Antifragile in 2022

Jan 1, 2022

It's been 661 days since COVID-19 was officially named a global pandemic. Much of our day to day life has completely changed. Zoom, handshakes, masks. But, maybe more interestingly some things haven't changed one bit.

Some things actually thrived even in the face of enormous shocks to the system. E-commerce pulled forward, SaaS multiples increased, and we accelerated much of our digital transformation in society and at work.

Nassim Taleb defined antifragile as a system that gets stronger as a result of stress. It goes beyond being resilient, the system gets better when shocked with negative input.

Nature provides us with antifragile systems. Exercise is antifragile. The more stress you put on your body, the more your body adapts. Evolution is antifragile – advantageous mutations are selected, and non-advantageous genes are penalized. Our immune system is antifragile.

Many machine learning algorithms are antifragile. Generative adversarial networks (GANs) pit two neural networks against each other, one generating candidates and the other evaluating them. Great system design is antifragile. Great organization are antifragile and learn from their mistakes. Lindy systems are often antifragile.

As we look at what didn't change during the pandemic, we'll find systems that are inherently antifragile. COVID-19 reinforced their dominance. Healthcare system will (hopefully) be better prepared for the next pandemic. Many organizations will be ready for their workforces to remotely work in the future. We learned lessons (hopefully) on large-scale contact tracing, the effectiveness of masks, and the effects of restrictions.

One of the most interesting corollaries of this idea is that optimization is inherently fragile. This goes counter to what we usually think. Systems are made stronger through their response to black swan events. These events can never be optimized for. Something to keep in mind.