How the Low-Key Billionaire Behind Surge Is Beating Out Rivals Like Scale AI (www.forbes.com)

Edwin Chen, a former Google/Facebook/Twitter engineer, has quietly built Surge AI into the dominant high-end data-labeling firm powering major LLMs. Surge reported $1.2B in revenue in 2024, counts Google, Meta, Microsoft, Anthropic and Mistral as customers (helping train Gemini and Claude), and is privately valued at roughly $24B with talks to raise $1B at a $30B valuation. Chen bootstrapped the company, retains ~75% ownership (making him a new Forbes 400 entrant), runs a lean core team of ~250 employees plus a global gig workforce of over a million, and insists on profitability and selective hiring rather than rapid, VC-driven scale. Technically, Surge differentiates by combining domain experts and vetted gig workers to perform adversarial, context-sensitive annotation: prompt chatbots to fail, craft improved responses, compare alternatives, and write scoring criteria that translate nuanced human judgments into labels for training and evaluation. That approach targets higher-quality signal vs. low-cost clickthrough annotation; Surge charges substantially more than legacy providers and claims its data is essential for robust model behavior and alignment. The rise underscores that data—like compute and energy—is now strategic infrastructure for AI, likely accelerating consolidation, higher-margin specialized labeling services, and renewed focus on annotation quality as models scale. Chen is also emerging as a thought leader pushing the industry to optimize for better human-centered objectives, not clickbait-style metrics.