Phyllis Schlafly dead at 92

[DGuller]

So you don't think a deterrent effect has been functioning the past 60 years to make a major war unlikely?

It has been functioning, but at a tremendous unseen cost.  We've traded high probability of a recoverable high severity event for a low probability of an end-of-the-world event.  In the long run, it may well be a very bad trade.

[The Brain]

It's important to nuke responsibly.

[The Brain]

FWIW I don't see a major nuclear war leading to the end of humanity. Civilization maybe (such as it is).

[alfred russel]

So you don't think a deterrent effect has been functioning the past 60 years to make a major war unlikely?

It has been functioning, but at a tremendous unseen cost.  We've traded high probability of a recoverable high severity event for a low probability of an end-of-the-world event.  In the long run, it may well be a very bad trade.

I agree with your assessment. However, that requires an assessment of the relative probabilities and costs, and I don't see why it is necessarily insane to conclude the probabilities of the two events have diverged far enough to compensate for an increase in their severity.

It is also arguable whether an all out nuclear war would really end the world. It probably would for us as we live in major cities of one of two likely combatants, but I'm not sure some people in other places would survive the consequences.

[Razgovory]

Yes, it does change the meaning.  The US is hardly the only one to build nuclear weapons, there are several nuclear powers.  So there are several players playing the game and none of them have died. 

The player is the entire world.

In Russian Roulette we know a person will shoot themselves after X number of rounds.  This is a mathematical certainty, and is born out in real life.

Not if you spin the drum after every round.  It's a damn near certainty, but not a mathematical one.

Several nuclear powers have weapons that would not directly affect in the US.  If India nuked Islambad, if Israel destroyed Cairo or if North Korea destroyed Seoul, it would have a limited effect on the US nor would they destroy the world. So it's hard to see it as one game with the entire world being a player.

[DGuller]

I agree with your assessment. However, that requires an assessment of the relative probabilities and costs, and I don't see why it is necessarily insane to conclude the probabilities of the two events have diverged far enough to compensate for an increase in their severity.

In the long run, any non-zero probability event will eventually occur.  Now, yes, if that non-zero probability is so close to zero that something else will likely end the world first, then okay.  But humans get locked all the time into situations where they know their actions will perpetuate a destructive system, but they're nevertheless forced to act in such ways anyway.  The decision by a major power to launch nukes may be one such situation.

[alfred russel]

I agree with your assessment. However, that requires an assessment of the relative probabilities and costs, and I don't see why it is necessarily insane to conclude the probabilities of the two events have diverged far enough to compensate for an increase in their severity.

In the long run, any non-zero probability event will eventually occur.  Now, yes, if that non-zero probability is so close to zero that something else will likely end the world first, then okay.  But humans get locked all the time into situations where they know their actions will perpetuate a destructive system, but they're nevertheless forced to act in such ways anyway.  The decision by a major power to launch nukes may be one such situation.

but you are implying a significant value judgment on the continuity of the world. im not saying that is wrong, but it isnt necessay either. suppose your ultimate value is existing human life. you estimate that a major war in a non nuclear world would kill 1 billion of us. a nuclear war would kill all 7 billion. if you assess the liklihood of the former at 50% and the latter at 5%, it isnt irrational to prefer the latter world.

[The Brain]

Surely the biggest cost would be all those smug preppers.

[The Brain]

Hey DG, this reminds me I have a question for you! In your profession, such as it is, how do you value risk at different points in time? For instance, one death now vs one death a hundred years from now? What's the system you use?

[Martinus]

Hey DG, this reminds me I have a question for you! In your profession, such as it is, how do you value risk at different points in time? For instance, one death now vs one death a hundred years from now? What's the system you use?

I will be interested in seeing DG's response but life insurance is quite different from other forms of insurance (that's why most countries expect them to be offered by separated insurance companie) because death is a certain future event so they work more like a savings/investment vehicle than classic insurance.

At its basics, you calculate life expectancy of a person and then make sure that at the end of the day they are more likely to put in more than they will receive at the end. 

[DGuller]

Hey DG, this reminds me I have a question for you! In your profession, such as it is, how do you value risk at different points in time? For instance, one death now vs one death a hundred years from now? What's the system you use?

I'm a property and casualty actuary, not a life actuary, so I only know the basic theory.  Since to a life insurer the risks of someone's death are financial, you use discount rates.  You can get a lot more investment income from a 20 year old's life insurance premium than you can get from an 80 year old's premium, so you price accordingly.  Of course, there must be a lot more to it, but I wouldn't know about it without working on it.

[The Brain]

Hey DG, this reminds me I have a question for you! In your profession, such as it is, how do you value risk at different points in time? For instance, one death now vs one death a hundred years from now? What's the system you use?

I'm a property and casualty actuary, not a life actuary, so I only know the basic theory.  Since to a life insurer the risks of someone's death are financial, you use discount rates.  You can get a lot more investment income from a 20 year old's life insurance premium than you can get from an 80 year old's premium, so you price accordingly.  Of course, there must be a lot more to it, but I wouldn't know about it without working on it.

What I'm interested in is how organizations assess risks at different points in time. Do they put the same weight on one death a hundred years from now as on one death today? For instance in a safety analysis.

[Martinus]

I am not sure if DGuller would know that - it's an accounting question more than an actuarial one. The Solvency / Solvency II regulations probably address it in detail but generally you are assessing the risks of a policy portfolio in bulk, not on an individual basis.

[DGuller]

Hey DG, this reminds me I have a question for you! In your profession, such as it is, how do you value risk at different points in time? For instance, one death now vs one death a hundred years from now? What's the system you use?

I'm a property and casualty actuary, not a life actuary, so I only know the basic theory.  Since to a life insurer the risks of someone's death are financial, you use discount rates.  You can get a lot more investment income from a 20 year old's life insurance premium than you can get from an 80 year old's premium, so you price accordingly.  Of course, there must be a lot more to it, but I wouldn't know about it without working on it.

What I'm interested in is how organizations assess risks at different points in time. Do they put the same weight on one death a hundred years from now as on one death today? For instance in a safety analysis.

That's as far as it goes for actuaries.  We don't things like cost-benefit analysis of safety initiatives.

[The Brain]

Hey DG, this reminds me I have a question for you! In your profession, such as it is, how do you value risk at different points in time? For instance, one death now vs one death a hundred years from now? What's the system you use?

I'm a property and casualty actuary, not a life actuary, so I only know the basic theory.  Since to a life insurer the risks of someone's death are financial, you use discount rates.  You can get a lot more investment income from a 20 year old's life insurance premium than you can get from an 80 year old's premium, so you price accordingly.  Of course, there must be a lot more to it, but I wouldn't know about it without working on it.

What I'm interested in is how organizations assess risks at different points in time. Do they put the same weight on one death a hundred years from now as on one death today? For instance in a safety analysis.

That's as far as it goes for actuaries.  We don't things like cost-benefit analysis of safety initiatives.

So you're basically useless?