data labeling was done through what Edwin describes as “body shops.” These companies relied on huge numbers of low-paid workers to complete simple tasks like “Is this a cat or a dog?” or “Draw a box around the car in this photo.”
While using low-cost labor is sufficient for basic image recognition, it falls far short when it comes to training generative AI. Edwin Chen explains, “If all you need is to draw a box around a car, the results from Ernest Hemingway and an elementary school student would look exactly the same. But in the era of generative AI, if you want a model to write poetry, the value of Hemingway is irreplaceable.” If people who know nothing about poetry are asked to annotate poetry data, they might only check for things like rhyme or line count, which means the resulting AI could only write at a high school level.
Surge AI’s core advantage is the quality of its human feedback. Rather than hiring low-paid temps, the company recruits experts with advanced degrees, legal backgrounds, or training in linguistics. These specialists teach AI to write code, compose poetry, and even solve complex math problems.
According to Forbes, Surge AI brought in over $1.2 billion in revenue in 2024 and has been profitable since its early days. Its clients include industry giants like Google, OpenAI, and Anthropic.
Remarkably, Surge AI has always operated as a bootstrapped company, never seeking outside investment. “A lot of people raise money just so they can send out a press release for their parents or show off to friends that they’ve landed ten million dollars,” says Chen. In his view, taking outside capital too early often leads to bloated teams and strategies that lose focus.
The Clickbait Crisis in AI
Surge AI’s rise brings to mind Scale AI, another well-known data labeling firm, which earlier this year sold 49 percent of its shares to Mark Zuckerberg, another member of the billionaire list.
Chen openly admits that there has been plenty of back-and-forth between the two companies. He criticizes Scale AI as a “body shop,” while Scale AI has fired back, claiming Surge AI’s barbs come from fundraising frustration or sour grapes. Yet both companies face similar legal troubles, including a class-action lawsuit in California over the misclassification of full-time workers as contractors.
For Edwin Chen, competition in the business world is less concerning than the direction the entire AI industry is heading.
He points out that many AI models today are optimized for “clickbait” metrics, aiming to top leaderboards rather than improve real accuracy. These models generate longer answers and add more emojis to win user approval, but often sacrifice factual correctness. For example, some developers try to boost scores on a benchmark called “IF Eval” by training models to capitalize five specific letters in “Abraham Lincoln,” even though this serves no real-world purpose and wastes valuable compute resources.
The full name of “IF Eval” is Instruction Following Evaluation. Its original intent is sound: to test whether AI can understand and carry out complex instructions. However, in practice, the requirements can become overly artificial, such as insisting on random capitalization. As Chen explains, “No real user would ever say, ‘Write my report, but make the name jump between uppercase and lowercase letters.’ Succeeding at this makes a model look impressive, but offers no real value.”
What matters most for the future, Chen believes, is ensuring that generative AI incorporates genuine human richness and emotion. The development of AI should always return to the golden rule: does it truly provide value for people? Only by relying on rigorous, expert human review can we prevent AI from becoming nothing more than a convincing but empty talker.
©2026 Business Next Media Corp. All Rights Reserved. No.102, Guangfu S. Rd., Da'an Dist., Taipei City 106, Taiwan