Tuesday Overnight Open Thread - August 26, 2025 [scampydog]

Would you look at that. It's Tuesday ONT time. Big science section this evening and some other random items. Pour yourself a tall water or adult beverage.

A bit of housekeeping before we get started. My sentence (or yours) has been completed. This ONT marks the end of our regular Tuesday night adventures. A hat tip to the Horde for enduring - whether reading or scrolling past with minor confusion. Thanks to bosses for not blocking the login and letting me get away with as much as I did. A special thanks to publius, Pete Bog – Bogs Rules!, Sock Monkey - Americana, Anonymous Rogue in Kalifonistan, and nurse ratched for their content and contributions. TRex, CBD, MisHum, and yes…Doof, cheers and thanks for the backchannel fun. It’s been a genuine honor to regularly contribute to what is, and has been the best place on the internet for a couple of decades.

Let's get started.

Who comes up with this wrongness? No way that Justified is only the 50th best TV series of all time.

Ranking the greatest TV shows of all time. pic.twitter.com/48UVVUWyyL

— Massimo (@Rainmaker1973) August 24, 2025

Science!

A quote from Feynman to kick off the science portion of the evening. “You have no responsibility to live up to what other people think you ought to accomplish. I have no responsibility to be like they expect me to be. It's their mistake, not my failing.” - Richard P. Feynman

***

Unleash the Kraken? Nope, but we are unleashing regular ONT commenter and AoSHQ contributor, publius. Whether this evening or later, highly recommended reading. Take it away, publius.

The Rotation of the Erf and its relation to the Significance of the Passage of Time:

You may have seen some stories that the earth’s rotation has been speeding up of late. Lots of click-bait articles have appeared (looking at you, Daily Mail) that sensationalize this. Of course, they don’t really know what they’re talking about. I got interested in this. Let’s unravel what’s going on. First, we need a brief history of time and its measurement. This is quite an interesting historical excursion.

From time immemorial, the notion of time has been intricately and almost inseparably linked to the motions of the earth and other celestial bodies. These cycles have been imprinted on just about every form of life on earth, plant, animal, and other. The main cycle is the diurnal one, the day/night cycle. The lunar cycle is also imprinted, through tidal action and the effect on certain marine life. And there's the cycle of seasons, the year, or more precisely the tropical year.

Life on Erf goes way back in time. The human species in modern form goes back maybe 180,000 years. However, modern human history goes back only five or six thousand years, a twinkling of the eye relative to astronomical and geological time scales. But from what we know of that history, our ancient ancestors were obsessed with observing and tracking the motions of the celestial bodies. Much superstition (think astrology as we know it today) abounded, but the basic reasons were practical, as they served as calendars and clocks, foretelling the seasonal changes, the time to sow, the time to reap, etc, etc. The ancient Egyptians, way back around 3000 BC and before were big on the heliacal rising of Sirius, and the ancient Greeks, c. 500 BC, were likewise big on the heliacal rising of the Pleiades. For the Egyptians this event heralded the annual flooding of the Nile. For the Greeks, the Pleiades rising signaled it was safe to sail the Aegean. In modern times, with our modern technology and all the time-keeping and navigational systems we have, the connection with the sky is lost. But to the ancients, observing the sky was a matter of life and death.

The main astronomical cycle is the day. 24 hours. To a first approximation, we think of that as the earth completing one revolution around its axis. This is not quite the case. One solar day is a little more than one revolution of the earth. Why? Because the earth moves around its orbit a bit in the meantime as the earth rotates. The earth must turn a little more than one revolution to line back up with the sun relative to a point on the surface. This is thus a "synodic period", known as a solar day. The actual rotational period, relative to (almost) inertial space, is known as the sidereal day. This difference resulted in the Gemini 5 mission splashing down 80 miles short of its intended location – someone not thinking naively programmed the earth’s rotation rate as 360 degrees per 24 hours, and not the correct sidereal value of ~361 degrees. The ancients knew this difference, although they had no understanding of a rotating earth or the heliocentric solar system. They thought the
sky rotated around the earth, and the “wanderers”, the planets, and the sun and moon moved in more complex ways against the background of the stars. Now, some heretics and crazy conspiracy theorists as far back as the 4th Century BC suggested the earth rotating could explain the motion of the stars, but this was rejected as crazy talk, mis- and disinformation and all that.

Thus, the synodic day, the diurnal cycle, is what has governed our notion of time of day. How did this business of dividing a day into 24 hours come about? It goes back as far as we know to the ancient Egyptians who first divided the day into 24 pieces, 12 for the day, and 12 for the night. That passed on through to the ancient Greeks and then the Romans. It was the Romans who first engrained this into mechanisms of civil society, religious, legal, and literary. The word "hour" and "horology" derive from the Latin "hora", which comes from a similar Greek word that meant roughly "period of time" or "season". The Woke lunatics who complain about rigid time schedules and demands of punctuality being "whiteness" can blame the Romans, I guess, if not the Greeks or Egyptians. They really need to blame Nature herself, as life on earth demands attention to the significance of the passage of time.

Now, where did minutes and seconds of time come from? That goes back to Ptolemy, in the 2nd Century AD, who formalized it, although he didn't invent it. He and his fellow Hellenistic astronomers adopted the Babylonian sexagesimal system. The hour, like a degree of angle, was divided into 60 smaller parts, which they called "mintua prima", the first small part, and then each of those was divided into 60 even smaller parts, dubbed "mintua secunda", or second small part. So, there you go, a second is just that, the second, smaller part of the hour.

And there we have it, the solar day is divided into 24 hours, which in term are divided into minutes and seconds, for a total of 86,400 second small parts per solar day. A second is thus defined as 1/86,400th of a solar day. But there’s a little problem there. The solar day is not constant, not uniform. Remember, the earth has to rotate a little more than one revolution to catch back up to the sun due to orbital motion. But, due to the eccentricity and the tilt of the earth’s axis relative to the orbital plane, this extra amount of rotation varies over the tropical year. Today, that varies from about 30 seconds longer to about 18 seconds less than 24 “uniform hours” over the course of the tropical year.

The ancients were aware of this as well, probably back to the Babylonians. Ptolemy and the boys were certainly aware of it and published crude tables to correct for this. But there was no understanding of the actual mechanism. Anyway, they came up with the concept of an average or mean motion of the sun. The difference between the actual position of the sun and this average position is called the Equation of Time and was used to correct sundials to “mean time”. This early understanding was crude, but they were aware of it that far back.

And that’s the way it was. Now, fast forward through the Middle Ages and to the Renaissance. There came Copernicus, Galileo, Kepler, and Newton, and the development of classical physics and understanding of the solar system and the mechanism of the motions of the celestial bodies. Also, we had the invention of mechanical clocks, which made using the “small parts”, minutes and seconds, practical. Mechanical clocks attempt to track “uniform time”, of course, and this made the Equation of Time notion must more apparent. Flamsteed (first Astronomer Royal) and Huygens worked out the Equation of Time precisely from Newtonian orbital mechanics and published accurate tables by the beginning of the 18th century.

And thus, the notion of “mean solar time” became theoretically defined and understood to high precision. This was very important for the development of precision navigation and the Longitude Problem and all that. But for the common man, it didn’t become that important until later. Clocks and watches were still mostly directly set by the sun, using apparent solar time to synchronize. It wasn’t until the concept of Standard Time and time zones came about that this became the standard for everyday use.

Thus, mean solar time was now the standard of timekeeping. The variable motion of the sun (more precisely, the earth’s orbital motion coupled with axial tilt) was averaged out. The length of the mean solar day and thus mean solar time depended only on the rotational speed of the earth. The earth was much more accurate over the long term than even the best mechanical clocks of the day. Precise astronomical observations were used to synchronize these clocks as needed. But then another little problem came up as astronomical observations and orbital mechanics became more and more accurate and precise. Based on Newtonian orbital mechanics, and calculations of the complex perturbations (look up the 3-body and N-body problem), the motions of the planets could be predicted to high precision. But they got to noticing some discrepancies there, which couldn’t be explained by the known perturbations. The moon, or Mars, or the Sun wouldn’t be exactly where it should be as a function of mean solar time.

They began to suspect the problem was with mean solar time, the earth as a clock. Maybe the earth’s rotation wasn’t as constant as we thought. This was strongly suspected in the late 18th century and was confirmed by the middle and late 19th centuries. The earth was not an absolutely accurate clock compared to the current precision of astronomical prediction and observation. Thus, the need for a truly uniform, or “Newtonian”, time scale was needed. It’s a bit hard to get your head wrapped around the concept, but the idea is to let the motions of the celestial bodies be your clock, and not the rotation of the earth. This is conceptually the same as that of a mechanical clock itself. We have some physical process, some motion involved, which by the laws of physics can be predicted as a precise function of time. The position of the hands on a clock are governed by this. The position of the hands tells us what time it is. And likewise, and to much higher accuracy, the positions of the planets, moon and Sun relative to the background stars can be the hands of clock that tell us what time it is.

In general, this is called a “dynamical time scale”. The astronomical version here became known as ephemeris time. This became well developed by the early 20th century, and allowed comparisons of mean solar time to this more accurate time scale. But soon thereafter, atomic clocks were invented, which were a game changer for timekeeping. The accuracy and precision became orders upon orders of magnitude greater than the most accurate mechanical clocks before. We now had a physical device to measure uniform time very precisely, far better than ephemeris time, down to the nanosecond.

The “second small part”, the second, had become our fundamental unit of time. We now got very precise in how we defined it. Previously, it was 1/86,400th of a mean solar day. With ephemeris time, it became defined as a certain fraction of the sidereal, and then tropical year. It was chosen to be as close as possible to the current solar day definition and was roughly about the length of one solar day second around 1850. Then with atomic clocks, it became defined in terms of the frequency of a particular transition of cesium atoms. IOW, it’s the period of a particular frequency of light, well microwave not visible light, waves. We now have an independent, uniform standard of time independent of astronomical processes. And thus, we can now measure to high accuracy what’s going on with the variable rotation of the earth.

Now, the accuracy of this, the estimates of the length of the mean solar day at any given time, depends on the accuracy of astronomical observations, which gets less and less as you can back in the past. Turns out, the accuracy is pretty good back to around 1830 or so, and with modern ephemerides, we have even more precise predictions of where things should be when they were observed. Timeanddate.com has a page with a graph of the mean length of the day, and hence the variation in the earth’s rotation here: https://www.timeanddate.com/time/earth-rotation.html
Since atomic clocks came into use, it’s possible to get much finer grained data and see the variation of the earth’s rotation on a day-to-day basis. IERS is the official source of all this, but a wiki article has a good graph in the link below of the LOD since 1962:

The green line is a yearly moving average which corresponds to the graph above, which we can call a “mean mean” solar day (one “mean” averaging out the sun, the other mean being the yearly earth rotation average). The gray line is the short-term variation in the length of day. You can see that over the course of year; the length of the day varies by around 2 – 2.5ms.

So, around 1870, the earth was spinning about 3 milliseconds (ms) faster than 24 hours of uniform SI time. And that’s a lot faster than even the speedup of today the Daily Mail and others were click-baiting about. But, by the early 1900s, it had slowed down to about 4 ms slower than 24 hours. So, it swung about 7 ms in 30 years. In another 30 years, it sped back up to around 0ms, then slowed down again. By 1972, when the UTC time scheme was put into place, the earth was about 3 ms slow. That’s the “root cause” of leap seconds, given the definition of the UTC scale. And then the earth started speeding again, with some bouncing around up until now. What’s notable is around 2020, the earth sped up enough that it was faster than 24 hours for the first time in nearly 100 years.

So, what is going on with this variation? A lot. A lot of geophysical and celestial processes are afoot there. You can see there are many cycles going on in that curve, and spectral analysis can reveal the frequencies involved. There are more than few periods there. Over the short term, the variation from day to day and month to month over the course of the year is driven by tidal friction and atmospheric and oceanic angular momentum exchanges. We’ve got fluids, air and water, sloshing around on this spinning ball. Angular momentum is exchanged between that and the solid earth below. Next are seasonal periods of the atmosphere and ocean over the course of a tropical year. Then we have periods longer than a year up to 30 years. Some of these are understood, while some are not so well understood. The 18.6 year lunar cycle comes in there for longer term tidal variation. Longer term climatic processes like El Nino and similar weather patterns are also in there. Core-mantle dynamics are in there as well, and this is one of the poorly understood ones. The 11-year solar cycle is also seen in there as well, not well understood either.

Now, what about the long term, the very long term, time scales of tens of thousands to millions of years? The main driver is the moon and tidal friction which is gradually and inexorably slowing the earth down (and, by conservation of angular momentum, causing the Moon to recede from the earth). Over geological time scales, this rate is variable, depending on how close the Moon (also the Sun, which is a secondary tidal driver, but the Moon is the main drive) is. It also depends on just how big the tidal bulges are, which depends on a lot of complex factors, including the rotation speed itself. Right now, (and this “moment” right now is probably 100,000 years or so, and certainly 10,000 years), the current rate of tidal slowing is about 2.3 ms per century. That is, absent all other factors, the length of the day should be increasing at 2.3 ms per century.

The curve on that timeanddate.com page graph goes back to 1830. It’s possible to go back even farther than that, but with less accuracy. 1623 is the limit, and below is an IERS page with yearly LOD values back to then:

Values over the 1700s are reasonably accurate, but the data before then is a bit sparse and scattered but nonetheless is not so bad as to be unusable. The first thing that jumps out is the earth seems to have been spinning pretty fast around 1623, 11 ms fast compared to 24-hour SI base time but slowed down pretty rapidly over the next 30 years. After that, things look more reasonable. Is that real, or just poor data? Well, it’s likely something real. The uncertainty there should be no more than 3 ms. And that time period is interesting. That was just a little before the Maunder Minimum, the little ice age. So, did the relatively fast spinning earth slowing down rapidly have any relation to that? Don’t know for sure, but it’s damned interesting.

And finally, remember the tidal slowing rate of 2.3 ms per century. While complex, that value is well known from current tidal dynamics. However, if we take that data since 1623 and calculate a linear trend line over that 400-year period, we find it was indeed slowing down, but only at 1.33 ms per century, not the 2.3 ms tidal rate. Either way, the earth was spinning about 6 to 8 ms faster than it should be relative to expected baseline in 1623.

But beyond that 1623 fast spin, something is counteracting the tidal braking at least over those 400 years. Estimates of this can go back much farther in geological time. The main driver of the difference there is thought to be GIA, global isostatic adjustment, or glacial rebound. The earth is still adjusting to the loss of the ice sheets from the last glacial period, those big ice sheets covering Canada and northern Europe and all that. The net effect, which is a gradual long-term process, is the earth becomes less oblate, and the moment of inertia decreases. This is sort of counterintuitive at first, but it has to do with how the mantle responds to the shifting weight of the ice sheets.

That’s likely one factor. Another is that elusive core-mantle coupling thing. And then something else might be going on, some unknown unknown. At any rate, over the last ~12,000 years ago, since the start of the Holocene, the long-term trend is about 1.7 ms per century. So, we’d expect the earth was spinning about 0.2 seconds faster in 10,000 BC than now – that doesn’t seem like much, but with the current definition of the UTC time scale, that would require a negative leap second every 4 days. Going back a lot farther, the day was about 23.5 hours (give or take a few minutes) about 65 million years ago, when the SMOD wiped out the dinosaurs. The dinosaur year would’ve been about 372.36 solar days per year. We could have 12 months of 31 days each and a leap year every 3 years (Feb. 32nd?). And finally, there is now good evidence that the earth’s rotation stalled at 19 hours for a period of 1 billion years, from about 1.8 billion years ago to 800 million. In a complex process, the solar atmospheric tide cancelled out the lunar tide. During this time the calendar would’ve required 461 days per year.

Big thank you, publius!

Cookie jar can be surprisingly noisy in the middle of the night.

😬 pic.twitter.com/rNlHCUqjrO

— The Vault (@TheVault1914) August 2, 2025

A bit more time related content. Watches and quartz. What makes them tick is fascinating. A good amount of science, persistence, and an attention span required to make them work properly.

Tick, tick, tick.

In theory, it works like this:

-Battery provides current to microchip circuit

-Microchip circuit makes quartz crystal (precisely cut and shaped like a tuning fork) oscillate (vibrate) 32768 times per second.

-Microchip circuit detects the crystal's oscillations and turns them into regular electric pulses, one per second.

-Electric pulses drive miniature electric stepping motor. This converts electrical energy into mechanical power.

-Electric stepping motor turns gears.

-Gears sweep hands around the clockface to keep time.

Theme music this evening.

Been reading and collecting a lot around Gen Z’s political attitudes and worldview. Not all of them are out of their minds.

Younger generation pushed to the right? A trend worth our attention?

***

Thompson writes for the Atlantic, so yeah.

Why are young people—in the US and around the world—shifting right?

I wrote about a phenomenon I call Generation C—the young people whose experience with COVID's health, political, and economic crises caused a strong shift right among young ppl across the developed world.

In… pic.twitter.com/ypaG2bguIF

— Derek Thompson (@DKThomp) February 18, 2025

***

AI and Jobs - A discussion with a Gen Z with a brain.

Had a youngin' over for dinner recently. He's in his mid-20s, has a degree in data science, and is heading into his final year of law school. His well of knowledge isn't yet deep, but his brain is impressive. He has the ability to distill ideas quickly, logically, and with the end goal in mind. I really enjoy hearing his perspective.

Young guy/oldish guy post-dinner conversation ended up mirroring many of the same themes discussed here at AoSHQ. We agreed that AI is a tool - more accurately a wrench/implement in the toolbox. Plausibly a very useful and common one, like a Phillips screwdriver or a 1/2" combo wrench. Not something obscure like a sink basin wrench.

We chatted about past "game-changing" technologies that were supposed to redefine or eliminate entire professions. We got sidetracked on how Excel was once predicted to make accountants obsolete. Obviously, that didn't happen. Instead, it created new opportunities - accountants now use it to crunch and present more data, and analyze more information to better serve their tasks.

Our next topic was that of job displacement. Yes, there will be some. But many of the roles most susceptible to automation aren't major pillars of the broader economy. That's not to dismiss the personal impact of job loss - just to say the macroeconomic effect may be limited. Think: proofreaders, entry-level bookkeeping, graphic design. And yes, coding came up.

AI can write syntactically correct code faster than humans. But that doesn't mean it replaces human coders entirely. The young man codes in Python (he hates HTML), and deploys AI for straightforward coding tasks, but builds out routines when specific logic is required - riddle solving. In his view, coding is just another tool to amplify human ability - not eliminate it. Smart kid.

Plenty of studies out there trying to predict which jobs AI will create or eliminate. Scroll through a few articles, and you'll see the same roles landing on both sides of the ledger. As usual, the alleged Smarties, simply don't know.

Which brought us to our final discussion point/topic: AI can assist, but it can't pull the mental levers that drive real value - strategy, negotiation, leadership, communication, creativity. Those remain the domain of Morons. And other humans.

Share your AI thoughts in the comments.

This ONT brought to you by: An inside straight draw (Ace of Spades high) and Three Dog Night.

Lurkers, as always - join the fray.

posted by Open Blogger at 10:00 PM

| Access Comments