Complexity engineering

Hi all -

Finally, finally, here is a short(er) newsletter.

+ mental models #

* Complexity engineering

I’ve written a lot about complex systems in the past. The reason I’m fascinated by them is because you simply get behavior that you don’t find anywhere else.

Want non-linear outcomes? Feedback loops? Virality? Emergent order?

If you can build complex systems, you can build these features.

And if you are working with complex systems - which includes virtually everything with other people - then you can also plan for these features.

Sometimes this stuff works against you: viscous cycles, financial contagion or sudden catastrophe. But other times it also works for you: serendipity, celebrity and renown, or financial windfall.

One of the biggest takeaways of complexity theory is that these things are not accidents. They are inherent properties of complex systems. With enough complexity and enough time, crazy things will happen - you just have to hope you’re on the right side of it.

Well then, two questions come to mind: (1) how do we build these complex systems, and (2) how do we get on the right side of them?

To answer (1), you could read all the stuff I’ve written in the past about it (like here or here or here).

Or! You could skip all of that and just read this one really concise, well-written essay by an actual complexity scientist, Joseph W. Norman. In it, you would discover fairly quickly how exactly to engineer a complex system, such as:

• Focus on variety, diversity & random mutation
• When resources are abundant, experiment widely
• When resources are scarce, be a little more careful
• Encourage and reproduce good traits
• Remove bad traits quickly
• Grow parts of the system together - don’t drop things in
• Grow bottom-up, not top-down
• Don’t grow faster than the surrounding ecosystem naturally allows

If a lot of this stuff sounds like evolution, you would be right. Evolution is an algorithm that has existed for billions of years - it’s probably pretty good! I think a wider appreciation of that algorithm can help us a lot as we work with and build complex systems.

Now how about (2)? How do you get on the right side of complexity?

Here, I’m on less certain ground, mostly armed with vague intuitions and unfounded opinions. But my guess is that it has to do with building complex adaptive systems.

In other words, you want to make sure the system you’re in is an adaptive one. It’s resilient, it’s stubborn and it doesn’t die easily. It’s deeply embedded in a mesh of relationships around it - it depends on them and they on it. And most importantly, it’s antifragile. The evolutionary algorithm is baked into its very existence, and so with every failure, it grows just a little bit stronger.

If this thing sounds like a freak of nature - something so powerful and relentless and uncontrollable - then you would also be right.

Because that would describe the human race.

Biology has been engineering complex systems for millennia, and we humans have benefited immensely from it. It’s about time we learn how they work so that we can engineer the next generation of complex systems ourselves.

Thanks for reading,

Alex