AI to solve the energy crisis?

Nuclear fusion and AI supercomputers

0:00This BBC podcast is supported by ads outside the UK.

0:06Decisions made in Washington can affect your portfolio every day. But what policy changes should investors be watching? Listen to Washington Wise, an original podcast for investors from Charles Schwab to hear the stories making news in Washington right now. Host Mike Townsend, Charles Schwab's Managing Director for Legislative and Regulatory Affairs, takes a non-partisan look at the stories that matter most to investors, including policy initiatives for retirement savings, taxes and trade, inflation concerns, the Federal Reserve, and how regulatory developments can affect companies,

0:41sectors, and even the entire market. Mike and his guests offer their perspective on how policy changes could affect what you do with your portfolio. Download the latest episode and follow at schwab.com slash Washington Wise, or wherever you listen. Running a business shouldn't feel like surviving a software group project. One app for accounting, another for inventory, another for sales. And somehow, none of them talk to each other. That's where Odoo comes in. An all-in-one business management software that brings every part of your

1:14business together. From sales and accounting to inventory and marketing, all in one powerful platform. No messy integrations. No bouncing between tabs. And best of all, no spreadsheets. Stop managing software and start managing your business with one unified system. Try for free today at odoo.com. That's O-D-O-O dot com.

1:37Welcome to Inside Science from the BBC World Service. I'm Tom Whipple. Now, last week, we might have left you with the impression that AI is going to kill us all. We did that by speaking to people who said that AI is going to kill us all. But what if it's not? What's the glass half full option? This week, we are exploring one of the many ways in which AI might actually help us, through, in this case, facilitating nuclear fusion. We're also going to discuss a more retro, old-school sort of techno threat. Remember

2:13that distant time several weeks ago when techno panics were all about social media? Well, we speak to the author of a book about how the internet might be making us all sicker. And how about better understanding the data infrastructure of HI? Yes, HI, human intelligence. How do we store data? Or, as we used to call it, memories. Then there is the formidable HI that is Lizzie Gibney from the journal Nature. And she's been looking at some of the other news. Lizzie, what have you got for us?

2:46I've got some exciting hints of an ancient Egyptian discovery. OK, fabulous. Well, amid all of the AI apocalypse news, you might have missed a rather different AI announcement. It concerns nuclear fusion and plans to bring together some of the computing and data on both sides of the Atlantic. But before we get to that, maybe a primer. Lizzie, nuclear fusion, bit of a long-term goal, would be quite an energy panacea if we get it.

3:19Absolutely. It would be incredible. It's one of those things where we know it works because it happens in the sun, right? So we have hydrogen atoms and they fuse to make helium and we get energy out of it. Doing that on Earth is actually very, very difficult. But there are a huge number of companies and governments who are trying to make it happen right now. So the reason we want to do this on Earth is because the sun manages to have a pretty low carbon footprint and make lots and lots of energy. Exactly. It's a way of getting out energy and it produces no carbon emissions whatsoever. So it sounds great. So why haven't we done it?

3:52Because it's really, really, really hard. So you need to make these conditions that are absolutely extreme, huge heats, talking like 100 million degrees, heat, pressure, you need to get these hydrogen nuclei to actually hit each other with enough energy to be able to fuse. And there are two different ways that they're trying to do it on Earth at the moment. One is you kind of zap these fuel pellets with a laser and they have a shock wave and you get fusion there or you superheat the fuels to make a plasma. And that's where the nuclei kind of

4:25separate from their electrons. And you have this very, very hot gas and you confine it with magnets and you can kind of control that and get fusion to happen in the middle of that melee. So controlling this plasma and understanding it is a huge problem. And that's where the announcement this week comes out. The UK and US are combining the brains of the nuclear fusion supercomputers, each of them fed with data from the national fusion experiments in the hope that by doing so, we can basically better use AI to model the behaviour of the

4:57plasma. Now, Tim Bestwick is the UK Atomic Energy Authority's Interim Chief Executive Officer. We spoke to him about why controlling plasma is so hard. Because it's a really big, complicated challenge, in short, all these complex processes in the plasma. We've talked about the fact it's held in a complex magnetic field. I mean, this isn't just a single magnet. This is lots of magnets which have to be controlled and produce a magnetic field arrangement that holds this plasma in the case of the machines we make in a shape of a ring

5:33donut, a torus. And they have to hold this plasma in the shape of a torus, all these things going on. It wants to do all kinds of funny things when the magnets are there to control it and hold it in that. So that control and stability and confinement is really important. So you can see that learning from those conditions is important and learning in such a way that you can make predictions is important. But there isn't a huge amount of data to go on. So each experiment produces a set. So, you know, clearly, if we can work with other machines that produce

6:05more data and share, pool our data, we could get to learn more. Now, fusion is quite challenging for that learning environment because in other parts of technology, we're used to the idea of machine learning on subjects where there's a huge amount of data to learn from. And, you know, you're swamped with data. Our data environment is a bit more sparse by comparison. So every bit of experimental data really does matter. So, you know, another reason for joining forces with another machine is developing similar experimental results we can feed into the models.

6:37Okay. So what do you mean by modelling the plasma? You've got this plasma that this happens in. What are your computers trying to do? And why is this so hard that you're linking up to the US to try to combine forces? So the plasma has to be contained. And in the case of what we are trying to do and our fellow labs in America, in the main, we're trying to do something called magnetically confined fusion, because it has to get so hot to make these fusion processes work that you couldn't hold it in anything physical.

7:11So it's so hot, it needs to be confined by magnet. So confinement is important. Now, you also have to get it very hot. I mean, really hot. Actually, in the jet experiment, which we operated UK AEA and produced fusion reactions, we were getting to well over 100 million degrees in the centre of the plasma. So probably near 150 million degrees, which as an aside, we were one of the hottest places in the solar system. And that includes the sun. So when you've got something that hot, so it has to be held in a magnetic field,

7:44it has to be stably held in a magnetic field, because a plasma is a bit like one of those children's party balloons, you know, you squeeze it in one place, and it pops out somewhere else. But there's then just the challenge of just, you might wonder how you get something as hot as 150 million degrees. And in the case of the jet experiment, we had a few grams of gas, and we were injecting more than 30 megawatts of heat into that. So huge amount of power going in, heat up this small amount of gas, gets very hot. It's in a plasma, it's ionised in a plasma,

8:15the fusion reactions start happening. So the modelling is about how you hold it in the magnetic field, how you hold it and keep it stable in the magnetic field. Actually, when you get in the power, how things move around inside this complex mix of electrons, positively charged ions, lots of heating flow coming in, the deuterium and tritium fuel, the helium and the neutrons coming out of it, there's just a lot going on. And so holding that all in a model is complex. Okay, I'm going to, I know that this question is going to annoy you a lot, but I know that when I

8:50go in to the Inside Science mailbox on Monday, I'm going to get someone emailing in saying, 30 years ago, I was told that fusion was 30 years away. It's still 30 years away.

9:03Fusion scientists are asked this sort of thing all the time. Is this going to make it happen? Or are we going to keep on finding more and more obstacles? Are we going to get fusion?

9:13So obviously, at this point, I have to say, what do we mean by it? Because that, I know it sounds pedantic, it's quite important, come back. So we've sort of demonstrated that you can make fusion happen. In more than one place has demonstrated that power comes out of a fusion reaction. What we haven't done so far, we, the worldwide community, is show that you can get more power out than goes in. So some sort of gain, that's the next step. So that's another landmark. Then there will be the landmark of making a prototype fusion power plant. And we can all say,

9:44oh, that kind of works. Then we'll have to make what they call the first of a kind. So a power station that you go, right, that's it. We'll make a whole lot more of those. Then we have to deploy that and roll it out and make lots of power plants. And then they have to make a difference to the global energy supply. So each of those is a different date. You know, we've sort of done, yes, you can make fusion happen in the lab. The next step will be fusion gain. The step after that will be, here's a prototype. And then, and then, and then. I think, you know, for a lot of us,

10:16what really matters is making a difference to energy supply. You know, actually some of the electrons that are keeping our lights on coming from the fusion power plant, that is some way off because we have to go through those steps. And there's time between each of those. I, I won't be drawn just simply because I don't know when fusion will make a difference to powering our world. But I, it's my belief it's in the second half of the century. That was Tim Bestwick. Lizzie, as I alluded, the sort of lazy thing to always say about fusion is

10:48it's 30 years away and always will be. And in that interview, we basically got a prediction that it's 30 years away. That slightly belies all of the stuff. There is a lot going on in fusion now, isn't there? There's so much going on. So there was a big result in 2022 that made the whole community very excited. So this was the national ignition facility in the US. They, they got more out than in this idea of gain, but it wasn't for the whole facility. Like it was the direct amount of energy from the laser. Yeah, there's a massive caveat on that. Yes, exactly. The direct energy in was less than the energy out, but that didn't factor in all the

11:20electricity it takes to run the laser and the whole facility. And so we haven't achieved that step yet, but that excited people a lot. And so we've had this huge injection of cash, about $10 billion worth investment into fusion companies in the last 10 years. And obviously that's, that's great in many ways. Like it's really pushed motivation. We actually have cash to do these experiments and for different companies to be trying different methods. But it's also slightly inflated expectations,

11:52I would say. And obviously, if you're a founder of one of these companies, you want to get more investment. And it's in your interest to say that it'll be within the next five or 10 years that you're delivering electricity to the grid. And that does seem to be unlikely, I would say. And I'd say I agree with Tim on that much more sober assessment. Well, look, it's unsurprisingly, it's hard to make a sun on Earth. But Lizzie, look, another man who flirted with the awesome power of the sun was Tutankhamun, whose name is linked to the sun god Aman Ra. And that link

12:26brings us entirely naturally to your first story.

Searching for Nefertiti's tomb

12:29That's right. This is a very exciting story. So this is a exclusive in Nature by Joe Marchant. And it's from a bunch of Egyptian researchers who published in the occasional papers of the Amarna Royal Tombs Project. The big news is they think they may have found Nefertiti's tomb. And this is a very kind of tentative may, a big maybe. So Nefertiti, if you remember, she was this super powerful Egyptian queen from the 14th century BC, possibly actually a pharaoh herself. And she's who that

13:01model, that bust is made of, you know, with the queen with the long neck, who's seemingly very beautiful. That's kind of why the world knows about Nefertiti. The mystery is her tomb has never been found. Now, this study follows up on a theory from 2015 that perhaps Nefertiti was buried with Tutankhamun. So he was her successor, possibly her son or stepson. It's a bit unclear. And the reason they think that might be the case is Tutankhamun's tomb is like oddly small and that the walls kind of have cracks in them. There's some suggestion that maybe there's a false wall there. And so the

13:34idea is maybe Nefertiti was buried first. And then when Tutankhamun died unexpectedly early, he died just aged 19. Maybe they kind of boarded up her tomb and expanded on the outer chambers to make his. So at the time, about 10 years ago, you know, geophysicists looked but didn't find any evidence for that. But now we have some tentative evidence that perhaps there is a network of corridors and chambers behind Tutankhamun's tomb. I mean, that's absolutely astonishing. Are they going to, what's the evidence and are they going to bash it open? So the evidence comes from

14:09microgravity surveys. It's amazing that you can do this. It's like you're looking for tiny little variations in gravitational field strength. So obviously, matter has a gravitational pull. And if there's less matter than you expect that you have a change in the field strength. So they map this area and they can see where there's areas where it seems to be less dense, which, you know, could be a chamber. It's not like just doing an x-ray. You don't get a nice clear picture. There's a lot of potential noise from traffic, any kind of vibrations. You've got to factor in the shape of the terrain and the resolution. So that's why it's a very tentative maybe. But what they are seeing is it

14:45looks like you can see some kind of some right angles and some walls that are parallel to other existing walls. So there's definitely some signs that there could be these rooms behind Tutankhamun's tomb. This has got everybody very excited. But what these researchers would like to do is carefully drill a really tiny hole, like a few centimetres wide, into one of the areas that they think, you know, that is an empty chamber they think is not actually the tomb itself.

15:17And they would carefully drill this tiny hole and put a camera down there and essentially see if their theory, if these areas of low density actually are the chambers that they think are there. So that would be the next step. Excellent. Well, I hope the nature officers are ready for the vengeance of the mummies.

15:38You're listening to Inside Science from the BBC World Service. Tell us what science you think we should be investigating. Our email address is insidescience at bbc.co. Hey, what's up? This is Cheekies from Soul Searching with Cheekies. You know those moments when you're out with friends? Everyone is catching up. Your food hasn't made it to the table yet, but your ice-cold Coca-Cola Zero Sugar arrives first. You tell yourself you're just going to enjoy one sip while you wait.

16:08Then the conversation gets going. Somebody tells a funny story. You take another sip here and there, and before you know it, your food shows up and you're thinking, wait, where did my drink go? Well, we've all been there. It's one of those little moments that just seems to happen when you're enjoying good food and even better company. Coca-Cola Zero Sugar has the real Coca-Cola taste you love with zero sugar. It's one of those drinks that's hard to resist, whether you're catching up with friends, grabbing lunch, or sitting down to dinner. And since it's zero sugar, you can always order another when you're ready.

16:39That's exactly why it's just too good to wait. Try it for yourself. Coca-Cola Zero Sugar. Real Coke taste. Zero sugar. It's too good to wait. Choiceology, an original podcast from Charles Schwab, is a show about the psychology and economics behind our decisions. Join host Katie Milkman, an award-winning behavioral scientist and author of the best-selling book, How to Change, as she shares true stories from Nobel laureates, authors, athletes, and everyday people about why we make the choices we do and how to make better ones to help avoid costly mistakes.

17:12Listen to choiceology at schwab.com slash podcast, or wherever you listen.

17:20UK.

17:22Now, there is under a fortnight to go. Ten sleeps left, if you're not too excited to sleep. At the Royal Society, Britain's premier scientific institution, they are getting in the champagne, they are hoovering the red carpet, they are bracing for their annual non-fiction bacchanal. Yes, it is the book prize ceremony that is to science-writing parties what backstage at Woodstock was to rock star parties. The evening that science celebrities speak of like actual celebrities speak of the Oscars.

17:56It is, of course, the Royal Society's Trevadi Book Prize. Each week, we interview one of the shortlisted authors. But to acknowledge its imminent arrival, this week we're doing two.

The internet and health trends

18:08First up is Deborah Cohen, who spoke to us about bad influence, her investigation of the internet's effects on our health. My book is basically about how the internet is changing the way we think about our health. Essentially, it asks what happens when you put health on platforms that are ostensibly designed for entertainment and shopping. And it's basically a poorly regulated advertising space. So it might be one thing if you're shopping for your trainers or your lipstick or whatever on social media. It's a totally different thing when you're looking for health.

18:41And we're surrounded by more and more health information than ever before. And it's increasingly difficult to know who or what to trust. So it covers everything from ADHD and menopause to weight loss drugs, fertility tests and AI. Let's go through some of the examples, because one of the really striking ones was something that you found in teenage girls. And it was these sort of tics that came from TikTok to not oversimplify too much. Tell us about that. Something I don't think we talk about enough in medicine is the mind and body.

19:14The way our minds and bodies are linked, we've tended to follow more recent terms, a very sort of biomedical approach to medicine. And actually, in the pandemic, there was something that was done, TikTok tics. And it sort of started when paediatric neurologists started seeing young women presenting with Tourette's-like tics, except they weren't your typical Tourette's. They were explosive. They tended to be in women, where Tourette's typically is in younger boys and different kinds of symptom presentation.

19:45It wasn't these explosive, florid tics. And they were repeating certain words, like beans and things like that. They really didn't have a clue what was going on. And then they found paediatric neurologists and other countries were seeing similar things and largely in Anglophone countries. And eventually, they traced it down to a trend on social media. So there were a group of influencers that were demonstrating these very florid tics that these young girls were picking up.

20:16And that wasn't to say they were faking it. The tics were real. And then they had to sort of work out what was really going on. Now, if you think about what the pandemic was like, it was a very abnormal situation. People weren't mixing with their peers. There was isolation. They were surrounded by death and dying and anxiety. And one theory was that actually these tics were a way of expressing what was going on psychologically for them. But what it showed was how what we see and watch

20:47can have a massive influence over how we express symptoms and how we express ourselves and how illness, perhaps, or health conditions are expressed. You talk about influencers and the role of social media. Some of those seem to be persuading people to take up full body scanners, which might be expensive, might be unnecessary, but it's not a bad thing, is it? So full body MRI and what we're talking about screening here.

21:18So people that are otherwise well and symptom free are being sold more and more screening, more and more of these interventions. So the title of that chapter is Sick Until Proven Healthy. Normally, we're perfectly fine. Then we've got symptoms. We might go and seek medical help. But increasingly, a lot of the marketing is about convincing us we might have something inside that is going to kill us and we just don't know about it and we need to detect it. And people will have heard during the pandemic about false positives and false negatives.

21:49And what's happening is tech becomes more and more sensitive. It's able to pick stuff up. But diagnosis and prognosis hasn't caught up. We still don't really know what a lot of this stuff means. And so the risks are that you pick up on it. And they're nicknamed incidental owners. That you pick up on these blobs or these shading, these patterns, and you have to, you can't ignore it. You have to then go and further investigate it. And that leads to more and more testing, potentially unnecessary interventions.

22:20And you can go down this whole clinical cascade that is potentially very expensive, although the burden often falls on the NHS. So the taxpayer might pick it up. And then no intervention is risk-free. And I think what's interesting is we're used to talking about side effects of drugs and the benefits and harms of drugs. We don't think about that so much when it comes to testing. There isn't one medical authority around the world that recommends full-body MRI for screening purposes. What's your conclusion? What do we do about this?

22:51It's time to ban all social media. I don't think there's a one-size-fits-all. And I think this is mass experiments in public health going on. We don't really understand the impact of it, how it's making us think about our health, our symptoms, the difference between ill and healthy, our relationship with the health service, our relationship with our health providers. It really is creating this quite messy space. Deborah Cohn talking about social media and influences

The science of memory manipulation

23:18and their effect on our health. Now, from some bad science to some utterly fascinating cutting-edge science, Steve Ramirez is author of How to Change a Memory. And he is another of the shortlisted authors at the Royal Society Trivedi Book Prize. The book is about the science of memory manipulation and all the ways that contemporary neuroscience has successfully been able to manipulate memories to restore health back to the brain and to understand how memory works.

23:49And throughout the book, I talk about my friendship with my late friend, Xu, who made these seminal co-discoveries with me. We teamed up together on trying to artificially activate memories. So my friendship with Xu is the beating heart of the book, and we use that beating heart to set the pace for all of the memory manipulation science that I unpack throughout. Okay, so you've mentioned that you're not just an author, you're a researcher who has made discoveries,

24:20quite important discoveries in this field. So you conducted these series of experiments about effectively manipulating the memories of mice. Can you tell us what you found? In 2012, we had found that there are cells in the brain that hold on to individual memories, which when artificially activated is enough to jumpstart that memory itself. So that was the first million dollar finding that Xu and I had made was that artificially activating a fraction of cells

24:52that held on to a memory is enough to jumpstart that memory. In 2013, we went on to show that not only could we activate a memory, but we could activate a memory and incorporate new information into that memory. And that's what we called creating a false memory. And the beauty of that is that it really set the stage, in my opinion, for one of the most potent and relevant and applicable findings in 2015 from our work, which was that could we activate memories

25:23in a way that would actually be beneficial for the organism, almost like a drug or a treatment that's used to try to restore health to the brain. And what we did was we were able to artificially activate a positive memory in the mouse brain to curb symptoms associated with depression and anxiety. And that worked. So that trifecta to us completed this trilogy of papers of activating memories, implanting false memories, and modulating positive memories to alleviate symptoms associated with psychiatric disease.

25:57And what is the... I'm going to ask a question which I think to listeners will sound a bit silly, but having read your book, I think it's not. I think it's quite... It goes quite deep. What's the purpose of memory? Why do you think we have this? If you recall a particular memory, there's areas of the brain that turn on, such as the hippocampus, which is the area that I study. Now, if you start to imagine yourself in the future, like what are you going to do tomorrow, or maybe the path you're going to take to work,

26:28or what you're going to have for dinner, that when you're imagining yourself in the future, that also taps into the hippocampus in some of our brain's memory systems, quote-unquote. So this is another way of saying that when we think of the past or when we imagine ourselves in the future, we're still engaging some of these processes in the brain that we think are involved in memory. So perhaps we use memories as building blocks that can help us imagine or configure an event that hasn't happened yet.

26:58Because if we can stay one step ahead of the present, we're one step ahead of the thing that almost ate us, and we remember not to go that way anymore. So there's some evolutionarily adaptive purposes, we think, for why we have the amazing cognitive faculty that memory is. You mentioned that this is more than just a science book. It's a part memoir. It's about your grief over a friend and collaborator's death. And it's about your addiction to alcohol. To what extent did grief and addiction

27:30inform your view of what memory is? And to what extent did your knowledge of what memory is help you analyse your grief and addiction? To me, I learned that grief and addiction and memory are all part of the same learning process in the brain, of the brain trying to adapt to extraordinary circumstances in this case. When I was addicted to alcohol, so much of my headspace was living in the past and trying to understand the unfairness of death,

28:01for example. That why would my friend, who was this extraordinary scientist, pass away in his late 30s, seemingly out of the blue? And then how do I continue in the field where I was used to tag-teaming with my lab partner every single project and presentation and paper that we had ever worked on? Even though these are really, of course, very fundamentally human and deep and heavy topics, the reason I couldn't finish the book until I had found some resolution in these

28:32is because I felt like my... I'm a very cup-is-three-quarters-full person, as are my parents. And to me, it felt like I hadn't yet found a reason for why I had endured what I had endured. And through this process of writing the book, it really felt like I'd not just extracted different reasons and different purposes for being as a scientist and for reconciling what life and death mean to me, but it was even more empowering to share not just this work and this story, but to share the...

29:03It sounds maybe off-color, but almost the upside of experiencing something so profoundly difficult, because that upside, I think, imbued me with a very different kind of respect and resilience for what life and memory really are. That was Steve Ramirez. And if anything today has sparked your interest, please do search for BBC Inside Science and follow the links to The Open University to discover more.

Closing remarks

29:27And that's pretty much all we have time for. You've been listening to BBC Inside Science with me, Tom Whipple. Until next week, it's goodbye from me and goodbye from Lizzie. Goodbye.

30:02Now, all of Sheeran's supporting acts have dropped out too. For more, listen to The Global Story on BBC.com or wherever you get your pods. THE BUSINESS THE BUSINESS MANDO WORRY THE BUSINESS Is the ultimate justice along with the oyster or to dimple or to the skull

30:32and the Like-Craft and theelia and theeese

30:50And you'll see each other