Pundi X and who benefits from AI data

Vitalik Buterin once explained blockchain to David Ben Kay by sketching it on a napkin in a restaurant in Suzhou, China. The simplicity of that moment underscored a principle Kay continues to value: technology should be clear, not obscure.

He now leads legal and strategic development at Pundi X and Pundi AI FX, where data and blockchain are being combined into new models of digital value.

Kay’s route into the tech world began with studying Mandarin, not programming. “I took a night course in Chinese during my junior year in high school,” he told The Crypto Radio, “because it seemed like something impossible to ever master in a lifetime.”

That curiosity led him through Brown University, UC Berkeley, and eventually law school. His Mandarin fluency placed him in China at the moment when global business, intellectual property, and local innovation were rapidly accelerating.

During this period, he helped draft some of the country’s early intellectual property laws and later joined Microsoft China as General Counsel. His work there intersected with government relations, anti-piracy efforts, and the emerging digital economy. But his career has always extended beyond legal frameworks. In Beijing, he converted a former factory into a digital arts and startup incubator, bringing together designers, artists, engineers, and founders long before such collaboration spaces became common.

“Among those meetings was one with Ming Chen, who became the first Managing Director of the Ethereum Foundation,” he said.

From crypto payments to data ownership

When Kay joined Pundi X in 2017, the mission was to make cryptocurrency simple and usable for everyday people. The team built retail payment systems that allowed merchants to accept crypto like they would a credit card. But as global compliance demands grew, scaling crypto payments across varied jurisdictions became increasingly complex.

The question became: what should be democratized next? “So in looking around, I asked 'What's the next thing that's of value that we want people to have access to?' And that was data,” Kay said. 

He frames the shift in a way many in the tech world will recognize but few follow through on. “We all talk about how data is the new oil. But the thing is that just like oil, it would have a tendency to be of little value when it's raw. It gets its value when it's refined.”

“What an AI asset is, is it's taking data and turning it into something that is an asset, something of value to the individual,” Kay said. His company enables people to contribute data, tag it, verify it, and transform it into refined datasets. Those can then be tokenized so contributors can share in the value they help create.

Kay often frames the principle simply: data is valuable – but the people who generate it rarely receive anything for it.

Why verified data matters for AI

Much of the data used to train AI today is produced in what Kay refers to as “data sweatshops,” where large groups of workers label information with little recognition or fair compensation. This also makes it difficult to verify the quality or origin of the data that ends up training large models.

Pundi AI FX aims to address both issues by enabling communities of subject-matter experts to refine and verify datasets collaboratively. “If you're building [machine learning models] that are concerned about medical issues, you'd like to have doctors being able to contribute to that, to be able to verify the data, to be able to tag it in a way that's going to be useful to other doctors,” he said. This approach improves the reliability of AI systems and ensures that those who improve the dataset are recognized and compensated.

Blockchain is the mechanism that makes this compensation and verification traceable. It records who contributed, tagged, or verified specific data, and how that data changed over time. This establishes provenance, making the dataset auditable and enabling contributors to receive fair value for their work.

“There’s a real synergy between them,” Kay said of AI and blockchain. “The end result becomes significantly stronger.”

The stakes and the human element

Concerns that AI will replace workers continue to grow, and Kay does not dismiss those anxieties. “It's a legitimate concern,” he acknowledged, emphasizing that the direction of development – who benefits and how – is shaped by design choices, not technological inevitability.

His guidance for young people entering the field follows the same principle: agency matters.

“I think the the most important skill set is flexibility – not having a fixed idea of where things are going, but to be able to bring in various skills along the way,” he said. Rather than planning a rigid career path, he encourages staying open to new directions and choosing work that feels meaningful.

He recalls the guidance his father gave him: “You want to be able to wake up in the morning and be excited about your job, and if you ever get to a point where you're not excited about it, then change jobs.”

Kay avoids singling out one guiding figure. “I always believe that everyone you meet has that potential of being a teacher for you,” he said. The conversation in Suzhou, where Buterin sketched blockchain on a napkin, was simply one example of that.