4 ms·

One risk is self limiting whereas the other is self propagating with unknown long term consequences.

arcticbull 6y ago
This is quintessentially the human propensity to fear the new/unknown more than the known. Over your life 3,500,000 people will die in car accidents in the US, 6,000,000 will die of the flu. 7,000,000 will die of opioid overdoses. Each of these has serious, long-term consequences. We shouldn't panic, we should make decisive decisions based on the best information we have, and we should seek to reject fear-driven reactions just because something's new and scary. R_0 really isn't relevant, we're all in all of these together.
- danharaj 6y ago
  As I have mentioned in another thread, when you're dealing with an extreme process with exponential growth and a large absence of information, """"fear"""", also known as respecting your fucking ignorance, is most rational. Even if after the fact it looks like the pandemic didn't have catastrophic long term effects, it was still rational to be extremely cautious. Car accidents don't have the potential to collapse society. A car crash has a chance to cause other crashes but that correlation dies off exponentially fast over relatively small time and distance scales. A pandemic in our modern connected world is basically scale invariant without mitigations. With exponential phenomena, if you wait for evidence of extreme downside it'll already be too late. The absence of evidence is not evidence of absence. You can't wait, you have to act on your ignorance. This is not "panic". Rationality is not just about reasoning about what you know, it's about reasoning about what you don't know as well. In typical, mundane processes you can act on ignorance by collecting evidence. In extreme, self-propagating processes, one needs to apply the rationality of ignorance: extreme value theory and the statistics of extreme distributions.

This is quintessentially the human propensity to fear the new/unknown more than the known. Over your life 3,500,000 people will die in car accidents in the US, 6,000,000 will die of the flu. 7,000,000 will die of opioid overdoses. Each of these has serious, long-term consequences. We shouldn't panic, we should make decisive decisions based on the best information we have, and we should seek to reject fear-driven reactions just because something's new and scary. R_0 really isn't relevant, we're all in all of these together.

danharaj 6y ago
As I have mentioned in another thread, when you're dealing with an extreme process with exponential growth and a large absence of information, """"fear"""", also known as respecting your fucking ignorance, is most rational. Even if after the fact it looks like the pandemic didn't have catastrophic long term effects, it was still rational to be extremely cautious. Car accidents don't have the potential to collapse society. A car crash has a chance to cause other crashes but that correlation dies off exponentially fast over relatively small time and distance scales. A pandemic in our modern connected world is basically scale invariant without mitigations. With exponential phenomena, if you wait for evidence of extreme downside it'll already be too late. The absence of evidence is not evidence of absence. You can't wait, you have to act on your ignorance. This is not "panic". Rationality is not just about reasoning about what you know, it's about reasoning about what you don't know as well. In typical, mundane processes you can act on ignorance by collecting evidence. In extreme, self-propagating processes, one needs to apply the rationality of ignorance: extreme value theory and the statistics of extreme distributions.