Science News October 10 (Patreon)

 [This is a transcript with links to references]

Welcome everyone to this week’s science news. Today we’ll talk about a new candidate for a cosmic string, the mysterious shrinking of planet Mercury, a nuclear clock, the first quantum engine, a simulator for human diseases, whether we can find new physics with spinning black holes, AI that wants to help find aliens, how to compute with photons, and of course, the telephone will ring.

Astronomers might have found evidence for a cosmic string.

A cosmic string is a hypothetical giant thread of pure energy, longer than a galaxy and with basically zero thickness. The existence of cosmic strings was first proposed in the 1970s, as leftover topological defects from the plasma in the early universe.

However, that idea was ruled out because it didn’t fit with the properties of the cosmic microwave background that were observed in the 1990s. The idea of cosmic strings was then revived by string theorists who argued that under certain circumstances, the normally tiny strings of string theory could grow to such huge proportions.

Because of its high energy density, a cosmic string would act as a strong gravitational lens, and because it’s a one-dimensional object, it would very cleanly duplicate images behind it. Basically, one image goes left around the string, the other one right. So if you’re looking for cosmic strings, you look for galaxies that seem to appear twice and there’s no visible cause of lensing.

In the new paper, the astronomers report that they’ve identified several cosmic string candidates in data that was collected by the Planck satellite and that they followed up on with further observations at the Himalayan Chandra Telescope. This way, they identified one candidate in particular that seems to be an almost perfect duplication and that could have been created by a cosmic string. They did a spectral analysis of both parts of the image and found that the spectra look almost identical. In case that sounds convincing, let me add that the spectrum mostly tells you what the galaxy is made up of, and nearby galaxies are quite plausibly made of similar star stuff.

We had a candidate for a cosmic string already in 2006 when the Hubble telescope spotted a similar double dot. It turned out to be a pair of galaxies. So don’t get too excited, it might just be one damn string after another.

Planet Mercury is shrinking. Geologists have known for some time that it has been shrinking in the past. New is that they’ve found it’s still shrinking .

Mercury seems to be undergoing a process known as “global contraction”. It’s becoming wrinkly, basically, kind of like an old apple. Scientists estimate that it’s lost a few kilometres in diameter in the past few billion years.  They first suspected this in the 1970s when the Mariner 10 mission sent images back to Earth, and the surface of Mercury looked very wrinkly. They got better data in 2015 from the Messenger mission which confirmed the suspicion.

Just exactly why Mercury shrinks is somewhat unclear. It could be that it’s cooling at a rapid rate or it could be that it’s to do with the slowing down of its spin, and there are some more ideas floating around in the literature.

In the new paper now they analysed the data from the messenger mission in detail. Mercury is covered in grabens, which are downward dips. Grabens have an upward counterpart, called horsts. The geologists say that these grabens must be “relatively young” by which they mean less than 300 million years. They think so because otherwise these grabens would have been covered up by sediments and rubble. So, they conclude, Mercury hasn’t stopped shrinking. ESA is about to send a spacecraft to Mercury in 2025, which will hopefully tell us more about what’s going on there.

So it’s not just us who are becoming more and more wrinkly, Mercury shares our fate. Though I’m very pleased that at less than 300 million years I can still be considered “relatively young”.

Researchers from Germany and the United States have found a way to build a clock that keeps time with a record accuracy of one second in three hundred billion years, and they did it with a big X-ray laser.

Atomic clocks are currently the most precise way of keeping time. They work by tuning a laser to a resonance frequency between atomic energy levels. At the moment, the most widely used resonance on is in Caesium atoms. The accuracy of these Caesium clocks is roughly one second in three hundred million years. These clocks are so precise because the resonance is an incredibly reliable pacemaker, determined solely by the properties of the atom.

Now, some atoms have more, others less suitable resonances, but the typical frequency of these resonances is determined by the mass of the electron, so there are no big gains to be made by switching to other atoms.

However, instead of using a resonance between electron levels, you can use a resonance of the atomic nucleus. This is an advantage because nuclear resonances are at much higher energies. Higher energies mean smaller wavelengths, and therefore higher precision for the pace-maker. Nuclei are also less susceptible to environmental disturbances, so they make more accurate clocks.

One of the nuclei that scientists have looked at is scandium which has atomic number 21. This is because the structure of its nucleus gives it an incredibly narrow resonance that would make for a very accurate pace maker. But if you want to use that as a clock, first thing you need to do is to measure the resonance. This is what the new study is about.

In the paper that was just published in the journal Nature, the researchers report that they were able to measure the resonance in scandium nuclei with unprecedented precision. It would allow timekeeping with an accuracy one thousand times better than current atomic clocks, roughly one second in three hundred billion years, or about three femtoseconds per second.

The difficulty in using resonances of nuclei instead of atomic energy levels is that they are at much higher energies, which means you need to excite them with lasers of higher frequencies. This is what they need the X-ray laser for. Scandium is not the only nuclei you can use to build a nuclear clock. There is also a European group working on building a nuclear clock based on a Thorium resonance.

They’re not just doing this because they like playing with X-ray lasers. Timekeeping accuracy is relevant in many research areas, notably navigation and space travel. Your tax office would also really like to be able to fine you for handing in your tax return 3 femtoseconds too late.

By the way this video comes with a quiz on quizwithit dot com. Even better! If you subscribe to quizwithit, you can collect points from all our videos, and you get free access to the transcripts with links to all references. It's an easy way to support both our channel and, hopefully, your memory, so go and check it out.

Scientists in Japan and Germany have worked together to build the first quantum engine.

This engine exploits the difference between bosons and fermions, that are different types of particles. Fermions don’t like to share the same space. Bosons on the other hand, find it cozy to sit in the same place. Electrons are fermions, and that they don’t like to share space is why we have atomic energy shells, otherwise electrons would just all sit on the same shell. But it’s not just elementary particles which are bosons or fermions, entire atoms have similar properties, depending on how their electron shells are filled.

These two groups of particles are not as distinct as you may think because you can combine two fermions to a boson. And that’s what they did in the new paper.

They used an ultracold gas of about 60 000 Lithium atoms which at the beginning were fermions and sat very spaced out. Then they applied a magnetic field that coaxed the fermions to combine to bosons and they moved closer together. Finally they relaxed the field and the atoms become fermions again. So they’ve converted the energy from the magnetic field to mechanical energy. It’s an engine.

This is indeed very similar to how car engines work, if you think about it. You inject petrol, ignite it, it expands and pushes a piston which drives a motor. In the quantum engine, the expansion comes from the transition from bosons to fermions.

This quantum engine isn’t going to move a car any time soon because it works at temperatures near absolute zero, but it’s another step on the way to putting “quantum” before every other word.

A new device about the size of a lunchbox can tell scientists how the human body would react to various drugs and diseases. The device is called Lattice, and it was built by scientists at Northwestern University in the US. They envision it as an alternative to conventional in vitro technologies.

The device consists of eight wells, which can contain anything from organ tissue to bacteria. The wells are interconnected by a series of pumps and channels that transport artificial blood or other liquids. This allows researchers to simulate interactions between tissue in various parts of the body, which is difficult to do with current in vitro technologies. They have already begun using it to study the origin of polycystic ovarian syndrome. Over time, the researchers hope, the device could become popular in laboratories that study physiology and pathology.

Next time you make jokes about physicists because they describe humans as approximately spherical,  keep in mind that for doctors you’re basically a box with pumps and channels.

I got a lot of questions about these headlines which say that spinning black holes can help us find new physics. They’re about something called a Kerr black hole, or so the news seems to say.

What’s a Kerr black hole? A Kerr black hole is just a spinning black hole. It’s named after the New Zealand physicist Roy Kerr who was first to write down the maths for it. Black holes almost always rotate, just because it’s incredibly unlikely that matter which collapses to a black hole had no initial angular momentum, and the angular momentum must be conserved. So if one could use spinning black holes to find new physics that’d be pretty cool.

But if you look at the news somewhat closer, or even better have a look at the paper, you’ll see that it’s not about just Kerr black holes, but extremal Kerr black holes. That they’re extremal doesn’t mean they like sky diving, it means that they rotate at the maximal speed that’s theoretically possible. Extremal black holes don’t exist in the real universe.

The authors of the paper estimate what quantum effects one should expect for gravity to appear near such extremal black holes, or black holes that are near extremal. They find that those effects would become very noticeable and would probably be measurable if those things existed.

What they say is almost certainly correct, though I admit that I didn’t check the maths. If I want to worry about stuff that doesn’t exist, I can just think about my pension savings.

Hello?

Ah, hi Elon,

No more headlines with the images? Yes, great idea, better not confuse people with too much information.

Ah, it’s much easier to comment on an article if you don’t know what it’s about. It’ll be great! Love you too!

Artificial intelligence is already helping us with a bewildering variety of tasks, from locating hard-to-find tumours to conducting beach cleanup missions. But it’s got a new mission: To help find life on other planets. That’s right, AI is joining the search for aliens.

In a recent paper, researchers in the United States say AI can help differentiate biological traces from non-biological samples. They trained their algorithm on more than one hundred and thirty samples which human scientists had already labelled to be of biological or non-biological origin. After that training, the program was able to judge other samples with 90 percent accuracy, correctly identifying recent biological traces that had come for example from shells, teeth, or rice, but also ancient ones such as coal, oil, or amber.

Identifying the biological origin of old samples is no small feat, because organic molecules tend to degrade over time. The researchers believe their program could help nail down the origin of ancient sediments, for example those on Mars.

So far they’re only proposing to use AI to look for alien life in sediments, so don’t worry, your little secret is safe a little longer.

Scientists in China have found a way to simplify a method used in photonic computing, saving both space and energy.

Photonic computing works by shooting laser light through microscopic optical components. Depending on how the light is routed with mirrors and beam splitters, it can execute different calculations. You then read out the result when the light exits the setup, or return it for the next step of the calculation.

Photonic computing is especially useful for matrix multiplication, where the matrices are represented by a grid of tiny interferometers. These photonic components are a very direct encoding of the matrix multiplication.

This is very energy and time efficient because in normal computers you don’t encode matrices directly in hardware, but indirectly in bits that stand for certain matrix elements, and transistors that operate on them. This might sound rather boring, but matrix multiplications are a common element of many real-world operations, from population studies, to economics and finance, to biology.

However, if you directly encode a matrix in a photonic computer, then this’ll generally be a complex valued matrix, that is, a matrix whose entries are complex numbers. This is because photons in an interferometer carry two types of information: which path they go, and how they interfere, that is, their phase.

Using photons for calculating with complex valued matrices has been done before. But for most everyday applications, one doesn’t need complex numbers—real numbers will do. So using an approach that encodes complex numbers is somewhat wasteful.

In the new paper now, the researchers  have found a way to simplify photonic matrix multiplication so that it works more efficiently for real numbers. They did this smartly removing some components which reduced the size of the chips and also made them more energy efficient.

This development is part of a general trend we’re seeing in computing, that certain types of calculations are being run on specialized hardware to speed them up. This is also the motivation behind quantum simulations and neuromorphic computing, it’s the computerized version of outsourcing to specialists basically.

When they said that photonic computing is getting real, they might not have meant what you thought they meant.