[ExI] libertarian (asteroid) defense

[Anders Sandberg]

Samantha Atkins wrote:
> On 02/27/2011 09:29 PM, Kelly Anderson wrote:
>> The total number of people currently employed in looking for asteroids
>> in the NASA Near Earth Object program is reportedly less than the
>> number of people working in a typical McDonalds. Since actuaries
>> indicate that we each have a 1:20,000 chance of being killed by such
>> an asteroid, that is a silly small number.
>
> What?? Are you seriously saying that 5 persons in every 100,000 will 
> be or have been killed by an asteroid?  The actual number historically 
> is closer to 1:1,000,000,000.  Or are you referring to the chance of 
> dying IF an asteroid over a certain size impacts the earth without 
> bothering to factor in the actual chances of that happening?

If you calculate the expected number of fatailities times the estimated 
frequency distribution of impacts, you get numbers like an annual risk 
of dying of 1 in 2 million (a lifetime risk of 1 in 30,000). These 
numbers jiggle depending on which dataset you use in the calculation, 
but basically the risk seems to be on the order of 10^-4 that we will 
have a asteroid GCR during our lifetime. It is megatsunamis or asteroid 
winters from 1+ km impactors that provide most of the hazard. 
Fortunately the size distribution has a rapidly declining power law tail 
and we have not seen any dangerous NEOs so far, but if you like to worry 
there is always the possibility that there are "black comets" that we 
are bad at detecting (sure, there might be just (say) 10% chance of the 
theory being right, but if it is right the risk might jump more than one 
order of magnitude).

I will be giving a keynote speech at the 2011 IAA Planetary Defense 
Conference this May with the title "The billion-body problem: taking 
human (ir)rationality into account for planetary defense". So I would 
love to mine this thread for good ideas. Here is my abstract:

    "Of all the types of global catastrophic risks facing mankind, NEO
    impacts almost represent a best case: they follow dynamical laws
    that are fairly well understood and deterministic, we have data on
    past effects, estimated probability distributions, there is some
    public understanding of the issue, and there is a community that is
    actually working towards ways of ameliorating the risk. Few other
    threats have this many positive factors working against them. Yet we
    should not expect planetary defense to be unproblematic even if the
    scientific and technical problems can be resolved. Part of this is
    due to the public goods nature of planetary defense: who will pay
    for something that benefits everybody? Another aspect is the long
    timeframes involved: how well can societies plan for predictable but
    remote threats when there are many more urgent issues? More
    fundamentally, human cognition suffers from several biases that are
    likely to both impair proper understanding and decisionmaking for
    this kind of rare but hazardous event. Can we get around the
    problems posed by human rationality for coordinating planetary
    defense? This is the billion-body problem."


Libertarian asteroid defense is a nice example of how a worst case 
scenario can be helpful for planning: if one can come up with a good 
private solution, then the public good issue might be solvable. The 
timeframe issue might be worse from our perspective than most, since if 
you are in a civilization hurtling towards singularity discounting of 
the future might go down: the next generation will be far more capable 
than the current one, and investing your capital in more growth gives 
better returns than investing asteroid deflection.

-- 
Anders Sandberg,
Future of Humanity Institute 
James Martin 21st Century School 
Philosophy Faculty 
Oxford University