Some Statistical Thoughts on the End of the World

Back in high school I remember Mr. Morrel telling us about the pastor William Miller who after a careful reading of the bible forecasted the world would end by the year 1844. Back then I thought it was absolutely ridiculous that a man in the 1800s could believe he was living in the end of history. It just seemed like it was in the such distant past. It seems less far away now that I'm older. Paradoxically as historical events drift further in the past they often loom larger in my mind. When I first learned about the halocaust it felt like it was done by people who were very fundamentally not like us in the present day. I don't feel that way anymore.

My teacher's main point with Miller was that people have been predicting the end of the world since the beginning of time. Every major modern day religion has something to say about the end of the world. Their predictions may be vague and allegorical but what is clear is that they are predicting the end of the world. Already something of a bold prediction from past data. With all these unborn out predictions my teacher posited that one shouldn't take them too seriously. And there is a sort of statistical sense to it. If someone correctly predicted the end of the world tomorrow, still only one out of thousands of end of the world forecasts were true.

Here's a statistical counterpoint, a Jake original. Right now about seven percent of everyone who has ever lived is alive. This is sort of an amazing fact to think about. When I think of all the times I could have been born it seems vanishingly unlikely that I would be born at the capital P, Present day. Or rather that the year 2019 would be my present day. But if I believe the world is ending in my lifetime. Then in fact its not that unlikely. It's about 7%. Its also not an accident of my time that roughly 7% of everyone is alive. Populations grow exponentially. This is in some sense the definition of exponential growth: The area under an exponential curve up to a given point is directly proportional to its height at that point. Now of course human populations don't really grow exponentially. Not all populations have the same fertility rate. There are wars and plagues. But at most points in the past about 7% of people up to that point were still alive.

So now if you believe human populations will grow roughly exponentially up until the end of the world then you should believe there is a roughly 7% chance you will be alive at the end of the world. Which is a dramatically more troubling figure than the one you'd arrive at by considering the preponderance of failed predictions.

Of course this is all bullshit. Who knows what will happen. These kind of statistical thoughts based on projected population at the end of the world or the number of people who have been wrong in the past seem like not the best way to understand the future. I have three visions of the future.

• The world essentially ends in the next 50 years. We're already past peak oil and peak production of a number of other important reagents in the global economy. This plus food scarcities from climate degradations lead to interstate conflicts which lead to nuclear war. That or the AIs kill us. Or synthetic plagues created from published virus genomes. Or an asteroid.
• The next 50 years look a lot like the last and nothing terribly bad or good happens.
• Amazing advances are made in longevity, space travel and AI. We live in a well organized, peaceful and civilized society.

When I write them all out they all sort of seem vaguely plausible and I don't know how to live my life in the uncertainty. Somehow the middle option seems unlikely. From my perspective we have big problems as a species and without big solutions we won't solve them. But maybe all the disasters I worry about aren't really big deals. And maybe all the technologies I'd like to be optimistic about just don't pan out.