Speed = Death?

Angst over AI is widespread, at least among early career white collar workers. Others pooh-pooh the worries, noting that lots of jobs have gone away in the past and this is no different. Oddly, a relevant example exists in the deep biological history of Earth.

At one point about 40 years ago, many paleontologists made a strong case that changing climates had no impact on extinctions. So in the Cenozoic, the end of the Paleocene was not at a big inflection in climate, which continued to warm well into the Eocene. The cooling from the Eocene through the Oligocene seemed just a part of a longer cooling that would lead eventually to the Ice Ages. And in a sense, these paleontologists were right: it wasn’t these long gradual shifts in climate driving species to extinction. But they were also wrong, because at least in a number of the extinction episodes that define the epochs and periods of the geologic record, it really was the climate. It was just that what the climate was doing was hidden by the blurriness of the geologic record as first read.

The most dramatic example is what we now call the Paleocene-Eocene thermal maximum (PETM), which was originally not identified to be on that boundary between Eocene and Paleocene. It took high-resolution examination of oceanic sediments and very carefully work with terrestrial sedimentary rocks to discern a very rapid change in the Earth’s climate. And that very rapid change–most clearly evident as a rapid increase in temperature–led to a notable number of extinctions, in fact the extinctions that had caused paleontologists to define a boundary where a number of species went away [that is where those funny names originated].

The harder you look, the more it seems that it isn’t absolute levels of change, it is the rate of change that causes species to go extinct. The bolide impact at the end of the Cretaceous would have been such a rapid change that it would have been a surprise if it hadn’t caused a mass extinction. [It is worth noting that a lot of other impacts have not had remotely as profound an effect, probably a combination of smaller size and a chemically less damaging target].

Now we usually bring this up in talking about climate change, because the rapidity of climate change today will drive extinctions even beyond the ones caused by us hunting or paving or polluting the world. But the same logic applies to just how extinction of jobs is likely to evolve. Back when electricity led to the illumination of the world with light bulbs, the penetration of electricity was slow enough that those engaged in creating whale oil for lamps still had jobs for many years. Similarly the blacksmiths of the late 19th century were slowly replaced by both railroads and automobiles. These changes took decades, and many participants had skills that could migrate to other jobs.

When you look more recently, you can get a sense of trouble. The furniture and fabric manufacturers in the southeastern US were pretty successful for a long time. GG recalls a family trip where we stopped somewhere in the Carolinas to buy cheap sheets that were seconds from the local manufacturer. But when China was accorded access to the international trade system, these US companies were wiped out in short order. Both because of the concentration of specialists in this smaller area and the rapid influx of cheaper items from overseas, a lot of folks lost jobs in short order, and many of these towns are still struggling to recover. [You might think coal a similar story, but there the big story is the mechanization of coal mining far more than the replacement of coal by gas and renewables].

So what? We now have extraordinary claims that large swaths of office work will be run by AI in the next 12-15 months. Now, maybe that won’t happen, but if it does, it will mean a massive reorganization of work in very short order, and that is the recipe for extinction of a sort. Roughly half of all workers are in white-collar jobs. If this occurs over the next couple of decades, it will barely make a ripple as people move through different parts of the system in their careers. But if we are talking a year or two when entire career paths are eliminated…well, the dinosaurs wish you well…

Let’s go dark…

GG seems to have found all the really depressing stuff lately, so you know, let’s indulge. How does the world end? What are the odds?

• 85%. Planet is burned into a cinder and demolished as the Sun expands a few billion years from now.
• 14%. Planet get roasted but survives Sun’s expansion to spin endlessly around a dead star until the energy death of the universe or a black hole wanders by
• 1%. Planet gets hit by a rogue extrasolar planet and is smashed to smithereens
• 0.0001% Planet eaten by the Doomsday Machine
• 0.000001% Planet demolished for Vogon hyperspace bypass

Now I hear you say, well, that is pretty far off (unless there are Vogons, hyperspace bypasses, and such). This is a geological view of things; humanity is pretty incapable of destroying the planet and might be incapable of destroying the biosphere. So what is so dark?

Well, let’s ponder how humanity will exit stage left, a more personal end, with some timescale and odds…

• 30% within next decade. The creation of true artificial intelligence destroys everything.
• 20% within next 50 years. Nuclear war.
• 20% within next 150 years. Environmental collapse.
• 15% within next 10,000 years. Evolve to a new species.
• 5% global plague
• 5% right-handed amino acid-based life (“mirror life”) consumes everything.
• 5% within next 1000 years. Hand off Earth to our robotic progeny.
• 0.1% population collapse to zero
• 0.01% within next 100 years. Enable ape takeover of Earth.
• 0.00001% ever. Zombie apocalypse.

We’re leaving out various religious ends-of-times as these seem to have frequent recalibrations. Our top three candidates are entirely self-inflicted. As such, these odds can be changed dramatically. So why so dark?

Old Fogey or Inbred?

Another year has passed and so the Grumpy Geophysicist risks toddling more and more into the realm of the Old Fogey. So with the new year, a post that stalled in limbo is completed…

One of the curious aspects of science is that dueling camps develop that seem to ossify. Often the two sides keep firing the same kind of observations at each other with little impact. Examples include the pro-plume and anti-plume camps (the latter most energetically defended by Gillian Foulger and the late Don Anderson), the Baja-British Columbia vs. small offset groups addressing the Cretaceous evolution of western Canada (mostly paleomag purveyors versus structural geologists). In many ways, an earlier example was the fixist versus mobilist debates of the middle twentieth century. So how much of these kinds of stagnant debates is simply resistance to change, how much is a result of echo chamber action, how much is genuine ambiguity, and how are these stalemates broken? And how do you avoid being an Old Fogey?

What’s it all about?

Ah, the end of the year. When we look back and wonder if we achieved what we hoped to achieve, appreciated that which is good, etc. We all have our personal ranking of things, whether it be family or friends or success of some kind of another. But what of civilization? Is there anything positive about civilization that makes it worthwhile as more than life support for several billion humans?

Certainly not warfare to gain territory. If civilization includes pretty much all human cultures, then warfare is, at most, a zero sum game. So while we often learn the names of conquerers from ages ago, this hardly seems to be a civilization-advancing behavior.

Eroding Confidence

When a student, GG felt that erosion was such a dominant force in earth science. You’d go to the Grand Canyon and be told, all this created in six million years (about 0.1% of Earth’s whole history). You might hear of the boo-boo that threatened to fill the Imperial Valley (and did create the Salton Sea); the redirected Colorado River started incising in soft valley fill, producing a waterfall that migrated upstream nearly a mile a day. Lava flows that dammed the Colorado in the Grand Canyon for tens of kilometers were wiped away in fairly short order.

But then you look at incision rates of the Colorado in the same places as the lava dams and it seems surprisingly low. Then there is low-temperature thermochronology interpreted to mean that the lower canyon has been sitting there for tens of millions of years with little change. Measurements of erosion rates on high Sierran surfaces are minuscule (~0.01 mm/yr, or 10m/Ma), similar to other bedrock erosion rates on mountain ranges, suggests that erosion is pretty feeble over large areas.

So let’s go on an expedition into erosion…with a few numbers and a lot of speculation.

Looking back at the Anthropocene

Hi, future geologists. Maybe you’ve stumbled across some million-year old writings that dismissed the idea of an Anthropocene Epoch. Yeah, you’re puzzled because that is a pretty solid division of time from your perspective…of course you also got rid of this Holocene/Pleistocene boundary too, because, really? The Holocene is hardly more than a small burp within the Pleistocene, at least that fragment not in the Anthropocene.

Those old writings focus on whether the line for the Anthropocene was when the first megafauna perished in Europe or Australia or North America. Or when forests were cleared for farming. Or when Roman smelters first led lead to waft into the Arctic. Or when carbon isotopes got really strange, or elements that should have been off the edge of the periodic table showed up. Or when calcite started dissolving at shallower depths in the ocean. From your end of the temporal telescope, this is all rounding error stuff. After all, the extinction event at this time was one of the biggest ever. Sure it took a few years; most of these sorts of things do, so those old arguments seem rather trivial disagreements.

Of course, the arguments in these old writings is about the Anthropocene Epoch, but you have the Anthropocene Era, a higher level distinction. Or maybe you prefer Anthropozoic, as after all, the big extinction events bounded the Mesozoic on each end and it seems these Eras like to be “-zoic”s. Of course your name is likely different. After all, the Cenozoic, the “Age of Mammals” is not the “Age of Chicxulub” or “Age of Deccan Traps”, so probably this Era shouldn’t be the “Age of Humans”…but from our end, it is hard to know who the winners will be.

So kind of just was well we didn’t adopt that Anthropocene title.

The Earth Transformed: Book Review

Reading this book took GG back to his grad student days. At one point GG decided that there was an important story in the geology of the Sierra Nevada, and so he read everything about that geology (both easier and harder than now, as the literature is now larger but the access so much easier). He then cobbled it all together in a manuscript that he gave to his advisor to read. And said advisor drowned the manuscript in red ink, but the main point was to say, how did all this combine to produce some new insight? While many advisors would have simply shrugged and allowed such a thesis chapter to go forward as an unpublishable demonstration of having read a lot, GG’s advisor saw some utility and pointed out ways to take all this reading and make a scientific paper out of it.

Having plowed through Peter Frankopan’s encyclopedic The Earth Transformed: An Untold Story, GG is very much reminded of his red-ink-soaked draft from years ago. Frankopan in essence is recovering all human history in slightly under 700 pages. With all footnotes banished to the web (what would be a normal addition of 212 pages), GG is impressed that Frankopan is described as a best-selling author (he is suspicious that the metric is pounds of books more than numbers of books, but even so, far beyond GG’s reach). Now online reviews laud this as scholarly and timely and vast. It is all those things. GG would add, it is also in need of both an editor and a real sense of how to develop an argument.

“How the Mountains Grew” book review

Hmm..well maybe review is not quite right, as GG will be a bit petty at times here. To be fair, this book would be better described by its subtitle “A New Geological History of North America”. Mountain building is not actually a focus–it isn’t until page 250 that there is any attempt to explain how mountains are built–instead this is more like a popular geologic history that could well be used in a general historical geology course. Much like those texts, this book starts at the beginning and moves forward through time. A large part of those first 250 pages has to do with evolution of life and some speculations on changes to the earth’s climate and oxidation state. Author John Dvorak often finds a spot to describe as an introduction to a broader theme. The text is a pretty easy read, though part of that might be that GG knows the material pretty well.

It does differ from a textbook in trying to capture some of the more recent ideas in the literature. And there is an interesting dance around the K-Pg extinction, with a chapter on the Yucatan impact and then another on the Deccan Traps, in the end preferring a repeated role for volcanic causes of other extinctions. So there is a little editorializing going on, that should be expected in a book like this. And there is a warning of sorts buried on page 298: “Never travel with more than one geologist at a time. Otherwise you will be hopelessly confused.” [You almost wonder if that thought was picked up from the approach of John McPhee, who conveyed geology through profiles of geologists]. So the clarity of the text is only possible because the author only listened to one voice?

So what of the mountain building?

Geology is Mostly Disasters

Well, in some ways. If you recall the old argument between the catastrophists and uniformitarianists, the geologic record was either mainly special events or the result of the daily plodding of earth processes dutifully stacking up sediments to become rocks. Usually this conflict is described as ending with the catastrophists being run off the stage and “the present is the key to the past” becoming engraved on the hide of every geology PhD ever since. (Wikipedia’s version is “Since the 19th-century disputes between catastrophists and uniformitarians, a more inclusive and integrated view of geologic events has developed, in which the scientific consensus accepts that some catastrophic events occurred in the geologic past, but regards these as explicable as extreme examples of natural processes which can occur”).

But stand back and ask, just what qualifies as a catastrophe? Consider, for instance, floods in California. A recent New York Times Magazine article describes efforts by some to get dam managers in California to try to plan for really incredible flooding. And repeatedly those efforts encounter incredulity: the floods they are asked to anticipate are beyond their experience. Frequently dam managers would claim such a flood is unrealistic. The response has been, look back to 1862, when the Central Valley reverted to a lake. Would that one year’s flood constitute a catastrophic event or just a typical one? Most people would look at such a flood and call it a catastrophe, but most geologists would be content to just say that was part of the usual runnings of the climate. But how much of the rock record comes from these more exceptional events?

Finding Importance

In the memorial symposium for Peter Molnar, Phillip England suggested that Peters brilliance in many ways was recognizing the next important problem to address. And Phillip asked, how do you know? Is it internal, that you just know what is important? Do other people tell you? What criteria might you use? One possibility is that in opening up a line of research, many other follow you in and the research that follows is long and fruitful. In Peter’s case, it was (initially) continental tectonics. But just how you identify important problems is, though, itself a knotty problem.

Consider two problems: earthquake prediction and paleoelevation. Earthquake prediction has been the subject of decades of research, some of it very productive (being able to recover the history of earthquakes on many faults) and much of it not (as in, virtually every prediction actually put forward). Almost nobody would say that prediction is unimportant–provided it is successful. But how about if it isn’t possible? Is an insoluble problem an important one?