5 ms·

Silly question: If I'm understanding this correctly, doesn't this essentially mean that statistically it's possible to live forever? Or is there a point when

by j_baker 15y ago

Silly question: If I'm understanding this correctly, doesn't this essentially mean that statistically it's possible to live forever? Or is there a point when you statistically have 100% probability of dying?

jbri 15y ago
Sort of, depending on the model. For most models, if you take your "probability of dying in year X", and sum that over all the years from 0 to infinity, you'll get a 100% probability of dying at some point. The interesting thing is that there's no individual year with a 100% probability of dying - the certainty of death is just because "forever" is a really long time. There are hypothetical distributions, though, where the sum total of probabilities is less than 100% - where some proportion of the population will, statistically, never die. Of course this brings us to the real issue, which is that what the model says doesn't really matter - if the model disagrees with reality in extreme cases, reality wins.
- usaar333 15y ago
  On high ends it definitely is. It predicts no one could have ever lived past 116, but clearly some have: http://en.wikipedia.org/wiki/Oldest_people http://en.wikipedia.org/wiki/Oldest_people
tikhonj 15y ago
In practice, a sufficiently low probability is indistinguishable from 0.
- Sharlin 15y ago
  Yeah: http://lesswrong.com/lw/mp/0_and_1_are_not_probabilities/ http://lesswrong.com/lw/mp/0_and_1_are_not_probabilities/
Jimmie 15y ago
Think of it like entropy. Statistically a broken egg could place itself back together but that's never going to happen.
- danvet 15y ago
  Entropy _is_ statistics ... it's just another measure for the probability of a given state.
redthrowaway 15y ago
Yeah, it doesn't quite account for the hard limits. Even by his equation, there should be a few 140 year olds strutting around.
- usaar333 15y ago
  No there shouldn't. The odds of living to even 130 under his equation are 1 in 4.8*10^47. That's just a ...tad.. more people than have walked this planet.
usaar333 15y ago
Well, sure. Except the equation gives a 2 in 7 billion chance of even making it to 114 (note that in reality your odds are empirically better). The odds of reaching 136 are 1 in 10^86. For comparison, there are only 10^80 atoms in the universe...
Sharlin 15y ago
This is just a model - a map of reality which may give some helpful predictions and insight about the territory, but is not the same thing as the territory (reality) itself. See, for instance, http://wiki.lesswrong.com/wiki/The_map_is_not_the_territory http://wiki.lesswrong.com/wiki/The_map_is_not_the_territory
- j_baker 15y ago
  Sure. I'm just curious how out of sync the map is from the territory.

by j_baker 15y ago

jbri 15y ago

Sort of, depending on the model. For most models, if you take your "probability of dying in year X", and sum that over all the years from 0 to infinity, you'll get a 100% probability of dying at some point. The interesting thing is that there's no individual year with a 100% probability of dying - the certainty of death is just because "forever" is a really long time. There are hypothetical distributions, though, where the sum total of probabilities is less than 100% - where some proportion of the population will, statistically, never die. Of course this brings us to the real issue, which is that what the model says doesn't really matter - if the model disagrees with reality in extreme cases, reality wins.

usaar333 15y ago
On high ends it definitely is. It predicts no one could have ever lived past 116, but clearly some have: http://en.wikipedia.org/wiki/Oldest_people http://en.wikipedia.org/wiki/Oldest_people

tikhonj 15y ago

In practice, a sufficiently low probability is indistinguishable from 0.

Sharlin 15y ago
Yeah: http://lesswrong.com/lw/mp/0_and_1_are_not_probabilities/ http://lesswrong.com/lw/mp/0_and_1_are_not_probabilities/

Jimmie 15y ago

Think of it like entropy. Statistically a broken egg could place itself back together but that's never going to happen.

danvet 15y ago
Entropy _is_ statistics ... it's just another measure for the probability of a given state.

redthrowaway 15y ago

Yeah, it doesn't quite account for the hard limits. Even by his equation, there should be a few 140 year olds strutting around.

usaar333 15y ago
No there shouldn't. The odds of living to even 130 under his equation are 1 in 4.8*10^47. That's just a ...tad.. more people than have walked this planet.

usaar333 15y ago

Well, sure. Except the equation gives a 2 in 7 billion chance of even making it to 114 (note that in reality your odds are empirically better). The odds of reaching 136 are 1 in 10^86. For comparison, there are only 10^80 atoms in the universe...

Sharlin 15y ago

This is just a model - a map of reality which may give some helpful predictions and insight about the territory, but is not the same thing as the territory (reality) itself. See, for instance, http://wiki.lesswrong.com/wiki/The_map_is_not_the_territory http://wiki.lesswrong.com/wiki/The_map_is_not_the_territory

j_baker 15y ago
Sure. I'm just curious how out of sync the map is from the territory.