The limits of complexity

In our daily lives, we are engulfed with complexity. It is all around us: just think about business transactions, global trade, health care systems, jet engines, the energy grid, the tax code, computers, … Many would even add the remote control of their stereo or the radio in their car to that list.  Yet despite that almost permanent exposure, we are not comfortable with complexity. The underlying reason for this uneasiness is the “behaviour” of complex systems: they seem to lead a life of their own; they respond to their environment; not always according to our intentions or expectations. They seem beyond our understanding, so that we don’t know what’s coming next. Such systems seem to prevent us from making solid predictions, they deny us the certainty about the future that we desire. Hence we try to control complex system, to contain their complexity.

The question is: can we actually achieve such control? Should we even try? For it often seems that we are fighting an unwinnable war against the boundless growth of complexity. Well, are there really no bounds? Time to revisit the limits of complexity. I’d offer that there are two: an outer limit as well as an inner limit.

The outer limit – energy capture

Seen through the lens of Joseph Tainter, complexity is society’s strategy for solving problems. But that strategy is constrained by an outer limit: the amount of energy available to that society. The story goes like this: over longer periods of time, as problems keep arising and solutions keep being implemented, complexity will steadily accumulate. This increase is actually an unintended consequence of our innovative endeavour: we tend to add new features to existing systems, to use quick fixes and simple approaches, rather than considering a full redesign of the system in question. Thus overall, our own actions are the driving force behind the increasing complexity that we can observe –much to our dislike– in any human system (e.g., entire societies, economies, legal systems, and our techno-sphere as well). And as the available energy is limited for any society, the situation resembles an accident waiting to happen.

If the limit is reached, society cannot solve its ensuing problems anymore. It must –at least partially– collapse, and in the course of that downturn it will reduce its complexity back a level that is sustained by the available energy capture. Energy capture as the outer limit thus defines the maximum level of complexity that a society can sustain. And this limit does not only apply to entire societies; it works equally well on any other complex system.

The inner limit – connectedness

But even if you keep a safe distance to that outer limit, you can still run into existential trouble. And that is due to the inner workings of complex adaptive systems. The evolution of such a system over time is very well depicted through the adaptive cycle proposed by Lance Gunderson and Crawford Holling. This cycle presents a model that explains why complex adaptive systems usually go through longer periods of rather steady growth that are interrupted by short, intense periods of renewal. For today’s discussion, the key argument is the following: during the period of rather steady growth (usually much desired in human systems), the internal connectedness increases: the interactions between the different components and processes become more and more interdependent, while the overall resilience of the system decreases. Hence, the system becomes stiffer and stiffer (or less and less elastic) until it is not resilient enough to cope with change. In that instance (through external trigger events or partial internal failure) the system cannot bounce back anymore and will collapse.

In human systems, we seek to optimise for efficiency. In fact that is the path of ever-increasing internal connectedness. It is through our own action, though unknowingly, that we make our systems more and more susceptible to failure, gradually pushing them toward their inner limit. And this inner limit can be reached well before the outer limit: the system can run out of resilience well before it runs out of energy. Therefore, in order to stay on the safe side, you’ll need to keep a critical eye on both limiting factors.

What can we do? 

But apart from trying to keep a safe distance, what can we do about the outer and the inner limit of complexity? How should we focus our innovation activity, our quest for problem-solving?  We have to realise that we ourselves are origin of ever more complexity, despite our best intentions. That said, we cannot simply revert that trend, because we have to continue to solve the problems that arise.

There are three potential approaches. They are partially intertwined, and none of them is easy:

• The simplest idea (not the easiest path, though) would be to try and increase energy capture. But that’d be an eternal race to keep the energy ceiling above the increasing complexity levels. The longer-term sustainability of this approach is doubtful.
• The second would be dedicated efforts to reduce complexity. Rather than trying to lift the ceiling, that is the approach to keep complexity within the bounds of what energy capture can sustain. It would require a far more comprehensive approach to designing and our systems, taking into account many more influencing factors than we handle today. That would slow down systems design processes and make them more expensive.
• The third would be to limit connectedness within human systems. That would force us to accept an element of inefficiency in the way that we design and run our human systems. Again, that would eat into short-term profit margins.

All those approaches imply that we’d need to rebalance our thinking; moving away from a single focus on the short-term gain of our immediate next steps, and factoring longer-term considerations into our planning and doing. That’d be a very different way to think about the future.

A society that is fit for innovation retains sufficient resilience to respond to unexpected changes, while it keeps a sound distance to the invisible energy ceiling. Granted, that’s easier said than done, but a shared comprehensive understanding of the sources and the limits of complexity would be a promising step in the right direction.