How one founder’s bet on ‘the old school web’ is paying off

There’s a quiet rebellion in Campbell’s choice, and it feels significant. In an industry drunk on scalability and disruption, his decision is a vote for curation and depth. The siren call of AI, with its rivers of venture capital and promises of world-altering scale, is nearly irresistible for a founder with his pedigree. To walk away from that—to choose instead to build what is essentially a digital library and appreciation society for cartography—speaks volumes. It suggests a critique of the current tech moment, not through words but through action. Campbell isn't anti-technology; he's a skilled engineer applying his craft to a different problem set. The problem isn't "how do we generate more?" but "how do we help people connect with something specific and beautiful?" His site, Past Maps, isn't a platform designed for exponential growth; it's a portal for enthusiasts, historians, and the simply curious. It’s a project built for the long tail of human interest, not the hockey-stick graph of user acquisition.

This move feels like a return to an earlier ethos of the web, before the gravitational pull of a few platforms drew all innovation and capital into their orbits. It’s reminiscent of the "small web" or "indie web" movements, where the goal isn't to build a billion-dollar unicorn but to create a useful, interesting, or delightful corner of the internet that serves its community well. In rejecting the AI pipeline, Campbell is betting on the enduring value of specialized knowledge and craftsmanship. An AI could certainly generate images of historical maps or even, with enough training, invent plausible ones. But it cannot replicate the human judgment that curates a collection, the scholarly context that makes a map meaningful, or the simple, tactile delight of exploring a well-organized archive. He's investing in a kind of value that algorithms can't easily replicate: the value of human taste and context.

There's also an unspoken challenge here to the narrative of inevitability that surrounds AI. The investor pressure he faced—"we'll write you a blank check"—frames starting an AI company as the default, sensible, almost obligatory path for a founder with his profile. Campbell’s rejection demonstrates that agency still exists. It’s a reminder that founders can choose their own north star, even if it points away from the hottest trend. Perhaps he’s seen the inherent volatility and winner-take-all dynamics of the AI race and decided he’d rather build something with a more predictable, if smaller, impact. Or perhaps, having already had a successful exit, he’s using his freedom to pursue a passion project that aligns with his personal interests, a luxury many founders don’t have until much later, if ever.

In a broader sense, his project is a hedge against digital homogeneity. As Google's search algorithms increasingly prioritize AI-generated answers and perhaps sideline niche, authoritative sources ("Google Zero" as the article hints), websites like Past Maps become more valuable, not less. They are repositories of curated expertise, the kind of deep, specific content that becomes harder to find when the web is flooded with AI-produced filler. Campbell might be building a small boat in the face of a coming storm, but if he stocks it with enough genuine treasures, it might just find its harbor. His work is a testament to the idea that not every meaningful technological endeavor has to scale to the heavens; some are best served by staying grounded, rich with context, and quietly, stubbornly human.