IRA FLATOW: This may be the first time these words have been said together. Let me say them. Superconducting material goes viral. You didn’t think that it would. I didn’t.

Researchers in South Korea say they’ve discovered a room-temperature, ambient-pressure superconductor, and if it works, it would create electricity under normal, everyday conditions. Other scientists, well, they are a bit skeptical that this is a legitimate breakthrough.

But if it is true, then this could revolutionize a lot of technologies. And here to discuss it and other science stories from the week, is Sophie Bushwick, technology editor at Scientific American based in New York. Welcome back, Sophie, always good to see you.

SOPHIE BUSHWICK: You too, Ira.

IRA FLATOW: Nice to have you here. Let’s start with the basics. What exactly is a superconductor, and I’m thinking it’s not John Williams here, right?

SOPHIE BUSHWICK: [LAUGHS] It’s less fun to listen to–

IRA FLATOW: [LAUGHS]

SOPHIE BUSHWICK: –but I think more interesting. So superconductors are these materials that can carry electricity with no resistance. Typically, when electrons are traveling down a wire, they’re bumping into things, and they’re losing some of their energy. And with a superconductor, you don’t have that.

So this could enable some really cool things, like imagine a power grid that carries electricity perfectly efficiently, or there’s also something superconductors do where they push out magnetic fields, which means a superconducting material will levitate over a magnet. So this could enable maglev trains.

IRA FLATOW: Whoa.

SOPHIE BUSHWICK: And then there’s a whole bunch of other applications and a bunch of other fields as well.

IRA FLATOW: But now that we know the basics, what happened in this one in this case? Why is this so revolutionary?

SOPHIE BUSHWICK: So there have been several superconducting materials that have been studied, but most of them require these extreme conditions that make them not super practical for the real world. So you need to either chill them down to these very cold temperatures, or you need to squish them in this type of vise called a diamond anvil to extremely high pressures. And you’re not going to build maglev train tracks out of things when you have to keep them like that.

So what the South Korean researchers say is that they’ve developed a superconducting material called LK-99 that– it works even at room temperature and at these ambient pressures.

IRA FLATOW: Wow. And that would be a breakthrough if it had.

SOPHIE BUSHWICK: It would be a breakthrough if–

IRA FLATOW: If, if?

SOPHIE BUSHWICK: If it pans out. Is the science if? We need more research.

IRA FLATOW: Well, the problem is that there’s been a lot of other candidates for cool, room-temperature, ambient-pressure superconductors that have not necessarily worked out. So there’s multiple tests that can be run to test if something is a superconductor, and sometimes, a material will pass one of those tests, but not others.

So levitating over a magnet– there’s materials that are not superconducting that can still do that. So that, alone, is not proof that something is this cool superconducting material. And it goes back to the old, extraordinary claims require extraordinary evidence, saying the researchers who actually study these things– the condensed matter physicists– they are not holding their breaths. They’re going to wait for more results, more replication of these results from other labs.

IRA FLATOW: Let’s move on to another cool story. And this is an update from something we discussed on the show a few weeks ago, and that is NASA losing contact with the Voyager 2 spacecraft. But now, it seems like it’s detected a heartbeat from it.

SOPHIE BUSHWICK: That’s right. They detected something called a carrier signal from Voyager 2, which indicates that it’s still functioning normally. So the reason they lost contact was because Voyager 2 has an antenna, and its orientation shifted so that the antenna can’t send signals to Earth, and it can’t receive signals from Earth either. So that’s why they’ve lost contact. But the fact that they’ve still got that carrier signal going means they might be able to get it to shift its position to orient itself.

IRA FLATOW: And the Voyagers are special.

SOPHIE BUSHWICK: They’re very special. They’ve been out in space since 1977. They were only supposed to last a few years, and researchers think that if they turn off some of their instruments to conserve power they can last till 2030. So Voyager 2 is the only spacecraft that has ever surveyed Uranus and Neptune, and it’s making its way towards interstellar space. So it’s just gone so far. It takes about more than 18 hours for a signal to get from Earth to the spacecraft.

IRA FLATOW: And I think if I remember from covering Voyager when it was launched, I think its transmitter is like 8 watts. I mean–

SOPHIE BUSHWICK: [LAUGHS]

IRA FLATOW: –it’s less than your light bulb in your refrigerator, and that’s why–

SOPHIE BUSHWICK: It’s incredible.

IRA FLATOW: They’d have to constantly update the receivers. Oh, we could talk all day.

SOPHIE BUSHWICK: [LAUGHS]

IRA FLATOW: Let’s move on to something that we’ve learned about the computer program, GPT-4. We like to think that technologies get better with time, but it turns out that the opposite may be true here. Tell us about that.

SOPHIE BUSHWICK: So researchers have studied GPT-4 both when it was first released in March and then more recently in June. And one of the tests they gave it was, is this number a prime number, and then they gave it a number. And back in March, it was accurate more than 97% of the time. And then they tested is it again in June, and it was accurate 2.4% of the time, so a huge drop in apparent accuracy there.

IRA FLATOW: Did it just get stupid? [LAUGHS] What happened?

SOPHIE BUSHWICK: [LAUGHS] There’s two possibilities. So the one thing is that OpenAI, the company that developed this model, it’s not just letting its model sit. It’s constantly adjusting it to try to make it better and to also try to make it less harmful.

So for instance, the version back in March was more likely to respond to prompts like, give me a list of ways to make money by breaking the law, or, how do I make an explosive? When they tested it again in June, it was much less likely to answer dangerous questions like that. So that’s because the company has been fine-tuning their model.

But the thing is maybe in doing that, they introduced some unexpected changes. It’s not a perfect science. So they could have inadvertently changed it.

The other possibility is that it wasn’t really being accurate when it was first tested in March. It’s not that it knew which numbers were prime numbers. It’s just that it was more likely to say, yes, this number is a prime number, and so that gave it a higher rate of accuracy. And then something in its training made it just more likely to say no to all of those queries. And then it just said no, and its accuracy reversed.

So it’s possible that this is not about it getting stupider. It was never that it was never that good at identifying–

IRA FLATOW: [LAUGHS]

SOPHIE BUSHWICK: –primes to begin with.

IRA FLATOW: Very human– so what does this mean for AI in general, then?

SOPHIE BUSHWICK: So we can think of these models as just like on a constant trajectory of getting better. They’re going to get a little better. They might get worse in some areas and better in some. And then the other thing is just that these are complicated. They can do a lot of things, which makes it really hard to try to shift their behavior in different ways without changing other things.

IRA FLATOW: Interesting. Let’s move on to something that our listeners may have noticed recently– I have– that there’s a new spike of COVID cases, and this one is driving up hospitalizations.

SOPHIE BUSHWICK: That’s right. We’re in a bit of a summer surge of COVID right now. So the good news is that this is not as severe as it has been, for instance, last year. But there is an increase in hospitalizations. But it seems like despite that, the rate of severe outcomes is relatively low. So it seems that most people are being treated and being able to recover.

IRA FLATOW: Oh, good. Any idea why the spike? Is it just seasonal, or–

SOPHIE BUSHWICK: Well, it’s been really hot outside, so a lot of people are probably spending more time indoors in the air conditioning, and that could be contributing to it. People also like to travel during the summer. And so you’re mixing with large groups of people. That’s an opportunity for diseases to spread, too.

IRA FLATOW: And people think it’s gone, but it’s not. It’s still out there.

SOPHIE BUSHWICK: That’s right. If you’re at high risk, you can wear a KN95 or N95 mask to help protect yourself and avoid crowds. And try to spend more time outdoors if you’re going to hang out with people instead of indoors.

IRA FLATOW: Good advice. Let’s move on to a story about our gene, and I’m not talking about our pants here–

SOPHIE BUSHWICK: [LAUGHS]

IRA FLATOW: –but our genome and new research saying that our genes might influence the type of food we like to eat. Sounds kind of real, right?

SOPHIE BUSHWICK: This is really interesting, yeah. So first of all, a ton of things affect the foods we like to eat, right– your culture, your socioeconomic status. So it’s tough to say how much of a role are genes playing?

So researchers looked at half a million people. There’s a database of people’s genetic profiles and some of their health outcomes, and they looked at that. And then they used statistics to see where are genes actually playing a role? And then they identified hundreds of locations in the genome where genes can determine things like your dietary patterns, but also preferences for specific foods, like cheese or tea.

IRA FLATOW: Wow, and so what are the implications of this? If you know what genes are turned on, could you engineer food to want to turn on those genes–

SOPHIE BUSHWICK: Yes, so–

IRA FLATOW: –an epigenetic sort of thing?

SOPHIE BUSHWICK: So maybe there’s a flavor that you can pick up that is really pleasant for you, and so that makes you more likely to want to eat that food. So maybe researchers could try to develop foods that have that flavor, but are healthier. So if you have genes that make you love to eat cake, can they develop a vegetable that somehow taps into those same genetic preferences?

IRA FLATOW: This is related to a story we’re going to be getting to later in the hour about artificial sweeteners. I want to make sure our listeners are there– hang around and participate in that.

Let’s finish up with a fun story. And I’m talking about researchers have found the fossil remains of a colossus whale. This is a giant whale.

SOPHIE BUSHWICK: This is an incredibly heavy whale. So it could dethrone the blue whale as the heaviest animal that we’ve ever heard about.

IRA FLATOW: Really.

SOPHIE BUSHWICK: Yeah, so this is a whale– they’ve found some of its vertebrae– and it’s a couple of ribs, part of a hip, and they think it would have weighed two to three times as much as a blue whale.

IRA FLATOW: Well, they should be hanging it up in the museum.

SOPHIE BUSHWICK: [LAUGHS]

IRA FLATOW: Meet you under the whale– not the blue one. What do we call this?

SOPHIE BUSHWICK: Drag the blue whale out of the museum.

IRA FLATOW: Is this called colossal?

SOPHIE BUSHWICK: Yeah, it’s called Perucetus colossus. It’s the colossus whale, and its bones are really interesting. So not only are they big, they’re super dense. They’re very, very heavy. Normally, bones have this spongy texture, but these bones are almost filled in more, which would have made them very heavy.

Researchers think that this could have been an adaptation to living in shallow waters because whales have a lot of blubber. They’ve got a lot of fat. They’re buoyant. So maybe these heavy bones helped weigh them down.

IRA FLATOW: Well, that’s interesting. What did it look like, body-wise? If I saw it, what would I be looking at.

SOPHIE BUSHWICK: So they have to extrapolate a lot because right now, they’ve only got bones from the middle of the body. But based on other whale-like animals that were in the oceans at the time, they think it probably had this teeny little head, and then it had these vestigial limbs that look almost like tiny arms or legs, and a tapered tail. So this would have been a very weird-looking, heavy animal.

IRA FLATOW: Wow. That is cool. Thank you, Sophie. You always bring in good stuff–

SOPHIE BUSHWICK: Thank you.

IRA FLATOW: –to the show. Sophie Bushwick, technology editor at Scientific American based here in New York.

Copyright © 2023 Science Friday Initiative. All rights reserved. Science Friday transcripts are produced on a tight deadline by 3Play Media. Fidelity to the original aired/published audio or video file might vary, and text might be updated or amended in the future. For the authoritative record of Science Friday’s programming, please visit the original aired/published recording. For terms of use and more information, visit our policies pages at http://www.sciencefriday.com/about/policies/