What is your field/career?

[Huinesoron]

I don't think "geologist who cried volcano" is a problem, is it?  The Philippines mass evacuated in 1991 for Pinatubo very successfully in response to the warnings of geologists -- that was almost Krakatoa-sized, but killed <1000.  Nevado del Ruiz killed >20,000 in 1985, and was only lethal because the Colombian government screwed up by ignoring geologists, and that's widely believed in Colombia.  And Mt. St Helens in 1980 is also pretty well known for killing people who ignored warnings, enough to have some penetrance in popular culture here.  (RIP, other Harry Truman.)

And I get the impression the real huge potential catastrophes are the ones that aren't expected at all -- the obvious example is the 2004 Indian Ocean earthquake/tsunami.  Other possibilities include:

• We haven't been hit by a serious solar flare since 1859.  One hitting the Earth would cause a serious catastrophe; that would screw around with modern technology something fierce.
  
• Los Angeles can actually be hit by a hurricane -- southern California got hit by a large one in 1858.  That would be incredibly nasty since they're instead prepared for an earthquake.
  
• Boston is vulnerable to a good-sized earthquake, and some think it's overdue.  It probably wouldn't be more than magnitude ~6.5; that may cause no deaths at all where I live, but in Boston quite possibly thousands -- it has a lot of brick buildings.
  

The biggest natural disaster to affect Britain in recent times was The Icelandic Volcano (Eyjafjallajökull, which is indeed the first Google hit for 'icelandic volcano') back in 2010. It grounded flights across Europe for a week, including every commercial plane in Britain, France, Germany... and as far as I can tell, that effect was completely unexpected. Apparently it was partially responsible for a mass closure of travel companies in the UK, which goes back to the economic effects idea.

I dunno, I had a volcano-related comment and just shoved it out there; I don't think I thought this post through.

[rho21]

For the purposes of learning the basic concepts of programming, you can do a lot worse than a bunch of games that require the same sort of skills. Spacechem, for example, has you thinking about complex algorithms and program flow, while Human Resource Machine provides a grounding in low-level programming concepts like memory and string manipulation. In both cases, once you've solved the problem, it will tell you how much better you could have done and challenge you to optimise your solution in various ways.

As far as maths goes, I can't say I've ever needed all that much of it to program, except when implementing specific bits of maths. On the other hand I do have a degree in maths, so perhaps I just don't notice when I'm using it. By far the most important thing anyway is creating algorithms, which is more an extension of the logic you learn through studying maths than the maths itself. Given that you like the logical side already, I suspect you'll be just fine.

[Dreylin]

Oh, I didn't mean to imply those were in any way causally related; what I meant is "these apparently happen more often than what modern society is accustomed to".

Except ... if that were the case, there should exist more "the world is screwed" eruptions between Baekdu blowing up in 946, Rinjani blowing up in 1257, and Huaynaputina blowing up in 1600.  Apart from that, I'm not sure, and would scientists/historians necessarily even know?

Yes, you can know. If a Vulcanologist can tell you that Yellowstone (sorry darrell) erupted 2.1 million, 1.3 million, and 630,000 years ago, they can certainly tell you whether it erupted 2,000yrs ago. I expect that most "recently" active volcanoes have had their history investigated, and most large scale eruptions of the type you are talking about will leave clear indications in the topography even if they have not been active in "Human history".

I get the impression that anything overdue in a geological timescale isn't that big a deal, because most are still unlikely to hit in a timeframe less than that of, say, how long writing has existed (~5000 years), which is still short geologically.

Actually most would be a huge, massive, enormous, catastrophic deal if they occurred; but as you say the discrepancy in the timescales involved causes us to dismiss them.

Of course Houston has had two <1,000yr flooding events in the last year - which is generally irrelevant apart from flagging how humans are bad at understanding probabilities.

I don't think "geologist who cried volcano" is a problem, is it?  The Philippines mass evacuated in 1991 for Pinatubo very successfully in response to the warnings of geologists -- that was almost Krakatoa-sized, but killed <1000.  Nevado del Ruiz killed >20,000 in 1985, and was only lethal because the Colombian government screwed up by ignoring geologists, and that's widely believed in Colombia.  And Mt. St Helens in 1980 is also pretty well known for killing people who ignored warnings, enough to have some penetrance in popular culture here.  (RIP, other Harry Truman.)

I do think it could be a problem, yes. Not counting the science deniers and the "I grew up here and damned if I'm not going to die here stubbornness", and certainly not in the Hollywood movie kind of way. Not all warning signs will lead to a major eruption, so all you need is a fizzle and people will have that doubt sowed the next time. I mean St. Helens was ~2months between the first signs and the eruption; that was a rural area so was relative simple to keep an exclusion zone, but what if nothing had happened after those 2months, what if the "warning signs went on for 4months, or 8 and nothing happened - now what if instead of St. Helens it was Vesuvius and you had evacuated a large number of people out of the potential risk zone? How long before people decided that they weren't willing to put their lives on hold indefinitely and moved back despite being told it was still a risk?

I mean if people can criticise Hurricane forecasters for suggesting that Irma was more likely to go up the East coast of Florida and then it actually went up the West....

[darrelljs]

a Vulcanologist can tell you that Yellowstone (sorry darrell) erupted 2.1 million, 1.3 million, and 630,000 years ago

You've killed us all !



Darrell

[darrelljs]

My two recurring nightmares are gators eating me and Yellowstone roasting me.

Darrell

[Mardoc]

Of course Houston has had two <1,000yr flooding events in the last year - which is generally irrelevant apart from flagging how humans are bad at understanding probabilities.

So, Houston has a land area of 10,062 mi², and Earth has an area of 196.9 million mi² (per Google). That means 20,000 Houston-sized areas on the planet. We'd expect somewhere in the world to have two 150-year events back to back every single year, if you're just looking at flooding and not all the other ways things can go extreme.

Also, it seems quite possible that we don't really have enough data on Houston to honestly say what a 1,000 year event is. I doubt we have more than 150 years of rainfall records.

volcanoes

So, I'm curious: It seems possible in theory that we could defuse volcanoes, perhaps by extreme levels of geothermal energy extraction, perhaps by controlled venting/drilling - basically any method that would remove the energy at a controlled rate instead of waiting for an uncontrolled release. Do you have any idea how possible that would be in practice?

[Dreylin]

Also, it seems quite possible that we don't really have enough data on Houston to honestly say what a 1,000 year event is.  I doubt we have more than 150 years of rainfall records.

Well I also don't think those "1,000yr flood" maps have been updated in the last 20yrs and so don't reflect the current situation. But yeah, that was an ironic reference rather than anything to suuport the discussion.

So, I'm curious: It seems possible in theory that we could defuse volcanoes, perhaps by extreme levels of geothermal energy extraction, perhaps by controlled venting/drilling - basically any method that would remove the energy at a controlled rate instead of waiting for an uncontrolled release.  Do you have any idea how possible that would be in practice?

It's been ~20yrs since I studied this in earnest, so I can safely say I have no idea. But my feeling is that the sheer quantity of energy that you'd need to siphon off to have a noticeable effect over a human timescale is outside the range of our current technologies.

[rho21]

So, Houston has a land area of 10,062 mi², and Earth has an area of 196.9 million mi² (per Google).  That means 20,000 Houston-sized areas on the planet.  We'd expect somewhere in the world to have two 150-year events back to back every single year, if you're just looking at flooding and not all the other ways things can go extreme.

That's the Earth's total surface area rather than its land area (57.5 million). It's kind of hard to tell what effect flooding has on the ocean.

Even so, you make the excellent point that we should expect 150-year events (or even two of them in the same place) to happen somewhere on Earth pretty frequently. I think the world would be a much better place if more people had this level of understanding of statistics.

However there is something of a reliance on the assumption that *only* Houston has suffered this fate this year. I think far more data is required to produce a useful quantitative argument here.

[Mardoc]

That's the Earth's total surface area rather than its land area (57.5 million). It's kind of hard to tell what effect flooding has on the ocean.

Dang it, I thought adding the word 'land' to my search would have gotten me the right number.  My google-fu is weak  

I think far more data is required to produce a useful quantitative argument here.

Honestly, this was sort of what I intended my point to be (apparently not made well).  People are way too confident about way too many subjects, compared to the actual data we have.

If AI puts us all out of work, maybe it'll hire us back to accurately estimate the risk of Yellowstone's explosion and then defuse it .

[Brian Shanahan]

Mm, I certainly wasn't suggesting them, only a possible outcome in the case that unemployment becomes unbearable (my point of reference was the French Revolution, which was partially inspired by high bread prices). If such a case doesn't occur, then for sure, we won't need to make any changes.

Fair enough.  I'll agree with the response to the hypothetical.

Anyway, I meant to respond to Adrien as well.  I don't want a UBI - but only because I believe that the world is such that we can't afford to pay it and have people stop working.  In the hypothetical situation where that changes, where automation really does produce so much that work becomes irrelevant, I would hope we'd land with a UBI instead of ensuring make-work.

Why? If the world's governments weren't so fixated on an economic system that gives the vast majority of the fruits of labour to the owners of capital, despite the minimal labour they put towards making those profits, we'd have well more than enough to be able to afford a decent standard of living for everyone. When 50% of the world's wealth is owned by a number of people so small they wouldn't populate a small village, a proper UBI is more than possible, given the proper will.