# Asteroid or comet impact

On average, there are “91 deaths every year from asteroid strikes”, which represents 1 micromort per lifetime; see: https://www.bbc.com/future/article/20120222-waiting-for-a-rock-to-fall.

Topics to discuss:

• all-things-considered chance of strike, broken down by type of object (large asteroid, smaller asteroid, comets, etc.)
• what percentage of each kind of object we’ve discovered, and how we would even know this percentage (i.e. how do we estimate the denominator?)
• base rate of collisions and particularly notable incidents, e.g. Cretaceous–Paleogene extinction event, Tunguska event.
• if something were to strike, how bad would it be? what’s the probability that it would lead to existential catastrophe? This should also be broken down by type of object.
• if something were to strike, could we deflect or destroy the object before it strikes?
• if we create new technology to deflect or destroy the object before it strikes, could that technology be used in a “dual purpose” way for malicious or other bad ways? (e.g. using it to intentionally make an asteroid collide with the Earth, or large explosions on Earth)
• talk about the social/epistemological value of asteroid strike prevention: relatively straightforward, good example of humans coming together to think about a risk before it happens, and how you have to combine hard numerical estimates of risk with fuzzy/subjective estimates of e.g. the same technology being used in bad ways.

# Problem

## Causal logic

An asteroid could strike the Earth and cause major damage to human society and the biosphere.

## Evidence

The statistical frequency of asteroid impacts and their basic energies and effects are well understood. However, most large near-Earth objects have been identified and none of those pose a threat.

## Scope

The damage from an asteroid impact depends on its size. Larger, more destructive asteroids are rarer. Small asteroid impacts are more likely, but they would do little damage except in the very unlikely event that they land on a metropolis.