How to determine if you’re dealing with a 'complex system'

A guide for embracing complexity & adaptive strategies

Hanna Barakat & Cambridge Diversity Fund / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/

I write about ‘systems,’ ‘sociotechnical systems,’ and ‘complex adaptive systems,’ which, is, uh, confusing. Let’s untangle these terms and help you determine what you’re dealing with.

A ‘system’ exists wherever people are trying to meet a need or solve a problem. (More) For example, a group of funders working to align the strategies of their grantees toward collective action is a system. If this system is mediated by technology — say the grantees are working to advance ‘fair and responsible AI’ — it becomes ‘sociotechnical.’ Sociotechnical is just a fancy word for how technology shapes people, and people shape technology.

But at a systems level, we’re not looking simply at how people shape technology, we’re looking at how systems of race, gender, status, and power shape technology (Untangled Deep Dive), and in turn, how technology shapes these systems (Untangled Deep Dive). This mutually shaping process creates dynamics within a system that contribute to its overall behavior. To use our example, if the funders and grantees don’t understand how these hidden dynamics affect the pursuit of ‘fair and responsible AI,’ they’ll either design programs that create ‘unintended consequences’ or worse, perpetuate the status quo in the name of technological innovation and progress.

But here’s the crux: most systems aren’t just sociotechnical, they’re complex and adaptive too. This means that they behave unpredictable and uncertain ways and participants adapt to each other’s actions. As detailed in the Cynefin Framework by Dave Snowden, complicated systems respond to cause-and-effect in a linear and predictable manner, wherein if I do X, Y will occur, and we can verify that X caused Y. By contrast, in a complex adaptive system:

• Data and/or knowledge are incomplete or contradictory.

• You can’t predict the outcome. Cause-and-effect doesn’t exist. Outcomes are more often the result of non-linear dynamics. (Untangled Deep Dive)

• The component parts of the system are interdependent (More) and the macro behavior of the system emerges from how the parts adapt to one another, and the dynamics this creates (More)

An oft used example is the distinction between a car and traffic. A car engine is complicated. Its operation is based on well defined principles of physics and engineering. If you understand the components of how they interact, you can predict the engine’s behavior and fix it if it breaks down. Whereas a car’s engine might be complicated, traffic is a complex system. Traffic changes its behavior based on information from its environment. Focusing on one car won’t teach you much about the entire system because what matters is the interactions between them.