I hosted a virtual salon last Wednesday. If you missed it, you can catch the recording on YouTube:

The hot topic of the Q&A was tsunamis. This was a bit off-topic – the essay I wrote for Summer of Protocols was based on an evolutionary analysis of coal mining safety. But now I kind of wish that I had written about tidal waves. Safety protocols for tail events like natural disasters are really important.

About 80% of Earth’s tsunamis happen in the Pacific Ocean. The word tsunami comes from the Japanese words for harbor (tsu) and wave (nami). One hundred and forty-three harbor waves have hit Japan since the year 684, killing over 130,000 people. There isn’t sufficiently strong technology to flat-out eliminate waves. So, the island’s denizens have had to come up with safety protocols. The most legible (and ominous) of these is the tsunami stone:

On the stones are warnings, like “Don’t build below this stone”, “A wave killed people here”, or “If the water goes out, seek higher ground”. One of the problems associated with managing tail risk is memory. Over time, we lose sight of just how bad things can get. If it’s been a long time since a disaster, new generations rely on stories instead of first-hand experiences. 

This poses a challenge, especially in small communities. For example, imagine an island town of 25 people where only one elder has seen a tsunami. It’s that person’s job to tell the story. But if the storyteller passes away, there’s a good chance that the story is lost. Myths are more durable than stories, but they require a larger social network to be passed on.

I think cultural memory plays a generally important role in safety. A pattern between company size and worker fatality rates has been found. Smaller companies have higher accident rates than companies with more employees. A company’s ability to “remember” protocols depends on the density of its neural net – its headcount. The same goes for cultures. 

Another factor that harms tail risk management is urgency. The costs of precautions are visible, but the corresponding benefits can take a long time to manifest. Precautionary measures can look paranoid or even inhumane. Many North Americans live in flood plains because it’s cheap to build housing on land that has historically (for good reason) been undesirable. To reclassify residential areas susceptible to disasters as no-build zones would, in the short run, appear to be a cruel policy.

Network size correlates with amnesia. Urgent needs justify the process of forgetting. With Remembrance Day around the corner, it’s a good time to look for social memory technologies. “Lest we forget.”

There’s a lot of discussion today of new threats (existential risks — “x-risks”) rather than managing old risks, like tsunamis. Some of these include genetically modified crops and bugs, artificial intelligence, bioweapons, and aliens. Many are legitimate systemic risks. Problem is: we don’t have protocols for these. So there’s a big temptation to put the lid on it and stop progress.

This makes sense in theory. If we keep doing what we’re doing, things won’t change that much. However, it appears to me that the wheel of technological progress keeps turning. I doubt that we will stop trying to make things - that will generate uncertain consequences. Keeping with the tsunami theme, most of us are just riding out the waves as things come up. Doing this well requires mentally athletic posture which allows us to protocolize and reprotocolize on the fly. Doomsurfers > Doomers