Science News May 24 (Patreon)

[This is a transcript with links to references.]

Welcome everyone to this week’s science news. Today we’ll talk about traces of your DNA in the air, a new telescope that will track down gravitational wave sources, a new quantum advantage claim, e-Fuels, a superconductor experiment that failed to replicate, nuclear fusion investments, image reconstruction from surfaces, EU regulations for artificial intelligence, and of course, the telephone will ring.

Many thanks to our sponsors on Patreon, especially those of you in tier four and higher. We really appreciate your help. And you too can help us. On our Patreon page you’ll find transcripts of all videos with references and also some news items that didn’t make it into the video. So go check it out.

Researchers in the U.S. and Europe have published two studies showing that human DNA is in the air.

We constantly shed tiny bits of DNA, mostly by dead skin cells but they also sometimes stick to hair, or are swept out of the body with sweat and other fluids. That’s nothing new, but new is that technology has become so advanced now that researchers are now able not only to easily capture that genetic material but also extract personal details out of it, including genetic predisposition to illness.

A group of Marine ecologists in Florida stumbled on the phenomenon when they were collecting environmental DNA to study the green sea turtle. The noticed that human DNA was mixed in with every sample they took, even those they collected in the wild, far away from human habitation. They got so much they called it Human Genetic Bycatch.

A second group of researchers from South Carolina collected their samples while flying a light aircraft. They found traces of chicken, cow, weed, bacteria, human, and even garlic DNA  thousands of meters in the air above the southeastern US.

There are huge potential benefits from these findings.  You could check on the status of wildlife, try to find endangered species, monitor for pathogens, solve crimes, find missing persons, discover archeological remains.

But, of course, there are a lot of potential ethical problems, too. You could collect someone’s DNA without them knowing about it, conduct genomic surveillance, or track ethnic populations.

The Florida group is calling on regulators, researchers, and funders to develop an ethical framework before the technology becomes more widely used, because as we all know asking people to think ahead works extremely well.

The new telescope BlackGEM  will help astronomers track down the sources of gravitational wave events.

Gravitational waves, aka ripples in spacetime, result from violent astrophysical events like mergers of neutron stars and black holes. They can be detected on earth with large interferometers that measure the tiny space-time deformations as they pass through our planet. But violent events like this often also have a visual part which could provide more information. BlackGEM will help find them.

The new telescope will collect images in the visible part of the spectrum and work alongside gravitational wave detectors, including LIGO and Virgo. When these detectors see an event, BlackGEM’s task is to pin down the exact sources. This information will then be used to turn larger telescopes into the right directions, for example ESO's succinctly named “Very Large Telescope”. I like scientists better when they’re not trying to be witty with acronyms.

BlackGEM is located at ESO’s La Silla Observatory in the Atacama desert in Chile. At the moment it has three telescopes, each 65 centimetres in diameter. The plan is to expand the array to a total of 15 telescopes.

The new telescope excels at having a high spatial resolution, which allows for the identification and precise localization of faint sources against the backdrop of the night sky. This is a zoom into the bulge of our Milky Way.

Besides identifying sources of gravitational waves, BlackGEM might also help with studying fast transient sources, such as supernovae  or gamma-ray bursts. At this point we probably understand better what’s going on in other galaxies than on our own planet.

The Vancouver-based company D-Wave and researchers at Boston University say they have shown that the firm’s quantum computer is faster than a classical computer on a specific type of problem. It’s known as realizing a quantum advantage.

The team worked on D-Wave’s annealing-based quantum computer. They programmed a 5,000 qubit spin glass system that they could control and used it to solve optimization problems. It’s the largest programmable quantum simulation that has been performed to date.

Quantum annealing is a type of quantum simulation. It’s basically a way to custom design an optimization problem. If properly set up, the system relaxes into an optimal state that answers your question. It’s kind of like settling into the most comfy position on a couch, but with quantum tunneling and entanglement and all. Since it’s got something to do with quantum and with computing, I think it’s fair to call it quantum computing. But the approach is very different from that pursued by IBM and Google and many others, who try to develop what’s known as a universal quantum computer that could execute algorithms.

Hidetoshi Nishimori of the Tokyo Institute of Technology, one of the inventors of quantum annealing, says in the paper’s press release that it’s a major step towards large-scale quantum simulations of complex materials.

Scott Aaronson from the University of Texas at Austin, says it looks like one of the strongest papers yet from the D-Wave group but that the central question will be how hard this system is to actually simulate classically. Hi Scott. I’m not sure how I ended up being a news anchor either, but thanks for chiming in.

A team of British and Australian researchers has had a close look at plans to reduce the carbon emissions of flying, and they are very, very sceptical.

The International aviation industry has pledged to get to net zero emissions by 2050, part of the Paris Agreement to keep global heating below 2 degrees Celsius. If aviation doesn’t go green and if it keeps expanding then, by 2050, it’ll contribute about 0 point 1 degree Celsius.

There are four ways to make flying more environmentally friendly: batteries, hydrogen fuel cells, biofuel, or synthetic fuel made using renewable energies. Batteries and hydrogen aren’t going to save the day because they’re too heavy, have a short reach, and the infrastructure doesn’t exist. Some might be used for short-haul flights, but at least for the foreseeable future they can’t be used for long-haul flights. This leaves biofuels and synthetic fuels. I talked about this in more detail in an earlier video.

These researchers now examined 12 roadmaps from various institutions for aviation to reach net zero. As you see in this breakdown, they all rely heavily on biofuel and synthetic kerosene.

The researchers say there are two main problems with this. First is that producing all that biofuel would require a lot of land. While the exact numbers depend on just how the biofuel is produced, one example they give is the plan from the Energy Transmission Commission, for which biofuel for the aviation industry would take up as much as 280 million hectare. That’s about the size of Argentine and roughly 8 percent of all current agricultural land.
If growing all that biofuel results in cutting down forests, that could actually make climate change worse.

The second problem is that there are barely any industrial plants to make all the required synthetic kerosene, and if they built them that would require a sizeable share of renewable energy. Just to reach targets set for 2030 would take between 300 and 400 industrial plants. Today, just eight are in progress. And even if you look at the most realistic plans, like those of the European Commission, that would make only 28 percent of aviation fuel sustainable by 2050.

So, yeah, I guess one could say the plans of the aviation industry are a little too high flying.

Hello?

Your majesty.

Charles, then. Well, I’m very flattered you watched my little joke about the coronation.

No no, if I’d wanted to be royal, I’d have been born into a royal family, right?

Ah, just don’t wear squeaky green and you’ll be alright.  Bye.

Remember the superconductor breakthrough that made headlines 2 months ago? A team of Chinese researchers has failed to replicate it.

Here’s a refresher on what happened. Ranga Dias and his team at the University of Rochester announced in a Nature paper in March that they had made a breakthrough in developing a high temperature superconductor. They had produced a compound, which they said is lutetium hydride, compressed it between two diamonds at 10 kilobars of pressure, and watched it change colour. Along with the change in colour, the electric resistance changed, and it became superconductive. And all at room temperature.

At least that’s what they said. If it was true, that’d be a really big deal. 10 kilobars of pressure is still a lot, but it would be a big step close to room temperature superconductivity. Superconductivity at room temperature could make the electric grid dramatically more efficient.

But other scientists met Dias’s finding with some scepticism. That’s to no small part because he was the main author of a 2020 paper on superconductivity – also published in Nature – that had to be retracted last year after other scientists questioned the data and said they couldn’t reproduce the results.

Now a group at Nanjing University has tried to reproduce these latest findings of Dias, and failed. They followed the process step by step and were able to make a compound that looked identical. It even had the colour changes and same vibrational frequencies. But it wasn’t superconductive.

The Chinese team says their findings aren’t a repudiation of the latest Dias paper but that maybe they didn’t use enough nitrogen to dope the compound, or something like that. I think they’re being polite, but stay tuned, we’ll keep you up to date.

The U.S. Department of Energy has announced it’ll pour as much as 45 million dollars into researching the physics and technology of inertial fusion energy over the next four years.

The announcement follows last year’s big breakthrough at the National Ignition Facility, NIF for short. For the first time, they got more energy out from a fusion reaction than they put in.

The basic idea of nuclear fusion is that you push two small atomic nuclei together. That creates a heavier nucleus with slightly less mass. The mass that’s lost is released as energy. It’s about a million times more energy per volume of fuel than for fossil fuel which makes it such an appealing source of energy. There are two major approaches to make nuclear fusion happen, field confinement, and inertial confinement. I talked about this in great detail in a recent video.

The NIF uses inertial confinement. They shoot lasers at a fuel pellet to ignite fusion and hope that more energy comes out than goes in. The big challenge for inertial fusion is that you have to very exactly produce the pellets and then you must reliably and very exactly reproduce the shots. But the big benefit of inertial fusion is that it uses about a thousand times less tritium than field confinement.

The new DOE funding announcement will support research into the physics of the target, as well as its manufacturing, the technology required for the shots and how to rapidly repeat them, and how to experimentally validate all those properties. If you’ve always dreamed of making a career by shooting big lasers at things to blow them up and save the world along the way, here’s your chance.

Researchers at MIT and Rice University have figured out how to reconstruct images from reflections on glossy objects.

They called their code ORCa, which stands for Glossy Objects as Radiance-Field Cameras  but seems to have lost a G.

It uses machine learning to extrapolate from the reflections of the objects, say, a coffee mug or a metallic paperweight, to see the world from the object’s perspective. Basically, the researchers figured out how to strip out the distortions caused by the shape of the object and then turn the two-dimensional images into a 3D environment.

The researchers are also experimenting with using the system in drones, recreating a scene on the ground from small reflections in the environment, say windows or cars.

The European Parliament has moved closer to passing the world’s first law to control Artificial Intelligence. Good news, I suppose, considering Terminator is set in 2029.

The committees governing consumers rights and civil liberties adopted a draft of the AI Act last week. It will go to the whole Parliament in June.

The rules are intended to boost trust in AI systems at a time when there’s widespread concern over how the systems will affect fundamental rights and values, and the safety of the vulnerable in society. Such as children.

If passed, the law would ban some functions it considers to hold unacceptable risk, such as predictive policing, which tries to figure out where crimes will be committed and by whom. Or emotion recognition systems.

The act would control other functions by the level of risk,  from minimal to limited to high. The greater the risk, the more transparency the activity would require and the greater the accuracy of the data. High risk areas include harm to people’s health, safety, fundamental rights or to the environment. They also include AI systems that could try to influence voters in political campaigns. But most systems are considered low risk, like video games.

The law would also require extra measures for generative AI, like ChatGPT. Those systems would have to disclose that the material was generated by AI, make sure the system doesn’t make illegal material and publish summaries of the copyrighted material used to train the system.

Fines would be up to 30 million euros or 6 per cent of annual revenue, which could add up for some companies.

Hi Elon,

A TruthGPT?  Well, that could be a bad idea, or it could be a really bad idea.

If people could just find out what’s true, 99 percent of twitter traffic would disappear.

I knew you’d see the problem. Talk soon.