I recently finished ‘First Contact: New Guinea’s Highlanders Encounter the Outside World’, and I can’t stop thinking about what it reveals about how systems evolve. Not just human societies, but technological ones too.
For 10,000 years, roughly a million people lived in the highlands of Papua New Guinea with zero contact with the outside world. They had invented agriculture independently. They had stable societies. They had culture, art, and complex social hierarchies. But they never advanced beyond subsistence farming. Meanwhile, the Egyptians built pyramids, invented writing, developed mathematics. 4,500 years of divergence.
The book doesn’t frame this as primitive versus advanced. It frames it as a problem of constraints. And that’s where it gets interesting for those of us building systems.
Constraints Shape Architecture
The highlanders had everything they needed to evolve except one thing: scalability infrastructure. They farmed taro and bananas, root crops that rotted within weeks. They had no way to store surplus. No animals to transport goods. No writing system. No ability to centralize resources or coordinate large projects.
Compare this to societies with grain-based agriculture. Wheat stores for years. You can tax it, trade it across regions, feed armies. You can build bureaucracies around resource management. You can plan beyond the next harvest.
The highlanders couldn’t do any of this. So they stayed as small groups of a few hundred people, constantly at war with neighbors, focused entirely on survival. No group had the capacity to explore beyond the next valley. Innovation requires surplus resources and time to think beyond immediate needs. Without storage, without trade infrastructure, without a way to accumulate and distribute goods, innovation stopped happening.
This is what I think about when designing systems. Constraints aren’t just limitations. They’re architectural determinants. The shape of your infrastructure determines what becomes possible.
Network Effects Require Connectivity
Here’s something that struck me: things that worked traveled even when people didn’t. A better crop variety would spread through neighboring tribes through marriage and copying, not through central planning. But this only worked at the margins. A new ritual could diffuse through adjacent groups, but it couldn’t propagate globally because there was no global network.
Now think about open source. Think about how the entire tech industry runs on shared, copied, remixed ideas flowing through GitHub and conference talks and blog posts. We take for granted that knowledge propagates instantly. The highlanders didn’t have that. Their innovation diffused at the speed of foot traffic and tribal intermarriage, which means it stalled after a few valleys.
I’ve seen this in tech teams. A brilliant architecture pattern locked inside one team’s codebase doesn’t scale. You need documentation, communication, a network effect. The highlanders couldn’t document anything. No writing system meant knowledge had to be transmitted orally, person to person. Systems that scale require the ability to transmit knowledge at scale.
Big Man Leadership Without Inheritance
The highlanders used a system called ‘moka’ competitive gift-giving where leaders had to earn status through constant public displays of generosity. It wasn’t hereditary. Leadership had to be re-earned in every generation.
This is brutal for long-term thinking. You can’t plan 20-year infrastructure projects if your power is ephemeral. Compare this to hereditary chiefdoms on the coast where you could actually plan across decades.
The Real Barrier Was Isolation
I want to be clear: the highlanders weren’t incapable of technological progress. They weren’t less intelligent. They were operating under structural constraints that made progress mathematically unlikely.
They needed: ways to store resources, ways to transport goods, ways to record information, ways to coordinate larger groups, and crucially, access to outside knowledge and ideas. They had none of these. And the geography made getting any of them incredibly difficult. You had to trek through malaria-infested jungles to reach them. To the outside world, the highlands looked like impenetrable mountains with nothing inside.
So knowledge from the coast never made it up. The highlanders never heard about metallurgy, writing systems, or centralized governance. They couldn’t import these ideas because they had no contact points.
Systems evolve through exposure to ideas from outside. Without that exposure, without external input, you hit optimization ceilings.
The highlanders in 1930 were using farming techniques identical to their ancestors 10,000 years ago. Not because they were incapable of innovation, but because their structural constraints created no incentive for it. They had solved the problem of survival. The only way forward would have been contact with external knowledge and resources.
When Australians finally arrived with goods, shells, and technology, everything changed in a single generation. Not because the highlanders were suddenly smarter, but because their constraint environment shifted.
I think this applies to how we build software. Isolation optimization leads to local maxima. Real progress requires external input, feedback loops, and the structural capacity to integrate new ideas.