Is the end of the world really at hand?

[Rach3]

It's The End Of The World!

There have been billions of humans in this world. Most of them are alive today; six billion, whereas in 1960 there were half that, and even fewer before then. Less than one in a million were among the first thousand, as an example. We modern humans are very common, I suppose about half of all humans ever born are alive today. So it's no mystery that we in particular are alive right now. But it would be truly surprising, if we were among the first thousands - it would be like winning a lottery!

However, if there were many humans to be born in the future, then there'd be more of them (all added together) then of us and all our predecesors. In which case, there'd be relatively few of us. Say the human population goes up to 10 trillion and stays there for a hundred millenia; then less than 0.0001% of all humans would be born by now! It'd be incredibly unlikely for us to be here now!

Since this is so extremely unlikely, it is very probably true that the end of the world is at hand.

 

[Rach3]

(I actually made the title all caps - IT'S THE END OF THE WORLD! - apparently the forum software automatically changes that. Way to stifle vital news about the future of humanity, vBulletin. :grumpy:)

 

[Ivan Seeking]

If the end of the world is at hand, then there will be no ten trillion and we are in the majority.

However unlikely it may be that we are here, we are here. I think you are trying to argue that our existence makes our existence less likely...:uhh:

 

[Rach3]

If the end of the world is at hand, then there will be no ten trillion and we are in the majority.

Exactly! And that makes us very likely. However, the world lasting much longer makes us unlikely. Since we are here, the proposition that we are likely is greatly favored - it is more likely. So it is much more likely that the world is at an end.

However unlikely it may be that we are here, we are here.

But it's more likely that we're likely to be here, than that we're unlikely to be here. Really, how likely is it that we're extremely unlikely?! Not very, I'd say.

I think you are trying to argue that our existence makes our existence less likely...:uhh:

Um, no, nothing of the sort. Our future existence would make us so unlikely, it would be improbable! So our existence makes a soon demise very likely.

(This argument does not apply to past, small generations, because we're not them. They were too unlikely to matter. In other words - even if they did use this argument, and it were wrong, it wouldn't matter, because the vast majority of people using this argument are here now, and will be right. So the argument is right more often than wrong, when invoked. In particular, it's incredibly unlikely that we're wrong.)

 

[Rach3]

Redux: there are two exclusive hypotheses, A and B. A says X is likely, B says X is unlikely. Thus, given X, it would be rather peculiar that B (which thinks X is unlikely), no? So X makes A likely.

A = end of world is at hand
B = end of world is not at hand
X = we are here

 

[Tomaz Kristan]

So called Doomsday argument.

http://www.anthropic-principle.com/primer1.html

http://en.wikipedia.org/wiki/Doomsday_argument

 

[Gelsamel Epsilon]

I never understood the Doomsday arguments logic. How does our position on on the graph of how many humans will ever exist determine how long it is till extinction? Especially when the total humans who will ever exist is just speculation.

Edit: See what Rach3 said, I have no idea how the hell that works. Wouldn't A say X is less likely? I'm pretty damn sure that if it's the end of the world then we're probably dead - and hence not here...

 

[russ_watters]

However, if there were many humans to be born in the future, then there'd be more of them (all added together) then of us and all our predecesors. In which case, there'd be relatively few of us. Say the human population goes up to 10 trillion and stays there for a hundred millenia; then less than 0.0001% of all humans would be born by now! It'd be incredibly unlikely for us to be here now!

Since this is so extremely unlikely, it is very probably true that the end of the world is at hand.

This would only be true if the growth rate were constant and it isn't - it is decreasing.

And it doesn't have anything at all to do if the end of the world is coming. It is just a misuse of statistics.

 

[Office_Shredder]

I'm sure humans 1,000 years ago thought the same thing

 

[Rach3]

I'm sure humans 1,000 years ago thought the same thing

And they were all wrong! But there were very few of them, and there are very many of us, so on average, people who use that argument are right. Because the ones who are wrong, there are fewer of, and they weight the average less.

 

[russ_watters]

And they were all wrong! But there were very few of them, and there are very many of us, so on average, people who use that argument are right. Because the ones who are wrong, there are fewer of, and they weight the average less.

But 1,000 years ago, they were exactly as "right" as you are now...

 

[Rach3]

But 1,000 years ago, they were exactly as "right" as you are now...

Yes, but there were too few of them to affect the averages.

 

[lo2]

We will make it of course we will!

 

[slugcountry]

Yes, but there were too few of them to affect the averages.

lmfao genious

 

[russ_watters]

Yes, but there were too few of them to affect the averages.

No, as far as they knew, they affected the averages then exactly as we affect them now.

The wik link gives a pretty good explanation of several flaws, but the most basic is the one I am highlighting: for any series of repeated events (such as human births), there is a 95% probability that you are not in the first 5%. But attach that to a hyperbolic curve with an end point and you get a near certainty that you are near the end of the curve if you use hindsight only - regardless of where you actually are on the curve.

edit: from the wik link

The probabilistic argument used in the Doomsday Argument can easily be applied to anything else; therefore, with 95% certainty, ANY event that has occurred a certain number of times has used up at least 5% of its existence, which is one twentieth. Any event that has occurred N times, then, will (probably) only occur 19N more (or even less).

This ratio stays constant, and therefore the estimate for how long an event will last will keep going up; if the event DID occur the 19N more times and then stop, as the DA advocate will predict, then the DA advocate will not say "Alright, it's probably over now." Instead, a new N will be calculated, and a new estimate 19N formed, despite the fact that things seem to be following their old estimate.

 

[Rach3]

How much does it matter when we are right now? If an omniscient third person looks at the whole timeline of humanity, most humans will be around at the very end of it (under typical models of growth, assume this much to keep the argument interesting). So a random human, on being born, should think to him/herself "hmm... I've just been born, so it's pretty likely I'm within a few generations of the end of the world. I might just happen to be one of the few early ones, but that would be truly improbable."

To rephrase things: let's say you have a number in an envelope - a real number between 0 and 1. That's all you know about it. You could bet that number is greater than 0.000001, just like you could bet that anyone lottery tick is a losing one. Now, that number in the envelope, tells you your position in the timeline of history; if you have very little information about the future (which is reasonable, we don't), then it's fair to say, a priori, that you belong to the very large and probable set, of people within a few generations of the end of the world.

 

[Rach3]

Isn't Bayesian statistics fun?

 

[russ_watters]

How much does it matter when we are right now? If an omniscient third person looks at the whole timeline of humanity, most humans will be around at the very end of it (under typical models of growth, assume this much to keep the argument interesting).

Yes...

So a random human, on being born, should think to him/herself "hmm... I've just been born, so it's pretty likely I'm within a few generations of the end of the world. I might just happen to be one of the few early ones, but that would be truly improbable."

Right - and the next human born would think exactly the same thing...and the next one the same thing...and the next one the same thing...

And at birth, each would be equally correct. Of course the next guy knows he's more correct than the last, but he also assumes the next guy after him won't be more correct than him and he's wrong.

To rephrase things: let's say you have a number in an envelope - a real number between 0 and 1. That's all you know about it. You could bet that number is greater than 0.000001, just like you could bet that anyone lottery tick is a losing one. Now, that number in the envelope, tells you your position in the timeline of history; if you have very little information about the future (which is reasonable, we don't), then it's fair to say, a priori, that you belong to the very large and probable set, of people within a few generations of the end of the world.

Correct. But how do you know the number is between 0 and 1?

 

[Rach3]

But how do you know the number is between 0 and 1?

There will be a finite number N of humans in all. The nth human is assigned the real number n/N; this is between 0 and 1. Then the numbers are evenly distributed over the unit interval.

 

[Andre]

How about balancing biomass? If the average person has a mass of 50kg, 20% is dry weight and 50% of that is carbon, then he would contain 5 kg carbon. So 10E+9 persons would have a carbon mass of 0,05 GtC. But at the top of the food chain, suppose we need some three orders of magnitude more biomass for that. That's around 50 GtC. We can manage that, that's about 10 years of fossil fuel consumption. No problem.

But it's also 200 Gt water. How much would the sea level be lowering with so many people?

 

[russ_watters]

There will be a finite number N of humans in all. The nth human is assigned the real number n/N; this is between 0 and 1. Then the numbers are evenly distributed over the unit interval.

But n and N are calculated based on how many humans are alive now, how many were alive in the past, and conjecture about how many will be alive in the future, are they not?

 

[Rach3]

But n and N are calculated based on how many humans are alive now, how many were alive in the past, and conjecture about how many will be alive in the future, are they not?

No, we don't know what N is, just that it exists and is finite.

 

[russ_watters]

No, we don't know what N is, just that it exists and is finite.

Right. So we have to assume what it is when we plug it into our calculations.

 

[russ_watters]

Here - let's use the calculation from the Wik link:

If we assume that 60 billion humans have been born so far (Leslie's figure) then we can say with 95% confidence that the total number of humans, N, will be less than 20·60 = 1200 billion. Of course, this would only be right if our assumption was 100% sure about that we are in the last 95% of humans to ever be born, but since it is just an assumption, the total number of humans, N, and further calculations based on it are just assumptions, too.

Assuming that the world population stabilizes at 10 billion and a life expectancy of 80 years, one can calculate how long it will take for the remaining 1140 billion humans to be born. The argument predicts, with 95% "confidence", that humanity will disappear within 9120 years. Depending on your projection of world population in the forthcoming centuries, your estimates might vary, but the main point of the argument is that we are likely to disappear rather soon.

That's 1140/(10/80)=9120 years.

But suppose this were the year 1800, with the population just below 1 billion and we did the calculation assuming the population would level off at 1 billion and still an average lifespan of 80 years. If 6 billion humans had lived before, the total becomes 120 billion and those left to be born, 114 billion.

That's 114/(1/80) = 9120 years.

The point is, anyone who has ever done the calculation using the same assumptions about their position as you do gets exactly the same answer.

In fact, if you updated your calculations every day based on that day's population and used enough significant figures, you'd find that the answer is the same every day despite the fact that we should be a day closer to our doom. The inevitable conclusion after 100 days is that you've been wrong 99 out of 100 times (if you always assume you are correct, you'll at least, always be right once :uhh: ), you should conclude that there is a 99% chance you are wrong today. You could plug that back into the calculations then...