Learn about why rainy days make you sleepy; why the “Great Unconformity” is one of the biggest mysteries in geology; and the problematic amount of energy it takes to power AI — along with a potential solution.
Learn about why rainy days make you sleepy; why the “Great Unconformity” is one of the biggest mysteries in geology; and the problematic amount of energy it takes to power AI — along with a potential solution. Plus: how do you pronounce "Colorado," anyway?
Why do rainy days make you sleepy? by Kelsey Donk
The "Great Unconformity" is one of the biggest mysteries in geology by Grant Currin
AI requires an enormous amount of energy, but we might have a fix by Grant Currin
Subscribe to Curiosity Daily to learn something new every day with Cody Gough and Ashley Hamer. You can also listen to our podcast as part of your Alexa Flash Briefing; Amazon smart speakers users, click/tap “enable” here: https://www.amazon.com/Curiosity-com-Curiosity-Daily-from/dp/B07CP17DJY
Find episode transcript here: https://curiosity-daily-4e53644e.simplecast.com/episodes/why-rainy-days-make-you-sleepy-a-fix-for-ais-energy-appetite-and-solving-geologys-mystery-of-the-great-unconformity
CODY: Hi! You’re about to get smarter in just a few minutes with Curiosity Daily from curiosity-dot-com. I’m Cody Gough.
ASHLEY: And I’m Ashley Hamer. Today, you’ll learn about why rainy days make you sleepy; why the “Great Unconformity” is one of the biggest mysteries in geology; and the problematic amount of energy it takes to power AI — along with a potential solution.
CODY: Let’s satisfy some curiosity.
You got a good night of sleep and you haven’t done anything too taxing. But it’s raining outside, and all you really want to do is go back to bed. We all know it happens, but why? Why do rainy days make you sleepy?
The truth is that we don’t know for sure. But it probably has something to do with the light, sound, and oxygen levels in the air.
First of all, just look outside on a rainy day. The clouds gather and everything looks dark and hazy. On clear days, the sunlight tells you that it’s daytime. That triggers your brain to stop producing the sleep-producing hormone melatonin, so you’re more alert. But on cloudy and rainy days, the sun doesn’t really come out in the morning. With less sun exposure to tell your brain to stop producing melatonin, the body might not know it’s time to wake up.
Then, there’s the sound of falling rain. It’s not a mystery why so many people play rain sounds to help them sleep. The sound of rain is comforting! It’s the audio version of being wrapped in a warm blanket! That’s because rain sounds are what experts call "pink noise." That’s like white noise but with fewer high frequencies. And evidence suggests that pink noise can help people sleep more soundly. So if you’re listening to pink noise all day, it makes sense that you’d start to feel a little sleepy.
Finally, as any good morning-news-meteorologist will tell you, there's the fact that rainy days have low barometric pressure. The lower the pressure in the air, the less oxygen it contains. Just think about climbing a mountain — the lower pressure at those higher altitudes reduces the oxygen and makes it harder for mountaineers to think clearly. The low pressure of a rainy day has a much, much milder effect, but it’s similar. That lower oxygen doesn’t pose any risks to your health! But getting less oxygen to the brain could make you a little more lethargic than usual.
So the next time you can’t seem to shake that sleepy feeling on a rainy day, just remember that it’s normal. Try turning on a bright light or playing some upbeat music. Or just wait it out. The sun will come back out eventually!
The deep history of Earth is written in the layers of rock that make up the planet’s crust, but a lot of that history is missing. Geologists call this puzzling gap the Great Unconformity. Earlier [this month], researchers used an exciting technique to hunt for clues. It turns out the Great Unconformity might have more than one cause.
The Great Unconformity is kind of a confusing name, but here’s the deal: throughout Earth’s history, new rocks have formed on top of older rocks. There are a lot of techniques for figuring out how old a layer of rock is, but scientists can usually look at the fossils in a layer or analyze the physical make up of the rock itself to figure out that a particular layer of rock is, say, 500 million years old. They’d expect the next-deepest layer to be a little older — maybe 600 million years old. But there are quite a few rock formations where that isn’t the case.
The researchers behind this study look at one of these formations at the base of Pike’s Peak in Colorado. At their field site, there’s a layer of rock that’s about 500 million years old that sits on top of a layer that’s one billion years old. That means there’s no apparent record of half-a-billion years. Where’d all that history go?
Scientists have a handful of different answers. Some say it’s a result of all the geologic chaos that happened during the formation of Rodinia, a supercontinent way older than Pangea. Others say that glaciers shaved off all that rock at once during a later ice age called Snowball Earth, creating the Great Unconformity in one disastrous event.
These researchers found an answer by peering deep into the atomic structure of the older rock. A technique called thermochronology helped them figure out what temperature extremes the rock had been through in its tumultuous billion-year lifespan. The hotter the rock was, the deeper it’s been.
From their measurements, the team concluded that the layer was forced up to the surface between 1,000 and 720 million years ago — before Snowball Earth, right around when Rodinia existed. At that point, the rock was vulnerable to erosion by rain, wind, freezing temperatures, and a host of other natural processes, which could have removed its upper layers from the geologic record. But importantly, this means that the geologic gaps of The Great Unconformity probably didn’t happen all at once everywhere. Instead, each one likely had its own cause and its own timing.
Of course, this is just one gap in one location. To confirm that The Great Unconformity was actually Several Great Unconformities, other geologists will have to take a closer look at a bunch more gaps in time.
Artificial intelligence, or AI, can do some amazing things; I mean, it can diagnose diseases way before doctors can. It can create original art that’s indistinguishable from human creations. But there’s another thing it does that’s a huge problem: AI consumes a ton of energy. So it’s good news that a team of engineers has announced a breakthrough that could make the futuristic tech a little more sustainable.
Before I get to that, though, lemme give you an idea of much energy AI gobbles up: in 2019, a team of researchers taught an AI system to solve a Rubik’s cube with a robotic hand. Even though the system had more than a thousand computers at its disposal, the project still took several months to complete. All told, the effort consumed about 2.8 gigawatt-hours of electricity. That’s the same amount that three nuclear power plants can crank out in an hour.
Another study found that training a single natural language processing model — the kind of thing Siri and Alexa are built on — can consume as much energy as a car over its entire lifetime, including what it takes to build it. But AI is only becoming more important, so finding ways to make it more energy-efficient is becoming increasingly necessary.
Researchers at Purdue University in Indiana have taken a step in that direction. See, the intelligence part of artificial intelligence is mostly in the software, and as that software gets smarter, it takes more computing power to run it. So this team tried their hand at making intelligent hardware. They used what’s known as a quantum material, and it’s called that because it relies on weird quantum effects to do things that aren’t possible with traditional materials. Specifically, they used high-speed pulses of electricity to move protons around in a quantum material called neodymium nickel oxide. Usually, quantum materials require super-cold temperatures to work, but the breakthrough here was that these experiments all happened at room temperature. That makes it a lot easier for researchers to imagine using the new material in electronics.
The material demonstrated what’s called “tree-like” memory. That’s a way of organizing information by storing related data together in categories called branches. It’s based on the way humans learn and organize information in their brains. For instance, when you see a beagle, you think of the category “dog,” even though Snoopy looks nothing like a Chihuahua or a Great Dane. Current AI software can do this too, but by building it into the hardware as well, the hope is that the two components could work together to achieve more with less energy. Here’s hoping that it won’t take an automobile’s worth of energy to create the Siris of the future.
Let’s recap the main things we learned today
[ad lib optional]
CODY: Today’s stories were written by Kelsey Donk and Grant Currin, and edited by Ashley Hamer, who’s the managing editor for Curiosity Daily.
ASHLEY: Today’s episode was produced and edited by Cody Gough.
CODY: Join us again tomorrow to learn something new in just a few minutes.
ASHLEY: And until then, stay curious!