Please remember to return your library books before then.
October 10, 2017 7:12 PM   Subscribe

 
Next Tuesday at 9:30 Pacific? Shit. I really want to skip my commute that day.
posted by Abehammerb Lincoln at 7:22 PM on October 10 [1 favorite]


Thank goodness it's always NEXT tuesday.
posted by sammyo at 7:26 PM on October 10 [5 favorites]


Oh, Bayes.
posted by Apocryphon at 7:26 PM on October 10 [3 favorites]


Good analysis, in general. But I do think there is reason to believe we are living in a privileged time in human history.
posted by So You're Saying These Are Pants? at 7:26 PM on October 10 [4 favorites]


That was a surprisingly good read. I was going to make a joke, but instead I encourage everyone to read this.
posted by Literaryhero at 7:26 PM on October 10 [4 favorites]


Reminds me a bit of xkcd What If guesstimation, although as far as I can tell it is based on different kinds of principles.

It's okay, we just have to engrave the sum of human knowledge on durable materials, Exhalation-style.
posted by inconstant at 7:28 PM on October 10 [1 favorite]


But I do think there is reason to believe we are living in a privileged time in human history.

A four way simulpost?

More seriously, it is just ego to think that we are living in a special time. What makes now more important than any other time?
posted by Literaryhero at 7:28 PM on October 10 [1 favorite]


Nuclear weapons, global warming, and twitter?
posted by So You're Saying These Are Pants? at 7:29 PM on October 10 [5 favorites]


People don't normally think like this? I mean, There's the conversation about death with kids like 6 threads down... if you are unwilling to extrapolate out the implications of that to human existence... well... sorry that this comes as a shock... I mean, 5000 years more would put us at a pretty short run, but I mean 8 mil as a high bound? That's also a pretty shit run comparative to a lot of mollusks and dinosaurs.
posted by Nanukthedog at 7:29 PM on October 10


Oh god, it's not loading on my phone. It's already started, hasn't it. We'll be extinct by tomorrow morning. Am I right?
posted by potrzebie at 7:33 PM on October 10 [16 favorites]


That's quite fascinating, thank you for sharing
posted by smoke at 7:34 PM on October 10 [1 favorite]


Lucky escape for Arsenal, then?
posted by TheWhiteSkull at 7:37 PM on October 10 [28 favorites]


I look forward to his tackling the really big questions, such as whether Star Trek: Discovery gets six seasons and a movie. (Of course I know that most of the other series got seven, but I'm trying to adjust my expectations.)
posted by Halloween Jack at 7:39 PM on October 10 [4 favorites]


Since when does "wild-ass guess" mean "pretty good idea"?

We're talking about 3-4 orders of magnitude, after all.
posted by tclark at 7:41 PM on October 10 [3 favorites]




This sounds like the same reasoning by which it can be 'proven' that we're living in a simulation: there's presumably one real universe, but potentially a huge number of simulated universes, so the odds we're in the real one are extremely low. It's internally consistent, but still doesn't seem convincing.
posted by echo target at 7:56 PM on October 10 [2 favorites]


Sometime in the next 10,000 years
A comet's gonna wipe out all trace of man
I'm banking on it coming
Before my end of year exam
posted by incster at 7:56 PM on October 10 [7 favorites]


Gott has put his Copernican formula to the test in a number of different ways over the years, with some surprising results. For starters, on the day Gott published his Nature paper in 1993 there were 44 plays currently running on and off Broadway in New York. He applied the Copernican formula to each of those plays to derive an estimate of how much longer they'd run.

As of 2016, 42 of those plays had closed within the time frames he predicted. Two more remain open. “I could even be wrong about those two and still get at least 95 percent right,” Gott notes in “Welcome to the Universe.”


Wait, he explicitly set up his formula to capture 95% confidence, and found that 42 of 44 plays, or 95%, closed within the interval he selected. That seems like the opposite of surprising.
posted by Existential Dread at 8:04 PM on October 10 [3 favorites]


To save some clicks for those whose brains follow the same tracks as my own:

It turns out there are, in reality, thirteen Land Before Time movies.
posted by kaibutsu at 8:10 PM on October 10 [5 favorites]


This sounds like the same reasoning by which it can be 'proven' that we're living in a simulation

Yeah, it is related. One of the philosophers who has defended the "doomsday argument", Nick Bostrom, applied similar reasoning in his "simulation argument".
posted by thelonius at 8:12 PM on October 10 [3 favorites]


I have always found this argument deeply dubious. Bayes' Theorem is applicable to probability distributions. It's not obvious to me that the distribution on "fraction of the lifespan of the humanity that this experience is happening at" is well-defined --- and if it is, I don't see why it should be uniform, which seems to be the only way the numbers work out. (OK, strictly speaking it only requires that the CDF evaluated at 2.5% is 2.5% and at 97.5% is 97.5%) And in fact that distribution is definitely not uniform --- there are over twice as many experiences happening right now as were in 1960.

More generally: There's a problem that the probability of that prediction being made is not independent of the fraction of the lifespan of humanity we're looking at --- I'm much more likely to make that prediction that average person living in 5000 BCE, sheerly because I have mathematical tools that they don't. That means that the accuracy of the instances of the prediction being made, averaged over said lifetime, can't be 95%: In extreme, consider an 'alpha and omega point' species, who are only capable of making the prediction in the first and last 2.5% of their species lifespan - every single time they made such a prediction they'd be wrong. I'd even accept some attempt to fudge the effect of progress over time; shit just throw in a logistic curve with an inflection point around ~1900. But to ignore the endogeneity of the fact of the prediction being made is criminal. (I'm fundamentally invoking the anthropic principle here, except as pertains to statements instead of entities)

You can call this pedantry, but the only thing that distinguishes this from stoner dorm room bullshit is the claim to mathematical rigor. I call shenanigans on that claim. The Copernican Principle is a heuristic, not some mathematical law. A lot of times it takes you somewhere useful but it sure as shit doesn't prove anything on its own. And this is just goofiness. Which, fine, if it were presented explicitly as playing around. But it's not, it's basically treating numerology seriously.
posted by PMdixon at 8:19 PM on October 10 [27 favorites]


Shouldn't population growth factor into this somehow? Humans have been around for a long time, but it's only within the last century or so that our numbers have reached into the multiple billions. If you (pessimistically) assume that a dramatic spike in population would likely presage a global collapse, then wouldn't picking a random perspective in time be more likely to land you in those bustling twilight decades, rather than some prehistoric era with only a few million humans?
posted by Rhaomi at 8:22 PM on October 10 [6 favorites]


There’s a 95% chance that I will die between a year from now and 3900 years from now.
posted by chrchr at 8:22 PM on October 10 [5 favorites]


Accounting for population is hard, because we don't know the shape of the population distribution in the future. For example, if things flatten out (or even decrease a bit) and then hold steady for the rest of human existence until we get destroyed by a passing planetoid, it could be that we've got relatively low serial numbers...
posted by kaibutsu at 8:27 PM on October 10


Oh god, it's not loading on my phone. It's already started, hasn't it. We'll be extinct by tomorrow morning. Am I right?

ASK AGAIN LATER
posted by Greg_Ace at 8:39 PM on October 10 [9 favorites]


I think the answer to this question is fairly obvious, barring any major physics breakthroughs and we as a technological species make it relatively intact for the next hundred or two years. If so, the greatest likelihood is that the last human dies out when the sun goes red giant.

Maybe some humans could eke out an existence in the outer planets somewhere, but I think it would be hard to keep it going indefinitely without any terrestrial planets left.

The other option is that we never leave Earth and we all die of some plague or comet or massive geological upheaval or gamma ray burst or global thermonuclear war or whatever long before the age of the sun matters one bit.

Other possibilities seem remote to me.
posted by wierdo at 8:40 PM on October 10


Since when does "wild-ass guess" mean "pretty good idea"?

My own WAG, completely unsupported by any reasoning or probability, is "real soon now in geologic time, within just a few million years." Pretty much in line with the supposedly good idea. I say we take the average of all our guesses, pool our money, and bet on that.
posted by sfenders at 8:55 PM on October 10


The problem comes from his attempt at removing the "specialness" of ourselves from the universe. Or perhaps that's merely the article author's framing (Copernican, yada yada).

Yet, he says there are the beginnings and ends of things that are usually not observable so we give a broad likely 95% probable range. with 2.5% likelihood that we're at the beginning or end (am I stating the properly)?

In the example of the wedding, he says "Ah this is but a momentous occasion" - he has to define some sort of hierarchy of privilege and centrality to any given point. And at that point, we need to define, in fact, what IS the "beginning" and how do we determine for any given participant in a subunit of this framework how important that thing is in their relationship to it.

Like - ok, the wedding, fuck all if I care how important it is. Or even if I drove by it. Extrapolating this thing is, in itself privileging a sort of Anthropic principle. I mean, yes, I guess we ARE talking about the existence of the human race, here. But in the sense of the goal of decentralizing the importance of the human race to itself in relation to this statistically, it seems fundamentally bound up in... the definition of what importance is itself. And how that determines the boundaries and edges.

In that regards it's a bit astrological, I think. Or, even like my own fave, Terence McKenna's time wave. It's really nice and easy to create broad terms without precise definitions on just WHAT constitutes something important or relevant to the given frameworks, and further, these lack of boundaries allow a "bleeding over" so the edges will always be fuzzy.

Maybe there's something... probabilistically quantum in that, I dunno. I'm tired, and not high enough, at this point.

Maybe there's a 95% chance that between the start and end of what I've written there might be information valuable enough to extract. But then wouldn't that make it part of a new boundary? I guess that's what I'm trying to get at. But then I suppose that makes the point - each point is its own new horizon. And if so, privileging a specific event (Sure, the wedding was a culmination and "an event" but there were previous "events" leading two entities to meet, engage in some socially contractual behavior....

Christ we're verging on creating a universe from scratch here. Every moment and place in time and space contains within itself the seeds of the future unfolding and only in the process of those unfoldings and emergences does anything take place, and to privilege any given moment over another (to even say "this is the start, or this is the end" really destroys the point of destroying an anthropocentric world).
posted by symbioid at 9:49 PM on October 10


The short book “Paradoxes in Probability Theory” by William Eckhardt gives convincing refutations of this argument, the simulation argument, and similar games people play with probabilistic reasoning.
posted by vogon_poet at 12:38 AM on October 11 [12 favorites]


I also, like PMDixon, call bullshit on this.

Maybe the article is just badly written, but if we assume that Gott is a bona-fide scholar and did his math, then we still need to ask the question, what worth is in a prediction that is precisely 50% right? It's actually the worst possible prediction, even accounting for the fact that a mistake may have been made when defining the two possibilities themselves.

So from a completely useless prediction, Gott goes on to a "sharper" probability of 95%, which is actually compensated by widening the time window, or reducing information content of the prediction, i.e. making it even less valuable.

Similarly, I can say with a 99.9% probability that I will live from 0 to 15 million years. How useful is that for future planning? Even less than saying that with a 50% probability I will die sometime in the next 20 years.

The article concedes this when it links this weak premise to the zingy "we must colonise other planets", thereby taking for granted the sci-fi notions of terraforming and travel close to the speed of light, and for what? The jaded me looks for hidden interests behind every such message and the cry of 'we'll get off this planet' may just reinforce the public feeling that, 'meh, this one is already ruined anyway, so why bother'.
posted by Laotic at 1:15 AM on October 11 [3 favorites]


the definition of what importance is itself. And how that determines the boundaries and edges

Yeah this is where I see a possible mistake as well.

He got inspired by the Berlin Wall and used that as an example. It was built in his lifetime.

How about if he was a guy using the Great Wall of China similarly, back in ~1000BC, that had also been completed maybe 30 or 40 years prior? He would have come up with a similar range for its life. How do we know that in the Berlin example we are statistically at any old time in the process, while in the China example we are in that very lucky momentous startup point?

We have been around as a species for a few million years. But just this very moment we have gained all of this awareness - of our place in the universe, of the ability to travel to other planets, of the existential threat to our environment and possible solutions. So the current moment does not seem like "any old time", it seems like a very momentous nexus point - this is either the very end or the very beginning of us.
posted by Meatbomb at 1:29 AM on October 11 [4 favorites]


And not to abuse edit - it gives us a range of times heavily weighted to either "a few million years, ending now" or "indefinite and potentially at cosmic timescale".
posted by Meatbomb at 1:35 AM on October 11


Bender: I can't believe it! I'm gonna die!

Fry: How much time does he have left, Professor?

Professor: Between a minute, and a billion years.

Fry: [puts a hand on Bender's shoulder] Well, at least you can plan accordingly.
posted by Rhaomi at 2:06 AM on October 11 [18 favorites]


The jaded me looks for hidden interests behind every such message and the cry of 'we'll get off this planet' may just reinforce the public feeling that, 'meh, this one is already ruined anyway, so why bother'.

Or we could await, with shivering anticipation, the technological advances that the simulators have in store for us.
For example, I just developed bionic implants and anti-gravitiy devices for my faithful Anno 2205 population! They were really excited about that and multiplied like rabbits after that.
posted by sour cream at 2:29 AM on October 11 [1 favorite]


Gott believes that because our species' time on Earth is very likely to be finite, we should be doing everything we can to colonize nearby worlds — particularly Mars — to increase our odds of survival. Putting a permanent colony of humans on Mars would be an insurance policy against a civilization-ending catastrophe here at home, like an errant comet or an accidental outbreak of thermonuclear war.
It would be ironic if the race to colonize Mars led to a nuclear war that wiped out humanity - especially if it happened 5,100 years from now, making Gott's prediction true.
posted by clawsoon at 3:56 AM on October 11 [1 favorite]


it seems like a very momentous nexus point - this is either the very end or the very beginning of us.

Science and math are not magic, they need datapoints, at least three to have any kind of line/trend, we have one, us. Is the appearance in the local environment of an animal with self consciousness and the ability to make tools to make better tools significant? Significant in the long term geologic time frame? No way to know.

But as wacko as us space nuts sound an element is based on basic logic. Diversify in area and a single crisis can be survived. The planet is fine, the state of the planet could change badly for us, so learn to live in many environments. (and move all the harmful factories bad for the environment where it doesn't mater)
posted by sammyo at 3:57 AM on October 11


I liked the presentation of the concept in this article, simple as it might be. It's tricky to communicate mathematical ideas to a general audience, and I think the fact that nobody seems confused as to what the idea actually is is a sign of how well it was illustrated.

I was familiar with Gott's idea from a popular science book he wrote a few years back, and I'm still not sure how to feel about it. I appreciate the commentaries from other mathematicians that have been mentioned so far. Even if it doesn't apply to the end of the human race, the application to more mundane lifespans is neat (for example, websites like Metafilter -- who wants to work that one out for us?).

I guess one simple thing to note is that the longer something lasts, the longer you expect it to go on. So the 95% estimates aren't fixed, but constantly increasing as you move forward in time.
posted by rollick at 4:26 AM on October 11


Even if it doesn't apply to the end of the human race, the application to more mundane lifespans is neat (for example, websites like Metafilter -- who wants to work that one out for us?).

Metafilter will last somewhere between another ~5 months and ~700 years.
posted by clawsoon at 5:10 AM on October 11 [2 favorites]


It's a weird argument for colonizing other planets because each other planet we colonize would actually have the end of humanity sooner than here.
posted by miyabo at 5:22 AM on October 11 [5 favorites]


Gosh this is charming but given recent trends there's no WAY we're lasting that long
posted by fuzzy night at 5:58 AM on October 11 [2 favorites]


Interesting, but his whole approach could be flipped around to become Zeno's Arrow, and we never reach that end at all.
posted by Capt. Renault at 6:33 AM on October 11 [1 favorite]


Other possibilities seem remote to me.

We're already living in a remote possibility.
posted by Flexagon at 6:43 AM on October 11 [2 favorites]


Metafilter will last somewhere between another ~5 months and ~700 years.

So either way, Cortex needs to get to work on succession planning.
posted by notyou at 6:50 AM on October 11 [4 favorites]


I’m always amused that this type of article always supposes a binary: the human race will either vanish or succeed forever. The assumption is that we’re exempt from evolutionary pressures somehow. The human race also has the possibility to change gradually into something distinctly not human, after which point “humans” will have died out same as every other species on the Homo tree.

An interesting question is how or whether successive species would reckon with a past intelligent civilization with a detailed historical record.
posted by BuddhaInABucket at 7:50 AM on October 11 [9 favorites]


The argument assumes that the period of time in which we live is unexceptional. But along many different metrics, the 20th century saw changes that were vastly disproportionate to historical trends. Take population growth, for instance, or GDP, or world energy consumption. If someone in the year 1900 tried to use statistics to predict what any of those measures would be in the year 2000, their answers would be wildly wrong. Maybe none of that stuff matters in the long-view geological sense. Or maybe it does.

Let’s say humanity has been around for 200,000 years. For roughly 99.95% of that time, the only factors likely to cause the extinction of humanity were natural—asteroids, climate change, plagues, etc.—and their probability of occurring at any given time was more or less constant. But in the last 100 years, we’ve developed the capacity to potentially cause or prevent extinction-level events. That means that we can no longer necessarily assume our probability of going extinct at any given time is the same as it was 100, 1000, or 100,000 years ago, or for that matter, that it will remain constant in the future. I think we are in genuinely uncharted territory.
posted by dephlogisticated at 7:53 AM on October 11 [8 favorites]


I’m always amused that this type of article always supposes a binary: the human race will either vanish or succeed forever. The assumption is that we’re exempt from evolutionary pressures somehow. The human race also has the possibility to change gradually into something distinctly not human, after which point “humans” will have died out same as every other species on the Homo tree.

An even more likely third possibility is that civilizational collapse reduces drastically the number of humans, but doesn't make us extinct.
posted by thelonius at 7:54 AM on October 11 [5 favorites]


Correction: An earlier version of this story listed an incorrect date for the fall of the Berlin Wall

So delicious.
posted by chavenet at 8:53 AM on October 11 [3 favorites]


An even more likely third possibility is that civilizational collapse reduces drastically the number of humans, but doesn't make us extinct.

Skeletons and Skulls.
posted by lagomorphius at 10:09 AM on October 11


ASK AGAIN LATER

BETTER NOT TELL YOU NOW
posted by Philofacts at 10:49 AM on October 11


This sounds like the same reasoning by which it can be 'proven' that we're living in a simulation.

Don't worry, they shot that one down. It's impossible to model the physics of our universe, so we're not living in the Matrix.

on the other hand, I used to bullseye womp rats in my T-16 back home, and they're not much bigger than two meters.
posted by Naberius at 10:57 AM on October 11 [2 favorites]


All they proved was that it can't be a classical computer, it would have to be quantum in nature. Which everyone assumed anyway, because anything running a universe simulation with quantum mechanics wouldn't be doing it on our style of computers any more than we run our weather simulations by banging rocks together.
posted by Infracanophile at 11:11 AM on October 11 [2 favorites]


It's a weird argument for colonizing other planets because each other planet we colonize would actually have the end of humanity sooner than here.
Clearly the trick is to colonize other planets in such a way that at least some of their inhabitants are prevented from considering this question for at least several billion years. So long as there's always at least one frozen zygote left who hasn't yet read this article, the civilization they will come to inhabit is likely to survive forever.

On the other hand, if you destroy the records of when your colony was founded and assume that your colony is unlikely to be an outlier, perhaps you can simply claim that every colony is likely to be as old as the Earth.
posted by eotvos at 11:17 AM on October 11 [3 favorites]


Metafilter: But to ignore the endogeneity of the fact of the prediction being made is criminal.
posted by sammyo at 1:12 PM on October 11


This article is either right or wrong. Therefore, there is a 50% chance this article is wrong.
posted by speicus at 3:27 PM on October 11 [6 favorites]


Yes, but how long with the article last?
posted by Greg_Ace at 3:52 PM on October 11 [1 favorite]


Ugh, this theory came up in a sci fi novel I read as a kid and it's been mildly pissing me off every time I remember it ever since. Humans aren't souls wandering in the ether until we get randomly assigned a body. Sorry, I thought the article was well presented, but of course he's correctly predicted the ends of a bunch of contemporary things with his formula, the end point shoots off past the maximum lifespan of a human as soon as something's been around for a couple of years.
posted by lucidium at 4:10 PM on October 11 [4 favorites]


of course he's correctly predicted the ends of a bunch of contemporary things with his formula, the end point shoots off past the maximum lifespan of a human as soon as something's been around for a couple of years.

This is a really good point - if the model had predictive power the errors should be symmetric. I'd be pretty shocked if they were; I would expect a massive bias towards falling short of the predicted interval versus surpassing it
posted by PMdixon at 4:14 PM on October 11


Yes, but how long with the article last?

Aargh, *will
posted by Greg_Ace at 4:18 PM on October 11


On the other hand, if you destroy the records of when your colony was founded and assume that your colony is unlikely to be an outlier, perhaps you can simply claim that every colony is likely to be as old as the Earth.

As I recall, in one of Asimov's possible futures the Earth had long been forgotten as the original home of humanity, and there were various hypotheses as to where the original world was. Earth was known, but it was a mostly irradiated backwater and nobody seriously thought it could be the origin.
posted by Philofacts at 4:33 PM on October 11


The Foundation series.
posted by Greg_Ace at 5:11 PM on October 11


[calculates]

There's a 95% chance that I'll die sometime between a year from now and 1,500 years from now.
posted by clawsoon at 6:11 PM on October 11 [1 favorite]


[calculates]

There's a 95% chance that I'll die sometime between a year from now and 1,500 years from now.


My frequentist approach uses the percentage of days I've died to estimate my personal death rate.

I have to say things are looking good for me so far. And logically each day I wake up alive increases my estimate of how many days on this planet I have left.
posted by mark k at 7:55 PM on October 11 [6 favorites]


I’m always amused that this type of article always supposes a binary: the human race will either vanish or succeed forever. The assumption is that we’re exempt from evolutionary pressures somehow.

I don't get that either. Humans have been around for 200,000 years, but millions of years from now we'll still be humans?

But I don't really get being concerned about what happens millions of years from now anyway.
posted by bongo_x at 8:44 PM on October 11 [3 favorites]


I just happened upon this thread so I'm pretty sure this comment will be followed by as many comments after it as there are before it.
posted by chortly at 6:13 PM on October 15 [4 favorites]


« Older “SCARY!”   |   "Books from 1923 to 1941 Now Liberated!" Newer »


This thread has been archived and is closed to new comments