I’ve always been fascinated by how connected we are. I’ve met some famous people in totally random situations, like Mike Love from the Beach Boys at a hotel in Washington, DC, and George Shultz, who was Reagan’s Secretary of State, in an elevator in San Francisco. Also, one day while running along The Charles River in Boston I had a conversation with a homeless man who was from Bari, Italy, where my grandparents were from. On business trips around the world, I’ve met random people on trains and planes who grew up near me in New York. 

When I moved to Sun Microsystems in 2000 in Silicon Valley, that’s when I really learned about the scale of massive human networks via my work on Free and Open Source Software Communities (FOSS). That experience hooked me. So when I watched Derek Muller’s Veritasium video, “Something Strange Happens When You Trace How Connected We Are,” it really hit home. This is not just math. It’s the invisible webs linking people, ideas, diseases, and behaviors. With simulations and real-world examples, Muller unpacks the six degrees of separation concept, showing why our world sometimes feels small despite eight billion people. It’s a deep dive into network science that reshapes how I view communities, opportunities, and my own connections. As someone who’s spent decades building FOSS communities, I see parallels in how software developers form networks. Here’s my take.

The Classic Hook: Six Degrees Explained

The video opens with a striking story from 1999. A German newspaper, Die Zeit, challenges a falafel salesman and former theater director, Salah ben Ghaly, to connect to his favorite actor, Marlon Brando. Through a chain of friends, family, and acquaintances, all on a first-name basis, they do it in six steps. Ben Ghaly’s California friend works with a woman’s boyfriend, who’s linked to a sorority sister of the daughter of the producer of Don Juan DeMarco, starring Brando. Six hops. Muller’s point is clear. This is not a one-off. In a world of eight billion, any two people can likely connect in six steps or fewer. But how does this work? And what does it mean for us?

Muller frames it as a puzzle. If connections were random, say each of us with 100 friends worldwide, the math checks out. One person’s 100 friends, times their 100, and by step five, you’re at 10 billion, overshooting Earth’s population. Straightforward. But real life is not random. We cluster geographically. Most friends live closely nearby and know each other personally. Picture eight billion people in a circle, each linked to 50 neighbors on either side. Connecting opposites takes 40 million steps on average. Six steps? No, you’re nowhere. Yet the world feels small. That’s the mystery.

Network Scientists and Shortcuts

Muller introduces Duncan Watts and Steven Strogatz, network scientists who tackled this in the mid-1990s with their small-world model. They start with a regular network with nodes in a circle tied to nearby neighbors, which mimics real social clusters. Then they add shortcuts, which are random links to distant nodes. In a 1,000-node simulation, rewiring just one percent of links drops separation from 50 to 10. Scale to eight billion? Only three shortcuts per 10,000 friendships bring it to six degrees. Clustering stays high, but the world shrinks jet quick.

Strogatz sums it up. The question is not why the world is small, it’s how could it be otherwise? I’ve felt this myself. A random meetup or conference chat creates a shortcut, putting me one step from a developer in Amsterdam or a startup founder in Tokyo. The video connects this to a concept in sociology called the strength of weak ties. Jobs often come from acquaintances, not close friends since weak links bridge distant networks. Muller suggests attending random events to boost these connections. I’ve landed opportunities this way, showing up to a meetup where I knew no one, only to meet someone who opened new doors. However, randomness matters. I got my job at Sun Microsystems by applying blindly to HR! I didn’t know anyone at Sun at the time. You never know what happens when you move in one direction or another. 

Proving the Small World

Watts and Strogatz tested their model on real data. In 1996, with no Google, they analyzed the nematode Caenorhabditis elegans worm’s neural network, which has 282 neurons with an average of 14 connections. Arranged linearly, it would take 14 steps to connect extremes. In reality? Just 2.65 degrees. Hollywood’s 200,000 actors average under four degrees, cue that old Kevin Bacon game. U.S. power grids also exhibit small-world properties, with key substations acting as connectors. Their 1998 Nature paper became a sensation, now with 58,000 citations, outpacing landmark works on the Higgs boson or DNA. The irony? It seems that a paper on networks does indeed spread like wildfire.

Muller and Strogatz explore unexpected applications of their small-world model, which reveals its reach beyond theoretical networks. For instance, they recount a surprising call from the FBI, which sought Strogatz’s expertise to apply network science to forensic criminology. The agency was investigating “hair and fiber” networks to calculate probabilities of secondary transfers in cases where, say, a fiber from a suspect’s clothing might end up on a bus seat, then transfer to another person, which can complicate crime scene analysis. By modeling these connections as a network, investigators could better estimate the likelihood of such transfers and refine their approach to evidence. Another example touches on epidemiology, where small-world principles help trace contact networks in disease outbreaks, which helps researchers identify key individuals who might unknowingly spread pathogens. These examples highlight how our interconnected world, while powerful, can amplify both solutions and problems.

Hubs and the Dark Side of Connectivity

The video shifts to disease spread. Simulations show a regular network takes 73 steps to infect 100 nodes. Add 10 percent shortcuts? It’s 26 steps. Fully random? 25. For billions, though, less than one percent shortcuts suffice. This echoes global pandemics with air travelcreating shortcuts that push spreading viruses fast.

Albert-László Barabási enters the discussion, drawing from his groundbreaking work on the internet in 1998. At that time, the web had around 800 million pages, yet any two could be connected in just 19 clicks on average. This was puzzling because it didn’t align perfectly with Watts and Strogatz’s small-world model, which focused on ordered networks with random shortcuts. Instead, Barabási identified a different structure: scale-free networks. These connections don’t follow a normal bell curve distribution, where most nodes have roughly the same number of links, like people’s heights clustering around an average. Rather, the web exhibits a power-law distribution — a steep initial drop followed by a long tail. This means most pages have few links, but a small number of super-connected “hubs,” like Yahoo or early search engines. They link to thousands or even millions of others. These hubs make the entire network navigable in so few steps and turn vast systems into small worlds not through random shortcuts alone, but through unequal connectivity.

Barabási explains how this emerges: through two key principles — growth and preferential attachment. Networks don’t appear fully formed. They expand over time. The web started small in the early 1990s and grew to trillions of nodes one page at a time. When a new node joins, it doesn’t connect randomly. It’s biased toward popular nodes. You’re more likely to link to a well-known site like Wikipedia than an obscure blog because it’s visible and useful. This “rich get richer” effect, called preferential attachment, naturally creates large hubs. In simulations with his colleague Réka Albert, they start with a few connected nodes and add new ones that preferentially link to those with more connections. Over time, a few nodes become massively dominant, mirroring real networks.

This scale-free property isn’t unique to the web; it’s universal in complex systems, from social media follows to airline routes.

Real-world hubs are everywhere. Take Chicago’s O’Hare airport, with over 200 direct flights. It’s a massive travel hub. But when storms hit in August 2025, 280 flights were canceled, impacting six other American airports. Ecosystems rely on keystone species like Atlantic cod. Remove them, and food webs collapse. In cells, ATP hubs drive reactions. In brains, the prefrontal cortex links functions. Barabási notes that once hubs are there, they fundamentally change the system. They’re efficient but fragile, useful for drugs targeting bacteria hubs but disastrous for pandemics.

Thailand’s 1990s HIV crisis shows hub power. Broad public health campaigns failed, but when officials targeted brothels as hubs and mandated condoms that cut infections by over 50 percent in military recruits, which prevented five million cases by 2013. Smart strategyto target the inherent weakness of large hubs. 

Behavior in the Net: Cooperation vs Chaos

The video also delves into how networks influence human behavior, using the Prisoner’s Dilemma, a cornerstone of game theory, to illustrate this. Imagine two players, each deciding whether to cooperate or defect in a scenario involving coins as rewards. If both cooperate, they each earn a moderate reward, say 3 coins. If both defect, they get less, maybe 1 coin each. But if one cooperates while the other defects, the defector scores big with 5 coins, and the cooperator gets nothing. From a purely rational, one-time perspective, defecting seems like the best choice — you either win big or minimize your loss. Yet, this leads to both players defecting, earning less than if they had cooperated. This paradox highlights a tension in human interactions: individual gain versus collective benefit.

The dynamics shift when the game is played repeatedly, simulating ongoing relationships like those in social or professional networks. Here, a strategy called “tit for tat” emerges as powerful and effective. In this approach, you start by cooperating and then mirror your opponent’s previous move — cooperating if they cooperated, defecting if they defected. This simple rule fosters trust and punishes betrayal, which encourages long-term cooperation. Political scientist Robert Axelrod ran famous computer tournaments in the 1980s, pitting various strategies against each other in repeated Prisoner’s Dilemma games. His findings, detailed in his book The Evolution of Cooperation, showed that “nice” strategies like tit for tat consistently outperformed selfish ones. Why? Because cooperative strategies build trust over time, especially in clustered networks where players interact repeatedly with the same neighbors. These clusters, like tight-knit communities, create environments where cooperation can thrive, as defectors are gradually outcompeted or isolated.

Watts and Strogatz simulated this on a network. In a regular setup, cooperation spreads from clusters. Add shortcuts? Defectors dominate. A critical threshold exists, more shortcuts, zero cooperation. Strogatz explains that clumps foster trust, global nets amplify toxicity. This mirrors the phenomenon of online keyboard warriors as the internet erases community pockets and fuels negativity. Social media connects but often divides and spreads harm. It seems humans compete and cooperate depending on the situation. This is a key lesson for people who build large communities for a living. 

Watts’ real-world experiments puzzled him. Network structure seemed neutral at first with cooperation as likely in clustered or random setups. But digging deeper, he found that clustered networks amplify copying. If someone cooperates by chance, others follow, but defection spreads just as easily. Over many games, these balance out. Allow players to choose connections? Cooperation surges. Muller’s takeaway. Curate your network, avoid negativity. These insights resonate with my own experiences.

My Take: Power in the Weave

This video is an eye-opener. We’re constantly shaping our networks through shortcuts, hubs, and choices. One action can tip the system. Steve Jobs said those crazy enough to think they can change the world do. This video shows how. Simulations, linked in the video, let you explore disease spread and hub growth to illustrate these concepts. 

In my work building FOSS communities, like through Duke’s Corner Java Podcast, I see these principles in action. Software communities are networks. Hubs like the Java Champions and Java User Groups or OpenJDK contributors connect thousands of developers. Shortcuts? A student’s pull request on GitHub links them to a global project. JCrete’s unconference gatherings and hacking sessions fosters life-changing career connections since such high value developers participate. I’ve interviewed hundreds of developers, from user group leaders to enterprise architects, and their stories echo this video. Collaboration, contribution, and shared purpose drive vibrant ecosystems, whether a local meetup or a global project spanning dozens of countries.

But there’s a flip side. Online toxicity, like flame wars in project issues, thrives in unclustered, chaotic nets. The fix is intentional community building. Encourage participation, foster leadership, ensure transparency, just as Watts’ experiments showed choice promotes cooperation. Every developer can shape the network. A single commit or meetup talk can spark a movement, connecting you to mentors, peers, and opportunities. As I’ve learned from 20 years in FOSS, contributing to communities like Java’s is not just coding. It’s building networks that fuel innovation and careers. Next random invite? Take it! Contribute to a project, join a user group, share your story. It might shrink your world or spark something big. Watch the video and rethink your connections. Communities are networks, and we all have a role in weaving them stronger.