Internet Pioneer David Clark’s Lessons for the AI Age

What can the person who helped build the Internet teach us about building AI?

We're launching the first full episode of Building 32 with someone uniquely qualified to put today's AI moment into perspective. Host Karen Given sits down with MIT CSAIL Senior Research Scientist David Clark to revisit the early days of the Internet, when a handful of researchers were trying to solve what seemed like an impossible problem: getting computers to talk to one another. Looking back, Clark says the hardest challenges weren't technical after all. They were human.

As AI reshapes how we work, communicate, and make decisions, Clark argues that many of the biggest questions aren't about the technology itself. They're about power, incentives, regulation, and the ways people use technology, for better and for worse. Drawing on everything from the collaborative "bake-offs" that shaped the internet's earliest protocols to today's debates over AI governance, he reflects on why the Internet's biggest challenges turned out to be social rather than technical, and what that means for AI today.

This conversation explores:

• What it was really like helping invent the Internet
• Why the Internet succeeded by embracing "rough consensus and running code"
• The unintended consequences of social media and centralized platforms
• What AI can learn from the Internet's biggest mistakes
• Why David Clark remains optimistic, but a much more cautious observer of human behavior

"I'm still a techno optimist, but I'm much older and wiser... The problems we're dealing with are not technical problems. They are problems of society." - David Clark