Ask HN: What other low-probability high-impact events are we ignoring?
The gulf oil spill is a canonical example of a "low-probability, high-impact event." Everyone who cared knew that this sort of thing could happen, but the risk was widely perceived as tolerable.
What other high-impact scenarios are out there that are likely to create a meaningful "event" at least once in the next 30 years? How are we currently mis-pricing the risk?
[Edit for clarity. Low-probability, in this instance, means risks that are low when taken in discrete units. Getting on an airplane is low-risk. With 14M scheduled US flight per year, though, it is almost certain that one plane will crash once in awhile. We're looking for events here that have a perceived low risk (correctly or not), and that are likely to surprise people when they actually happen. In hindsight, it will be clear to us that the event was actually likely given the extended circumstances.]
86 comments
[ 2.8 ms ] story [ 162 ms ] threadIsn't that kind of like asking "What's an unlikely event that's likely to happen"?
Think serious car accidents: unlikely enough that most people never have one, yet many thousands happen each year. The effects can be devastating to the people involved, but everyone keeps telling themselves that "it won't happen to me", and many people forego actions that would decrease or mitigate their risk, or even engage in actions that increas it substantially.
On a scale of things unlikely enough that most people never have one a car accident is pretty near the bottom.
Even now, while it is unlikely, it is very possible for it to happen. It is even possible for it to happen through the democratic process (in fact, that may be more likely with the current makeup of the West). People tend to turn to absolutes when they become desperate and it is quite forseeable that some combination of economic, military, political, or natural disasters will make some country so desperate they will once again try anything.
http://en.wikipedia.org/wiki/It_Cant_Happen_Here
Edit: corrected link, sorry.
Edit 2: Or...not. Apparently I can't have a URL with an apostrophe.
:)
For god's sake. The difference between a major Northern California earthquake and the effing Singularity is that one of these has happened in the last century and is guaranteed to happen again.
Lay in an emergency supply of drinking water and strap your heavy shelves to the wall.
Another very real possibility is that the Cascadia zone could have a megathrust earthquake. This would be a 9.0 hitting cities like Seattle, Victoria and Vancouver that are completely not prepared for it. The last one in 1700 gave Japan a really nice tsunami as well.
That could also potentially trigger the lahar that is waiting on Mt Rainier. Which brings to mind all of the volcanos listed at http://en.wikipedia.org/wiki/Decade_Volcano - the eruption of any of which would qualify as low probability high impact event.
With such a flagrant history of safety violations, I'm surprised it didn't happen sooner.
Other events that were or will be likely to occur, but have a low perception of risk, include the recent financial crisis, and, looking forward a decade or two, ocean level rise flooding major cities, and a US debt default.
Some interesting events that (seem to) have a genuinely low probability, but would really change the world if they happened:
* Asteroid hit (either a big one, or a smaller one that hits just right).
* A 'supervolcano' eruption that puts us into nuclear winter
* Solar flares fry every piece of electronics on the planet
* Sudden tipping point in the Antarctic or Greenland ice sheets that raises ocean levels by a few meters over a year.
* A strong 'godlike' AI (benevolent or otherwise)
* A self-replicating nanoassembler
* Godzilla attack
I would hope the structural differences between the US situation in 1945 and now would be obvious - exporter vs importer, young vs old, short term debt (the war) vs long term (SS and Medicare, plus maintaining an ongoing worldwide military presence). And we don't currently have the advantage that every possible economic competitor of ours was recently carpet bombed and utterly drained by years of total war, which certainly didn't hurt our exports.
Adding for clarity: I don't currently believe either a default or major coastal flooding is inevitable or unstoppable, I just think there is substantially more risk associated with it than is generally perceived. Ironically, the perception of little risk can help make such events more likely, because after all there is little point wasting lots of effort trying to prevent something that probably won't happen, right?
The nastier example was in 1933 when the USA devalued the US dollar against gold by 40% and unilaterally rendered null and void all contract terms where the USA had agreed to pay back debt in the borrower's choice of dollars or gold. Technically not quite a default, but it was a pretty subtle distinction for lenders who got paid back what was effectively 60 cents on the promised dollar.
The last time that the government came close to failing to pay as promised was in 1995. Under Newt Gingrich, Congress refused to do any business, it brought government to a standstill and the government was unable to make payments. Luckily the deadlock was resolved just before the government would have been forced to not make bond payments.
When it comes to the current debt, I would again not expect an outright default. But I'd expect another fudge where some owed debt does not get repaid. And the prime candidate for how that will be done is that the US government will restructure Social Security so that the government won't have to pay back bonds owed to Social Security. (Bush tried to do this, but didn't manage to push it through.)
Here is that story. In past years Social Security has brought in more money than it has paid out, and the difference was invested in government bonds. Now that the government is facing Social Security paying out more that it takes in, the government is supposed to pay those bonds back but doesn't want to. Structurally, on paper, Social Security is fine for decades to come. However that depends on the US government paying back a lot of this owed debt, which nobody in government wants to do. Hence all of the talk about Social Security being broken. And the one feature that any "fix" is guaranteed to have is that the US government won't be paying back all of the money that was "lent" to the government from Social Security funds.
Cosma Shalizi illustrates this problem nicely here: http://cscs.umich.edu/~crshalizi/weblog/543.html
http://www.michaelmoore.com/words/latest-news/africas-oil-sp...
http://www.independent.co.uk/news/world/africa/niger-delta-b...
It very true that we operate with a "out of sight, out of mind" mentality, particularly here in the western world. Oil companies don't deal with many of the spill scenarios they create well enough simply because they aren't being held accountable. Social responsibility should be a requirement for corporations.
But do I really care? I drive a small SUV.
(answer: of course I still care!)
Let's suppose the event is an accident occurring during some industrial process. This process is going to happen very frequently over the next 30 years. While the probability of an accident during any particular instance of the process may be low, the probability of the accident happening at least once during that period may be quite high.
Some math:
p -- accident probability
n -- # of times the process happens in the period under consideration
Probability of accident happening at least once: 1 - (1 - p)^n
For example, suppose the process fails 0.01% of the time (1/10000) and happens twice a year for 30 years. Then an accident happens 1 - (0.9999)^60 ~= 0.6% of the time. If the failure rate is 1%, the probability of at least one failure jumps to 1 - (0.99)^60 ~= 45%.
http://www.earthmountainview.com/yellowstone/yellowstone.htm
W T F?!
There are billions of things that could happen, that might happen or that will eventually happen given enough time. We should reasonably ignore those or it starts to get silly :)
I think what you are really talking about are "Elephant in the Room" problems. Things which shouldn't happen, or are low probability, but end up being caused by bad management or human error.
If someone had asked this question 6 months ago would anyone have brought up deepwater oil drilling? I somehow doubt it. Finding these scenarios is difficult, if not impossible, because they rely on things that don't get reported till it has happened.
That being said, here a few high-impact (moderately low probability) events that could be on the horizon.
* Northern California Earthquake * Israel v Iran Nuclear Conflict that triggers WW3 * Singularity
Drastic life extension would count, as would 'a mature nanotech', matter edition on any scale,
The record-breaking deliberate oil spill happened during the Gulf War, in 1991.
In the past thirty-five years, there have been ten oil spills in which over a million barrels of crude were released. I don’t see how you can classify this kind of thing as “low-probability”, in the same category as asteroid strikes.
http://en.wikipedia.org/wiki/Oil_spills#Largest_oil_spills
Like the bee example, it just seems like something I dismiss because I can't imagine it actually happening.
reference: http://www.cbsnews.com/stories/2006/11/02/health/webmd/main2...
The cause is that CO2 in the atmosphere becomes carbonic acid after it dissolves in water, and if the ph of the seas changes even a little bit, then calcium carbonate shells (which is what everything from clam shells to coral is made out of) becomes unstable and starts to dissolve.
According to the best current projections this is now hard to stop given current CO2 levels, and a major die-off is likely around the middle of this century.
The effects of this on global ecosystems are hard to estimate, but will definitely be large.
And, humans are resourceful. It's likely that, even in the worst of scenarios, at least some humans will find a way to survive -- even if it's subsistence living.
The question is, would it be an environment that we would want to live in?
I think that's an important distinction that might help sort out some of the ecological "debates".
I wonder, how do we preserve our science across such gaps in ability to sustain it. Should we even bother?
If something doesn't spell our doom either in the short term or in the moderately-short term then we will quickly recover from where we left off. A generation or two of subsistence living will not entirely wipe out human civilization. No one is going to have to reinvent the wheel, because we've got so many present in our civilization that they could easily be scavenged for hundreds of years before the quantity became depleted and it would likely take a good thousand years before someone couldn't find an 'original wheel' to at least take inspiration from. Aluminium alloy wheels have such high proportions of aluminium in them that even the exceptionally slow oxidization (I work in siding and I can tell you that 20 year old aluminium siding barely has marks on it and it's paper thick, a few pounds would take millennia to degrade but would likely still be clearly recognizable).
If we're really concerned about our longevity I'd suggest manufacturing information arcs that can teach the major language of the geographical area allowing people 100 years down the road to access the knowledge stored when civilization recovers. It would have to be built for longevity and nothing else (IE it can't easily be recovered for materials by scavengers or used to draw power). Essentially you'd want a very long life RTG to power a multiple redundancy computer system and basically load encyclopaedias and patent databases and an archive of literature, magazines and journals.
What concerns me is not the transmitting of basic machines across the simplification of a society, to use Joseph Tainter's jargon, but, rather, difficult, contextual knowledge. Our knowledge of Mathematics, for example, was preserved by luck: the Byzantine Empire harbored enough learned individuals and archivists until such a time as the Crusades swept through. All of my work simply assumes an industrial, technological civilization with advanced computational capacity. Any textbooks I might produce in the future will have the implicit assumption of contemporary operating systems, compilers and hardware. If, through sheer chance, any of my potential works survived across a collapse they would be useless without further context. I imagine that the same holds true for advanced works of physics, especially that which requires hugely complex experimental devices.
Euclid held up well enough as his works were self-contained. What modern works can say the same?
???
This is supposed to be comforting, how? Try this paraphrase: we're twiddling the knobs on the ecosystem, we know we're likely to make big changes, but we're not sure what. Would you like operators of a nuclear plant to follow this behavior?
It's more like, if you found yourself in a universe full of potentially dangerous forces that you don't understand, would you try to harness them or just hide from them? So far, the human race has traditionally gone with the former. Some of the things we've done have had great positive effects. Some have been catastrophic. Most have been relatively inconsequential. The result is never clear beforehand — that's the whole idea behind experimentation and the scientific method. Otherwise you're left with fearmongering and superstition.
Inconsequential? Then you haven't been paying attention to history. We've had freakin huge impacts on our ecosystem throughout our history, even predating recorded history. The disappearance of huge tracts of North American forest. The same for British and European forests. The Colorado River dwindling into nothing. Virtual eradication of Bison. Eradication of the Passenger pigeon. Aborigines causing the unintentional desertification of much of Australia. The current desertification of Africa. Virtual disappearance of Salmon from the North American west coast. None of these have been inconsequential. Some have caused tremendous benefits to people, but all have had significant negative impact.
The result is never clear beforehand — that's the whole idea behind experimentation and the scientific method. Otherwise you're left with fearmongering and superstition.
Many scientists refrain from experimenting on their own bodies. This has nothing to do with "fearmongering and superstition." When scientists do experiment on themselves, they are generally certain that it won't have permanent consequences. Again, nothing to do with "fearmongering and superstition." You categorically state that we won't know what happens.
And nobody's talking about doing experimentation on their own body here, so I don't see how that's relevant, unless you're saying that all uncertain actions (experimenting with electricity, nuclear power, etc.) are like experimenting with your body and should not be done.
We're doing significant things to ecosystems. The magnitude of what we do gets larger as our technological capabilities in crease. We're still not close to understanding ecosystems well enough to ensure we won't do anything bad.
I think that what's going on is awareness.
There was a much larger deep-water spill in Mexico in 1979. There have been huge on-shore spills, some that ran into the sea. There have been huge tanker spills. And so on.
The difference is that this one is off the US coast, we have on-scene video, and folks who claimed to be uber-competent have, arguably, dropped the ball.
Yes, they've probably gotten more blame than they deserved, but they've also tried to exploit this incident for political gain, so ....
1. California earthquake
2. global economic depression
3. war involving developed countries(N Korea vs S Korea, India vs Pakistan, Iran vs Israel, China vs Taiwan, etc)
4. 2 leading into 3(maybe)
5. pandemic
6. The Next Big Thing(biotech? nanotech?)
7. car accident resulting in serious injuries involving yourself or a loved one
8. serious illness involving yourself or a loved one
It is worth noting that India and Pakistan have fought several recent wars, and are now both nuclear armed. Worse yet, everyone's wargame simulations have concluded that their next war WILL go nuclear. As one friend of mine (whose father is high up in India's air force) put it, "The only real question is whether the bomb that Pakistan drops on New Delhi works."
To be specific, the likely scenario is that they go to war, Pakistan, lacking reserves, knows it has to win fast. Pakistan throws everything into the battle front, India breaks through somewhere, and Pakistan, with nothing to stop India's breakthrough, drops a nuke on Indian troops in Pakistani territory. (There is a perverse logic to having all of the civilian casualties be your own.) India, having just lost major numbers of troops, has little choice but to escalate to nuclear war against Pakistani military targets. Pakistan, with a more limited supply of nukes, has only one way to escalate. And New Delhi is the easiest to reach major city.
There is speculation that the last military coup in Pakistan happened in part because the Pakistani army was scared that Pakistan was escalating into another war with India, and it was the easiest way to force the situation to de-escalate.
The trick to that statement being that we don't know where the baseline is...at .00001% likeliness this year, an exponential increase of 1.0001x every year isn't particularly likely.
Aliens arriving, being discovered, or merely communicating with us, there are so many possibilities, many of them with the potential for being as disruptive to human civilization on Earth as an asteroid impact. The mind boggles at the possibilities.
- Israel bombing all suspected Iranian nuclear facilities. Iran retaliating with a nuke.
- Low level or dirty bomb by any number of terrorists groups in the world. India/Pakistan is a pressure cooker right now. Could happen in Europe, the Mid-East or America.
You don't have to go further back than the 1918 global flu pandemic to see an example of what a global pandemic can do. 3% of the global population died.
You may think that we have better vaccines and better systems in place now than we did 100 years ago, but you'd be largely mistaken[1]. A vaccine takes so long to develop and produce that when it's ready for the market millions of people will already have died. Pandemics spread quickly. Adding to the problem is that the world is so connected and people so mobile that once the virus is out it could easily spread across the globe in a matter of days. One infected stewardess in Heathrow airport wil lmake sure of that.
[1] I'm no expert on biotech so please correct me if I'm mistaken.
Of greater importance is a disease that there is a vaccine for but which the U.S. government is actively suppressing: smallpox. DNA synthesizers are already at the point where anybody who wants to can resurrect smallpox. It is only a matter of time before that scourge is set loose. The results will be ghastly.
Odds are there is barely any.
EDIT: Apparently the U.S. has rather a lot stockpiled, but good luck getting any in advance of the end of the world.
As far as I can remember, most of the flu last year was of the swine variety and was spreading human to human. We were lucky it wasn't so deadly as it might have been. It spread far and wide though.