Wild animals forage. They spend calories and take risks to acquire food. If the required work and risk expended are more than the food they acquire, they go extinct. The goal is to get as many calories as possible for as little effort as possible. If there’s a surplus, their instinct is to have kids until there isn’t a surplus.

It’s a search for calories.

Cavemen were in a similar situation. Invest effort and get calories in return.

Organization and culture amplify the opportunities and create complex systems, all in search of energy. A few hunters working together could catch a different sort of animal. A system of cooperation creates a commons for sheep to graze, and private property creates an incentive to invest in farming.

Investment in search of a return in energy. Sometimes we call that ‘profit’.

Calories turn into energy, but more directly, there’s the energy of electricity or heat, which we can then use to plow or cook or fertilize, creating energy in a form we can consume. Researchers talk about this entire category as ‘energy’ but tend to focus on things like oil–but the thinking applies just as well to spinach and to customers.

You don’t need much organization to collect sticks for a fire, but once the sticks are gone, the next highest yield on effort requires focus, patience and coordination. If the culture has momentum, it continues to move through the energy gradient, investing more to get more.

Marketing is the same way. When there are lots of customers and little in the way of competition, the return on marketing is very high. As sales get harder, we build systems and promotions and scaffolding to reach people who need a different story. And when the structure is too complex and the customers we acquire don’t pay for themselves, the organization fades away.

The metaphor is everywhere we look.

Most of the people who work at Yale don’t teach classes, and airlines have a very small percentage of pilots among their staff. We create organizations to deliver value, and we add complexity to coordinate and amplify the skilled work that people trade time or money for.

The bursar or the custodian might not be teaching a class, but their work increases the return the organization receives, and so the complexity they bring to the system is worth it.

A bicycle is the most efficient way for a person to get around, but in order to get that return on effort, we also need to coordinate and invest in a system of mining, manufacturing and road building. Everyone who is part of the project increases society’s cost, but the benefits of mobility, at scale, make it worth it.

Society (a small company, a country, anything in between) can be seen as layers of complexity, created to maximize our return on energy investment. We spend more to get more, and the shape of that curve determines how or if we’ll grow and thrive.

If we look at food, the efficiency of beans is about 5 to 1. For every calorie we invest to harvest beans, we generate 5 calories of benefit. For beef, it’s less than 1 to 30–in the wrong direction. Every 30+ calories we invest get us 1 calorie in return. Beef isn’t an efficient form of energy investment–we can build systems to make it more efficient, but we’re always going to be behind.

When an organization gets too complex, it becomes unwieldy and inefficient at its job. Most people are supporting and maintaining the system, coordinating with each other, not doing the work. As time goes on, stasis and the status quo grind progress to a halt. Complex systems get less efficient and harder to coordinate. Big science, big medicine, big government–sooner or later, the leverage stops paying off.

There are some interesting problems, and also some brand new shifts that will change the energy yield curves.

Consider a car in 1950. It was a complex (and complicated) set of pulleys, wires, cables and steel. It’s unlikely we could have made cars more efficient or safer than we did then–until the computer chip arrived. Its ability to allow us to efficiently increase complexity gave us an entirely new generation of transportation.

For my entire lifetime, the computer chip has enabled us to build increasingly complex organizations and systems that extract energy and generate value in ways that were impossible before it was invented. But the yield of any complex system peaks sooner or later.

Back to the original source–the energy we use to power all of this. When there were giant pools of oil just below the surface (oil that took millions of years to produce) we could extract energy from the ground virtually for free. A hundred years ago, we had to expend one unit of energy to get 100 out of the ground. It was easy and cheap and needed very little coordination, technology or investment.

Today, the EROI (energy return on investment) for some kinds of oil has dropped from 100 to 3. We only get 3 units of power for every one we invest to get it processed. It’s generally understood that once an energy source is less than 8 or 9 by this measure, it’s no longer economically feasible. It costs too much to get the energy.

So: If an organization gets too complex, it stops working. And if the EROI gets too low, it’s not worth moving forward. Society is at a crossroads as we are challenged by too much complexity and not enough EROI–for food, for learning, for power. (And the same is true for the customers we seek to find or the projects we seek to build.)

And, and it’s a significant ‘and’, the EROI of solar has been climbing (wind is even higher, and nuclear, by some measures, higher still, but each has their own caveats and challenges). In the charts below, up is better than down.

And the final wildcard is AI. If the computer chip created new frontiers in building complex systems, proponents of AI are saying that it opens new possibilities for coordination and control of complex systems, systems that will give us a new, more efficient way to extract the value we need.

The dance continues. Complex systems increasing yield but then fade as they get too big. Energy sources that have high yields moving relentlessly lower as the easy wins are taken. And then breakthroughs in coordination and processing that allow us to lean into the process all over again.

Avoid systems that get stuck in scale. Seek out energy gradients that work in your favor. Mostly, pay attention to the changes in each, as they never stay the same.