Record-breaking neutrino detected by huge underwater telescope
34:56 Share

They say opposites attract. That's why the Sleep Number smart bed is the best bed for couples. You can choose what's right for you whenever you like. You like a bed that feels firm but they want soft? Sleep Number does that. You want to sleep cooler while they want to feel warm? Sleep Number does that too. Why choose a Sleep Number smart bed? So you can choose your ideal comfort on either side. And now save 50% off on the new Sleep Number limited edition smart bed.

Limited time exclusively at a Sleep Number store near you. See store or sleepnumber.com for details. With the Redfin app, you'll know the moment your next place hits the market. Whether you're looking to buy your dream home or rent a suite apartment, give Redfin your gotta-have-it wish list of property features, and you'll receive real-time notifications tailored just for you. Ready to see it up close and personal? Scheduling a tour is just a tap away. Don't wait to find your perfect match.

Download the Redfin app and start searching today. Nature.

Welcome back to The Nature Podcast. This week, the detection of a record-breaking neutrino and an update on the funding situation in the US. I'm Dan Fox. And I'm Nick Pachachow. Most telescopes studying the cosmos observe waves of light in different frequencies. But there is more than one way to see what's going on in space.

Astronomers can also look at the particles that fly out of energetic phenomena, such as supernovas. Though electrons and protons often get absorbed or deflected on their route, that's not the case for neutrinos. These ghost-like particles interact with almost nothing, meaning that they can reach all the way to Earth, bringing with them the secrets of the cosmic labs in which they were made.

But neutrinos are super hard to detect. To try to spot them, scientists have placed an enormous array of detectors deep under the Mediterranean Sea, off the Italian island of Sicily. The experiment, known as KM3Net, doesn't detect neutrinos directly, but rather the faint flashes of light given off by new particles they create as they collide with molecules in the water.

And now, amidst the fish and ocean currents, the researchers behind KM3Net say they have spied a very special neutrino indeed. Reporter Lizzie Gibney spoke to Art Heibel at NICEV, the Dutch National Institute for Subatomic Physics, and member of the KM3Net collaboration.

She started by asking him to describe this very unusual kind of telescope. So this is a telescope in the sense that we look for things from space, but that's basically where the similarity ends. We don't look at light. We look at particles. They're called neutrinos, and they come to us from space and bring us information about astronomy. So what we do is we go to the bottom of the Mediterranean Sea, where it's pitch black normally.

And we equip a large volume of water with light sensors located in blast spheres. They're about 40 centimeters in diameter.

And we put a lot of those glass spheres in a sort of matrix in the water. They're standing on the bottom of the seafloor at three kilometers depth. So we've got this massive net detecting light under the sea in the Mediterranean. And how does it actually detect these neutrinos then? Because they're really difficult to detect, right? Exactly. Neutrinos are quite famous for their reluctance to interact with any normal matter. So most neutrinos will fly through our detector without us ever having a chance to see them.

And just to catch enough of them, we built these huge detectors just to have the hope to see them. So the way it works is that we need to rely on a neutrino having an interaction. And if a neutrino interacts close to our detector or inside, a neutrino will produce a secondary particle. In the case of the paper we talked today about, it's a muon.

And when this flies through the water, it emits light. And so we have detected this light. And from the arrival time of all this light, we measure that with the accuracy of like a nanosecond. From this data, we can be very sure that it was a muon. And also we can measure quite accurately its direction. And

And how do you know that this muon was made by a neutrino interacting? Because we do get muons coming in from space and they are, you know, sometimes produced. How do we know that this is from a neutrino? Indeed, we see with this detector many, many muons. In the period we talk about here, we have detected 110 million of them. So they are produced in the atmosphere by neutrinos.

cosmic rays hitting air molecules, and we see them as downgoing in our detector. So what's special about this event we talk about here is that we saw a very horizontal muon.

So if we point back where it should have come from, it would have had to travel for hundreds of kilometers through the water. And we know that neurons cannot do that. So the picture is that we are sure that this neuron that we see must have been caused by something else. And the only particles that can traverse this amount of material that we know of are neutrinos. So the neutrino is the only thing that could have got all the way to the detector through the water, through Earth.

and then it must have interacted very nearby to create this neuron. Right, it must have interacted within a few kilometers of the detector. And so what's special then about this particular neutrino you saw? I think, gosh, it was almost two years ago to the date that this paper comes out that it was discovered. What was special about it? Well, what's special about it is that it's the most energetic neutrino that we or anybody else has ever seen before. The reason why it took two years for this publication to arrive after the actual detection was

is that first we have to calibrate the detector. And then to be honest, it also took us a little while to realize how extremely special this event was, that it was actually that we could trust what we had seen there. Of course, first we checked all kinds of mistakes, but we really believe now we can trust it. And so just how energetic is this neutrino? We've said it's unlike any other before, but how high energy is it? So our energy measurement that we use in particle physics is the electron volt.

And so this particle has an energy upwards of 100 petaelectron volts. That means 10 to the 17 electron volts. It's a huge number. But to put it in a bit of perspective, the LHC accelerator at CERN, it accelerates protons. It's a different type of particle, but up to 40 TeV. So that's tera electron volt. So we are 10,000 above that energy, roughly.

And what's the highest energy neutrino that had been found before this one? So far, the highest energy neutrino was observed by the IceCube experiment. It's a very similar idea, but implemented in the south polar ice. And that has an energy of about six petaelectron volts. So this event is roughly a factor 30 or 40 higher than that one.

And the question, of course, is what caused it? What could have made a neutrino that is just this energetic? Do we have any ideas? There's lots of highly energetic astrophysical objects that could be responsible for particle acceleration because you first need to accelerate particles before you can make neutrinos of this energy.

For example, a typical source would be associated to the massive black holes that are seen in the centers of galaxies called active galactic nuclei. Matter is falling on these black holes and we have reason to believe that some of this matter is ejected. So typically one would accelerate protons there that interact to cause the neutrinos.

I do have to say we have so far not been able to, with certainty, pinpoint the exact object that this neutrino has originated from. We did quite some studies seeing if at the point in the sky where it comes from, there is an interesting object. But we do not measure the direction of the particle with information.

enough accuracy yet to really pinpoint which object it came from. So you've had a look in the sky, kind of in the vague direction where it came from. There's no smoking guns. I mean, there's many interesting objects. There's a lot of these galaxies in the sky. And so if you believe it came from one of these galaxies, that's very possible. Another possibility is that actually neutrinos are also expected from related with other source. These objects are also sending protons and other atomic nuclei into space.

So they can interact at some point, for example, with a photon from the cosmic microwave background radiation. So you also expect a sort of diffuse source of neutrinos. So this has been realized a long time before that there should be this source. It's called cosmogenic neutrinos.

actually in more or less exactly the energy range where we have now measured this one. So it's very possible that in the future we will be able to say it came from a particular object, but it's also possible that it was actually the first of these cosmogenic neutrinos that has been detected. And you mentioned this is more energetic than other neutrinos seen so far, and KM3Net is still not full size, right? It's still actually being built, right?

How did you see this incredibly energetic neutrino? I mean, did you think you just got lucky and like there happened to be a really, really high energy one that you caught? How did this happen? Yeah, so this is indeed the question we asked ourselves because there are other detectors that has been looking for actually several years already and has not seen anything like this.

So one possible scenario is that it was just by chance that in our case it happened in the first year. It will be extremely interesting to see what we and these other experiments detect in the next decade that will tell us if this was an upward fluctuation

or whether something else is going on. I mean, we have thought about alternatives, but we cannot really think of anything. If we would see another one of these events next year, it would be really weird to explain why the other experiments have not seen anything. But so far, I think the statistical

probability is not crazy small. That's just basically by chance that we saw it and they didn't. How often would you expect to see a particle of this energy then? So we did a little bit of sort of statistics, gymnastics to answer that question. And then the answer comes that we would expect something in the order of one every 70 years.

So that's about, if you will, the luck factor that we maybe had. So this is obviously an extraordinary finding, a neutrino so energetic. What do you think other researchers are going to make of it, of this finding? I think the paper is really about the experimental measurements. How we really try to make it as simple and robust as possible with the energy measurement as the key part of it. What

people will ask. I mean, thinking about our colleagues in other experiments in previous conversations, this question came up about why have they not seen something while we see this, but we address that. And I think our answer is maybe not fully satisfactory to everybody, but it's a bet we can do at the moment.

Well, it sounds like Chem3Net is off to a very good start. Once you are clocking up more of these energetic neutrinos, what kinds of things do you hope to be able to study about the universe? You can really indeed study the most energetic particle acceleration processes in the universe. So I mentioned already these active galactic nuclei. And the nice thing about observing these neutrinos is that they give really a complementary way of studying such objects.

together with the photons. For example, it might be that with neutrinos, you can really look deep inside these objects, whereas photons are absorbed by all the dust and matter that is around there. KM300 is also hoping to study a lot of neutrino sources in our own galaxies. For example, supernova remnants. So a supernova is an exploding star,

If you wait like a thousand years later, then there's a shell of matter that's been ejected by the supernova that is plowing into the environment. If we could also detect the neutrinos associated with that, that would tell us really a lot more about what exactly is going on there. Well, so you're off to a good start. Who knows what will come in the future? Yeah, I think it's very exciting. Art Heiber there. To read his paper, look out for a link in the show notes.

Coming up, the impact that proposed cuts to NIH grants could have for US science. Right now though, it's time for the research highlights with Noah Baker. Bonobos, one of our closest relatives, can adjust their communication based on what they think other individuals know or don't know.

Researchers assigned a human partner to each of three bonobos. Then, both ape and partner watched the other person hiding a food reward under one of three cups. This reward would be given to the bonobo if its partner could find it, and the bonobos could help their partner by pointing to the correct cup. Now, sometimes the humans saw where the food was hidden, but other times their view of the action was blocked.

When the human partner did not know the food's location, the bonobos pointed more often and more quickly to the cup where the food was hidden, as opposed to when their partner already knew which cup concealed the reward. This suggests that apes, like humans, can spot when social partners are unaware of something and adjust their behaviour accordingly.

If you're unaware of where to find that research, I'll point you towards the Proceedings of the National Academy of Sciences of the United States of America. More than a quarter of a million beads have been recovered from a burial site in southwestern Spain. This hoard, the largest array of such ornaments ever found, adorned the attire of the powerful women who were buried there some 4,800 years ago.

In 2007, researchers discovered a burial containing the remains of several humans, mostly women. Some of these women had been interred wearing tunics and skirts bearing thousands of tiny beads. On the basis of their weight, researchers estimated that there were more than 270,000 of these minuscule beads. Nearly all of the beads were made from the shells of locally harvested cockles, scallops and other shellfish.

The authors estimated that 10 people working eight hours a day would have taken nearly seven months to make them all, a process that would have consumed almost a ton of shellfish. The researchers think the attire might have reflected the high status of these women, who could have been religious or political leaders in copper age Iberia. You can get a bead on that research in Science Advances.

Next up, Benjamin Thompson and reporter Max Kozlov are here with an update on the funding situation in the US. There's a lot going on in the US right now as science organisations and government agencies respond to President Donald Trump's executive orders. Now, Lizzie and I talked last week about the US National Science Foundation's freezing of funding, which then got reversed, but continued scrutinising of grants to comply with directives targeting diversity, equity and inclusion issues.

efforts. We said at the time that the situation in the US was a developing one, and it has continued to develop. And joining me to talk about some of the things that have happened is Nature's Max Kozlov, who's been across all of it. Max, how are you doing today? I'm good. Thanks for having me. Not at all. Well, listen, Max, there was a story that we'd planned to talk about, which actually has changed since I went to bed and woke up. And it's about the National Institutes of Health, which, of course, the biggest biomedical funder

Now, this one isn't related to DEI, but it is related to grant funding. And it sort of came in rather hot on Friday afternoon and caught folk fairly unaware, as I think it's fair to say. Yes, exactly. So what you're referring to is this policy that the NIH put out on Friday evening, and it would start on Monday.

basically slashing something called an indirect cost. Now, it doesn't sound particularly interesting and it honestly isn't, but it is a huge deal because what it does is it slashes billions of dollars of funding each year to be paid.

hundreds of research institutions that NIH funds. Basically, when you get a grant from the NIH, you get money that's called a direct cost. So that's you paying for your research staff, your reagents, your equipment that you use in the lab. But then there's a lot of other things that you also need to do science. You would ideally like to have electricity. You would like to have a building. You would like to have the lights on.

And other things like you need IRB approval, so you need an ethics board to review your study. Things like that that the university has to front the cost of because it's kind of beyond the scope of your specific study. So what indirect costs are, are...

the NIH and other funders trying to pay for some of those other important things that go into a research study. So what happened was they set a cap on the indirect costs rate.

So every university in the United States negotiates their own rate with the NIH's parent organization, because in some places like New York City, rent is more expensive than other places. So they negotiate their own rate. The average is around 40 percent.

It can go as high as 75%. And what this policy does is slash it to 15%, meaning hundreds of research institutions are each missing out on hundreds of millions of dollars, effective two days after the policy was announced. And so there was a huge outcry from the biomedical research community. And it's worth saying that we're recording this on Tuesday, and this was due to come in yesterday on Monday as we're recording this.

And what was the justification then for the NIH reducing the percentage of indirect costs that would be paid to institutions? So basically, President Donald Trump has teamed up with Elon Musk to slash institutions

billions of dollars from the federal government. And so it's another part of that mission to cut billions of dollars from the federal budget. And the Trump administration alleges that universities like Harvard or Yale, these prestigious universities, they receive too high a cut of these indirect costs. But at

At the same time, research regulations for how we take care of animals, how we take care of human participants has increased over the last few decades, meaning that universities need to spend more to comply with all these different regulations. And by putting forward these examples of universities that do have maybe deeper pockets, that's not necessarily the same across the board for all institutions in the US. Exactly. So one scientist who worked in Trump's first administration, he testified in front of Congress

and found that there was really no problem with the rates the way they are before the policy was enacted. And he found that it would be some of the less wealthy institutions that would stand to lose the most from this kind of policy because they don't have a multi-billion dollar endowment to draw from. And also endowments, from what I understand, are difficult to just extract money from. So in writing this story this week,

I talked with many research administrators and researchers who were worried about literally keeping the lights on in their lab as a result of this policy.

But there is more to this story, right? This was expected to take effect on Monday, as you say, very short lead up time. And yet there seems to be, I mean, dare I say a stay of execution? Yeah, the policy was supposed to be enacted on Monday, but a group of states sued the NIH, alleging that the policy is illegal and that it should be blocked. And today, Tuesday, a group of universities also sued the NIH for the same reason.

A judge considered these petitions and issued a nationwide halt on the policy until they could consider both of the petitions. So it is frozen at least until February 21st at the moment. So we still don't know exactly how things are going to play out in this instance, but there is a lot going on surrounding funding at the NIH right now, as I understand.

Trump's NIH, at least, kind of has maintained something of a funding freeze in that somebody crunched the numbers. And in a similar timeframe this year, they've released just less than $5 million in a one-week span in February versus last year, $218 million. And so just to say, like, it seems like some version of a funding freeze is still in effect at NIH, according to this data. And in terms of funding, I mean, you've been covering what other places have been doing

as well, not just government agencies. And there was a story last week about non-government funders and the action that they're taking, in particular, the Howard Hughes Medical Institute, the HHMI. Yes. So researchers were alarmed when they saw a big $60 million program aimed at boosting diversity in science education terminated at HHMI last week.

this is a big deal because HHMI is a private funder, and so they should not be beholden to what's happening at the federal level. Not only was it alarming that HHMI cut this program, but they deleted all mentions of it from their website. And when I asked HHMI the reason for this termination, they never explicitly tied it to the Trump administration, but researchers could not help but notice the parallels and the timing

especially after every single government agency has taken down any websites that have the word diversity, racial inclusion on them. So you can see why researchers would be upset for an institution that has a $24 billion endowment to all of a sudden cut support from an area that researchers say is crucial right now. And they were hoping that private funders would be a safe haven amid all this chaos and confusion at the federal level.

And this comes among a broader targeting of DEI activities, but it's not just funding, right? Datasets, for example, are being looked at as well. This is something you've looked at too. Yes, entire websites, resources, datasets, especially from the Centers for Disease Control and Prevention have been taken down. And I think

It's abstract to think about data being taken down. Let me give you a tangible example of why researchers have been very worried about data going missing. So in 2022, there was an outbreak of mpox or monkeypox around the world. And part of the reason we were able to stop it so early in the United States is because we had data suggesting what populations were most at risk. And so we could target our resources and our attention to making sure that those communities had access

the resources they needed, they had vaccines. And you can see it in the graph, the number of infections peaked and then sharply declined right after. So researchers are saying that without some of these data sets or without some of the identifiers such as gender, sexual orientation,

you name it, we'd be kneecapping ourselves for no reason. And it would make it harder to stop infectious disease outbreaks in the future. And so folk are, in a sense, taking it into their own hands to try and back up these data sets on thumb drives in their office, that sort of thing. Yes, researchers have banded together on social media to work together to do whatever they can to archive

the old versions of these websites and to download the data sets before they go missing. I'm not sure if they've managed to capture every single data set, but I think they did a pretty comprehensive job trying to at least archive the data that existed prior to the new administration coming in. And the impact on DEI-related activities seems incredibly wide-reaching. I think that is intentional. Trump's executive order banning corporations

quote-unquote, woke policies and gender ideology was meant to be quite broad, and it's up to each agency to interpret that as they will. But one thing is very clear. The effects have been so far-reaching that researchers who study things that have nothing to do with, quote-unquote, gender ideology have been impacted. I'm talking about researchers studying women's health, the science of menopause, say. I've talked to researchers whose studies have been flagged

because they might include some of the same keywords that were included in Trump's executive order. And of course, there is a lot going on here and more stories to follow, no doubt. But you and the rest of our colleagues who have been covering this have spoken to many, many researchers online.

across the US and across the world, of course. What are they saying about the impact that all this is having? What's the general sense that you're getting? It's been a sense of chaos and confusion because news happens on a minute-by-minute basis. And so it's been like that for three weeks straight where huge numbers

wrecking ball of a policy gets announced and then it's tempered slightly by the courts. And we'll see to what extent that continues to happen. But I think the overall sense is chaos and confusion, even for scientists within some of the agencies. I think that there hasn't been very clear top down communication. And so it's hard for people within NIH, let alone the entire scientific community, to make sense of what's happening right now.

Well, Max, I know you and the rest of the team will be keeping abreast of all of this as it happens, but let's leave it there for the time being. Max Koslov, thank you so much for joining me today. Thank you for having me.

Nature's Max Kozlov there. To keep up with all of nature's coverage of the ongoing situation in the US, head over to nature.com slash news. Finally on the show, it's time for the briefing chat, where we discuss a couple of articles that have been highlighted in the Nature Briefing. Dan, what have you been reading this week? Well, I've been looking at a story originally published in the New York Times about asteroid 2024 YR4, which might be on a collision course with Earth, or not...

The probability of impact keeps changing. And that's what the story is about. It's about why this percentage chance that this asteroid might hit the Earth in 2032 keeps fluctuating. I mean, might hit the Earth does sound quite alarming. So maybe tell me what are the percentage chance that it might occur and why it keeps changing? Well, since it was observed in December, it's been...

fluctuating between 1.3% chance of hitting the Earth and as recently as February 7th, it was measured at 2.2%. So you might read that as the chance that this asteroid is going to hit the Earth has doubled. What the story pains to try and explain is that actually what's happening is astronomers are getting a better understanding of

of where this asteroid is going, what its actual orbit is. I mean, to me, 1%, 2%, that still sounds like quite a high chance, but you seem quite calm. So is the expectation that as they understand more about its path, that those percentage chance will go down further and maybe it won't hit at all? I mean, potentially. So how they describe this in the article is the asteroid's many possible future orbits are sort of like a spotlight, right?

shining out into space and with a sort of wide cone of where that's going and at the moment earth falls inside that spotlight but as more observations come in as the astronomers watch the asteroid for longer and they'll be able to observe the asteroid until april when it will be too far away and too faint that spotlight becomes more and more focused and hopefully it seems like optimistically earth will fall out of the spotlight and we won't be on a collision course with this asteroid in

in 2034. I mean, I've never been quite so happy to not be in a spotlight for something, I think. But, you know, there is obviously still a small chance of this occurring several years in the future. What would happen then? Well, the observations suggest that this asteroid is between 40 and 90 metres across, and

And this article suggests that the impact would be similar to a nuclear bomb. The current uncertainty over the orbit, though, means that the exact locations where that would fall aren't known. This article mentions a mix of uninhabited and sparsely populated areas, as well as some densely populated areas. So it could be the Eastern Pacific Ocean, or it could be South Asia. And so when might we know more about this asteroid? Well, as I mentioned, it'll be observable from Earth until April this year.

And if that's not enough time to kind of get a good understanding of exactly where it's going, the next flyby will be in 2028. OK, well, we'll keep an eye on it then because I'm sure many of our listeners will want to know if an asteroid is going to hit Earth. Although I must say, it seems like it's very unlikely.

For my story this week, it's sort of related to probabilities and mathematics because I've been reading a story in Nature about an AI that's gotten really good at maths. Isn't that what computers do anyway? How is this different from the regular maths, you know, my calculator can do? I mean, that's a good question. It was my first thought as well because ordinarily computers are great at calculation. You give them numbers, they can give you totals, sums, all sorts of things.

but this is about a specific kind of maths known as the International Mathematical Olympiad and in this challenge there are a lot of different problems that require quite deep reasoning and understanding in order to work them out because essentially they're like big logical puzzles and normally when we create computers we tell them how to work through logical puzzles but

This is something that's a lot more tricky. It's not just like adding numbers up and that sort of thing. And in particular, this AI, known as Alpha Geometry 2, has focused on geometry, which is difficult for computers and AIs because it involves imagining shapes and things in space. So what helps this AI fare better at these sorts of puzzles than, you know, we've all seen...

AIs fail miserably at the simplest of riddles or hallucinate answers. Why can this one crack these problems? Yeah, so this AI is quite different to large language models and other AIs that people might be familiar with. And it's actually an AI we talked about about a year ago on the podcast, although that was an earlier iteration of the AI. And basically it has two distinct systems that

that allow it to do sort of deep reasoning, so really logically thinking about things, and also allows it to have a little bit of creativity to try different things, to try stuff out and see what works. And those systems come together to allow it to tackle geometry problems.

The way in which they've improved it with this newest version, this Alpha Geometry 2, so it's now got Gemini, which is Google's large language model, incorporated into it. And this AI itself has come out of DeepMind, which works with Google on these problems. And this integration allows it to

basically speak a more formal mathematical language and check for problems in its rigour so it can basically weed out its own hallucinations by going through and sort of checking what it's done and it's also now able to move geometric objects around so it can really visualise how shapes are working and maybe move points on a triangle and that sort of thing and

And with all of this, it's now improved its ability to solve these geometry problems from the International Mathematical Olympiad by 30%. So a year ago, it could solve 54% of these problems. Now it's able to solve 84% of the problems worldwide.

which would put it at sort of a gold medal level standard for geometry problems at least. So what applications might this have apart from putting high school mathematicians out of a job? Well, one thing it could do is it could help mathematicians work through complicated mathematical problems. It could try to have the first go at them because it also generates a proof. So it shows how it's worked through a problem that mathematicians could then use.

The other aspect of this is by solving these sort of problems, DeepMind are showing how you can create an AI model that's able to reason and logically think about things, which many people consider as the first step to more sort of abstract reasoning and more general purpose AIs. So is this AI reasoning in a way that would be relatable to a human being working out these problems? Or is it, you know, able to

just brute force the solution by imagining a thousand different answers and testing itself against them. Yeah, so it's a bit of both. It does sort of brute force because it can try a lot of different things very, very quickly, much faster than a human could try. But also it does generate these proofs, these outputs that then a human could work through. And previously, when I talked to the researchers behind it, they said that

It does do that, but the proofs are very long. They're much longer than a human would come up with. So it's like a hundred steps to solve a problem versus maybe, you know, 10 or something that a human would come up with, but still interpretable by the humans. This is from DeepMind, so I...

assume it's not open source. Is anyone doing this with an open source model that we could actually have a look inside? Yeah, you're right. It is not open source coming from DeepMind. But if anyone listening out there fancies giving this a go with an open source model, you could win $5 million because there is an AI Mathematical Olympiad prize that will go to any AI systems that are open source that

can solve these international mathematical Olympiad problems at a gold medal level. Well, that's exciting for open source researchers. What's next, though, for this model? Well, the obvious next step is to broaden this out to more problems and not just geometry and try and solve all the problems of the international mathematical Olympiad. And there is a AI that goes along with this called AlphaProof, which is working on those problems in addition to the geometry problems.

And researchers expect that it won't be long before AIs are able to do this and tackle the entire International Mathematical Olympiad. The real next test will be the next test, because one of the problems with designing systems like this is maybe in the training data, you've accidentally given it the answer. So fresh problems that no one has ever seen before will be the real way to put these AIs to the test, and that'll be happening immediately.

in July this year. Well hopefully those results will get published and we'll be able to see how well this AI can do against brand new problems. Definitely it's one that we'll be keeping an eye on but for now I think that's all we've got time for on the briefing chat so listeners if you want more on those stories and you want to sign up to a nature briefing to get more like them direct to your inbox check out the show notes for some links. That's all for this week as always you can keep in touch with us on email we're podcast at nature.com you'll also find us on bluesky and x.

I'm Dan Fox. And I'm Nick Pachaciao. Thanks for listening.

With new line on my plan. Additional terms apply for trade-in and pay off your phone offer. See Verizon.com for details. To make switching to the new Boost Mobile risk-free, we're offering a 30-day money-back guarantee. So why wouldn't you switch from Verizon or T-Mobile? Because you have nothing to lose. Boost Mobile is offering a 30-day money-back guarantee. No, I asked why wouldn't you switch from Verizon or T-Mobile. Oh. Wouldn't. Uh...

Because you love wasting money as a way to punish yourself because your mother never showed you enough love as a child? Whoa, easy there. Yeah. Applies to online activations. Requires port in and auto pay. Customers activating in stores may be charged non-refundable activation fees.