We’re witnessing more and more instability factors such as climate change, economic inequalities within a country, decrease in key resources like oil and drinkable water, etc. Those factors are generally interconnected, e.g., a decrease in oil production could decrease the energy available to produce drinkable water. And lack of drinkable water could trigger a war, say, in The Middle East, and induce global fuel shortage. You get it.

These factors create giant self-reinforcing feedback loops we don’t fully control (and we’re not prepared). Of course, we can reduce worldwide gas emissions and shift to more renewable energies or nuclear plants but the climate change flywheel has already started to move. As the likelihood of one of those factors blowing up increases, there’s a high chance that some of the businesses we see today will evolve.

In his book The Wealth of the Commons, David Bollier proposes to ask, “How would this [development, business, community] function if oil cost $200 a barrel?”. Spoiler: most companies would be wiped from the surface of the Earth overnight.

The human world is more fragile and uncertain than ever. If things continue in this direction, it could lead to more war, violence, nomadism, poverty, etc. Quite bleak, isn't it?

Fortunately, the solution already exists and can be drawn directly from:

• Nature through evolution and natural selection theory: species have evolved to endure and adapt (with some limits) to threats and terrible events, allowing them to survive other ages.

• Architecture through earthquake-resistant buildings in Japan: because of the high concentration of strong seisms in Japan, building engineering has been adapted to resist such events and avoid the complete destruction of entire cities.

That solution is resiliency.

What is resiliency?

First, let me tell you what resiliency is not.

Banks are not resilient, and the current financial system is not either. We noticed it in 2008 and more recently with the bankruptcy of SVB. States come to the rescue when the economy cracks down. Fair enough, but that doesn’t help banks grow stronger.

Another example: floods caused many casualties and damage in 2023, especially in India, where human infrastructures are unsuited for such dramatic events.

So what’s resiliency?

According to Wikipedia, resiliency has numerous definitions depending on the sector it applies:

• For an ecological environment, it’s the capacity to recover from perturbations (e.g. after a wildfire, a forest is capable of regenerating itself thanks to the protected seeds in the soil).

• For a computer network is its ability to maintain service in the face of faults.

When you expand these different definitions to a more general one, you get (in my own words) the following: it’s the capacity for something to maintain or regain its capabilities over time no matter what happens.

Resiliency as a requirement

Resiliency is what people are craving for. They don’t want their house to be destroyed by the first flood. They don’t want their lifetime savings to turn to dust, blown by the wind of the next crisis. They want their computer or dishwasher to last at least a decade. They want stability.

Resiliency is also crucial for a company’s success. It brings confidence and allows for long-term planning. Without resiliency, a project may only focus on short-term growth, which is not the goal for major companies or ambitious startups looking to redefine the world. In fact, a lack of resiliency is the shortest path to failure. This is why startups try to have a runway of several years. With money, you can adapt.

If a product or service wants to last, it must be resilient.

How to become resilient?

First, make something that people care about. It seems dumb, but if the product or service is useless, it won’t be used, maintained, or fixed. It is doomed to fail.

For the rest, I’ll speak about what I know best: web3 which takes resiliency at heart.

In web3, resiliency depends on:

• The number of critical parameters (things that define how the solution behaves) to monitor and manage to make the solution to work;

• The number of code dependencies the web3 protocol relies on. They change often.

The fewer parameters and dependencies, the better.

One advantage of web3 projects is their reliance on decentralized public blockchains. Even if there’s an outage in a specific region, the solution continues to work. On the opposite, if a solution depends on ChatGTP’s API and the next day OpenAI shuts it down, you’re screwed. More generally, If you rely on a centralized service and there are no easy alternatives you’re screwed.

Open-source decentralized services do not evolve as centralized ones. This is at the same time a bug and a feature. It is slower so it can’t adapt to new competitors (in theory) but building on it is much easier since you know the rules, how it works, and how it behaves.

One of the most successful protocols in DeFi (Decentralized Finance) is Uniswap. It has no dependency whatsoever. It’s immutable meaning it can’t be changed anymore. Rules are clear. It’s easy to plug into and to trust. It’s built to be built on. It just works. This is much more reliable than any other alternative that could have dependencies and third parties.

On some other protocols, rules are clear but can be updated. Occasionally it reinforces the resiliency of the protocol against unexpected events. But in general, it’s a flaw. The problem with such a design is that any users or integrators using the solution must closely monitor parameters’ changes. The more parameters, the more painful, the less resilient.

Thus, the best design solutions rely on the following:

• as few modifiable and critical parameters as possible.

• as few dependencies as possible. The solution should be able to respond to all changes that could erect from the different dependencies. By recursion, apply this to the dependencies of dependencies...

Ideally, the solution should have no dependency. If it has one, it must survive if the dependency fails, disappears, or blows up. Not conforming to this design pattern exposes the solution to huge operational costs (monitoring) and poor adaptability (it must react faster than its dependencies).

What we did wrong at Morpho

We faced that exact problem in the first iteration of Morpho. It has been built onto other protocols (Compound and Aave) with hundreds of risk parameters to update depending on market conditions. Operational costs skyrocketed due to ever-increasing parameters monitoring. This is not what we imagine as a final state for a lending protocol to be used by billions of people for sure. That’s why we’re designing a new protocol — one with minimum parameters and minimum dependencies.

How to transpose this to other areas?

Just translate dependency to: third party, subcontractor, fuel, material, humans, machine, things that fail, things that break, things that are fragile, things that require a lot of maintenance, you name it. I leave this as an exercise for the reader.

Closing thoughts

Resiliency is longevity. You wanna last? Be resilient.

I bet resiliency will become a buzzword in the next decade. The trend is slowly increasing. Let’s see how it plays out in the coming years.