Beyond AI: the top-10 tech of 2023 set to change our lives

This transcript has been generated using speech recognition software and may contain errors. Please check its accuracy against the audio.

Greta Keenan, Lead, Strategic Impact and Communications, Centre for the Fourth Industrial Revolution, World Economic Forum: The top ten Emerging Technologies Report brings together the insights from over 90 experts across 20 countries to list the breakthrough technologies poised to have dramatic impact on people and planet in the next 3 to 5 years.

Robin Pomeroy, host, Radio Davos: Welcome to Radio Davos, the podcast from the World Economic Forum that looks at the biggest challenges and how we might solve them.

This week: what are the technologies that are about to change our lives? We have a top ten.

Mariette DiChristina, Dean and Professor of the Practice in Journalism, Boston University College of Communication: Spatial omics.

Bernie Meyerson, Chief Innovation Officer Emeritus, IBM: A Phage is a delivery vehicle of information.

Mariette DiChristina: Wearable plant sensors.

Robin Pomeroy: When it comes to life changing tech, everyone’s talking about artificial intelligence - and yes, there’s plenty of that in the World Economic Forum’s top ten tech.

Mariette DiChristina: The power of data science, AI and computing is really throughout the entire list.

Robin Pomeroy: That list, compiled every year by the Forum, has tech that could give us big advances in climate change mitigation, agriculture, and mental and physical health.

Bernie Meyerson: The impact would be almost unimaginable to people who are essentially quadriplegic

Robin Pomeroy: We hear from the two people who led the work compiling the list as they talk us through the top 10 technologies that have met stringent criteria.

Mariette DiChristina: We talk a lot about is it truly novel? Is it really building in a significant way? Does it have the power to disrupt?

Robin Pomeroy: Subscribe to Radio Davos wherever you get your podcasts, or visit wef.ch/podcasts where you will also find our sister podcasts, Meet the Leader, Agenda Dialogues and the World Economic Forum Book Club Podcast.

I’m Robin Pomeroy at the World Economic Forum, and with this look at the top 10 emerging technologies of 2023...

Mariette DiChristina: It is amazing technology.

Robin Pomeroy: This is Radio Davos.

Robin Pomeroy: Welcome to Radio Davos, the weekly podcast from the World Economic Forum that looks at the world's biggest problems and how we might solve them.

In this special episode on technology, we look at the technological breakthroughs that could help us solve some of those problems.

To look into that, I'm joined by my colleague, Greta Keenan. Hi Greta, how are you?

Greta Keenan: Hi, Robin, thanks for having me.

Robin Pomeroy: You work at the Centre for the Fourth Industrial Revolution at the World Economic Forum. You're very heavily involved in the report we're going to be talking about today. Tell us about this report.

Greta Keenan: Absolutely. So the Top Ten Emerging Technologies Report is an annual report from the World Economic Forum that brings together the insights from over 90 experts in 20 countries to list the breakthrough technologies that are poised to have dramatic impact on people and planet in the next 3 to 5 years.

This report builds on a long history of ten reports in the past. So since 2011, it's identified little-known technologies that then went on to have a huge societal impact. One such example would be genomic vaccines or mRNA vaccines, which then became just a few years later, the technology underpinning the majority of COVID vaccines worldwide.

Another such example would be AI-led molecular design, which we put on the report in 2018. And then a few short years later, we saw those first AI-designed drugs enter clinical trials.

So we're really excited to build on the legacy of that report over the last ten years by bringing the 2023 edition of the report out, which we launched just a few short weeks ago in collaboration with Frontiers.

And the report, as it has each year, really helps to tell experts or give them insights across sectors and industries into what are these emerging technologies, what's the potential impact they might have, and then how they can help shape a positive impact for industry and society.

Robin Pomeroy: Yes, and we've looked at this on Radio Davos in the past. We have done two episodes before on previous editions of this one. People can go back and listen to those and see if we were right in those forecasts.

And, I'm delighted to say we've got the same guests on this time as we had on last time. Greta, could you introduce them to us?

Greta Keenan: Absolutely. So behind this report is a fantastic steering group of experts who have the hard job every year of whittling down almost 100 different technology nominations down to the ten that make it into the report each year.

We're really pleased today to be joined by the two people who've really been championing that effort over the last decade as the co-chairs of our steering group. We have today joining us, Mariette DiChristina, Dean and Professor of the Practice of Journalism at Boston University College of Communication; and Bernie Mason, chief innovation officer emeritus at IBM. Thank you both for joining us

Robin Pomeroy: Great to see you both again. We're going to go through this list one by one, these ten technologies that you've helped pick out. Bernie, could you tell us what are the criteria you use when you're whittling down just to ten technologies each year?

Bernie Meyerson: At the end of the day, it's scale and impact. If you think about it, you're looking at what's going to have a profound impact on society, on technology, on science, on business. It really is going to be the scale of that impact that we look for, because at the end of the day, remember the word we use in this is 'emerging'.

Emerging literally means that the technology has to grow to some significance within the three to five-year period. So those are our metrics.

Then based upon that, you have to do the very difficult job of forecasting the future, which Mariette and I have struggled with for years, where you look at all of these excellent proposals and say which ones are going to quote "win" and emerge. And the group we have is fabulous, they've really done an extraordinary job in terms of helping us whittle it down.

Mariette DiChristina: I'd just like to add quickly to what Bernie said. He got that absolutely right about the scaling, and one way we probe that is we talk a lot about is it truly novel. Is it really building in a significant way? Does it have the power to disrupt? We call it depth. But what we mean by that is, is there more than one company or more than one group exploring it, which will help add to that?

Robin Pomeroy: That must be a tricky job because you've got technologies which are still relatively small and you're having to think three or five years in the future to what it could become. Greta over to you.

Greta Keenan: Now that we know what the criteria are to get onto this top ten list, let's get into it.

Artificial intelligence is constantly making headlines these days and two of the technologies on the 2023 list are AI-related. Bernie, perhaps you can tell us about what those technologies are?

Bernie Meyerson: If you take a look at what's going on in the world, generative AI, you know Chat GPT, is just something that's hitting it out of the park in terms of people's interest in it.

It's truly extraordinary, the ability of it to basically help you go through and actually say this is an answer that as a human being might have taken me months to generate or I might not ever have the skills, Generative AI and Chat GPT basically can come back and give you this answer you might have sought for months or years or maybe never done.

That's the good news. But we tend to be very balanced here in terms of looking at what we call lights and shadows. Generative AI basically trains upon the existing data that's out there in the world and says, from that data, I can then parse together this answer that you're looking for, whether it's a description of a Shakespearean epic or it's something as straightforward as what would a painting look like of somebody who was basically ice skating in a competition? I mean, literally, it has that broad a set of capabilities.

The caveat and the quote 'shadow' that people worry about is how far can it emerge? In other words, how far can it go in terms of actually displacing human effort? And that has been a contentious area, which we're not ignoring. But there are such positive things it can do at this time that it deserves to be on the list.

If you think about it, for instance, in health care -- forgetting about just it can help a dentist, for instance, identify from an image there's an issue in your mouth, that's a fairly obvious one, but what about areas where you go deep into the hinterlands of a country, where you don't even have medical services and you actually have, however, a connection to an AI that has actually been trained in medicine? It has actually read literally the entire body of medical evidence. And if given some very basic information on symptoms, for instance, can actually come back and say, "We're probably dealing with this." That's access and basically levelling the playing field globally in terms of accessing health care and getting treatment, therefore, it's an extraordinary evolution and it's taking place today.

There are places in India that did not have medical services which are getting them through this means. It is clearly something that is going to be emergent over the next three to five years in terms of bringing up the standard of medical care globally, even where the educational system does not get caught up creating the ability for the local institutions to provide such care.

So, these are just two obviously disparate things. But when you look at them, and perhaps Mariette could comment, if you look at AI in Chat GPT, the challenge that you're faced with, of course, is also when misused it can generate answers to questions where you are attempting to train a student and the student just gets out on the web and basically pops up an answer that looks absolutely great and has nothing to do with their own ability. I mean, I know that's a setup Mariette, but it is a challenge you face in particular.

Mariette DiChristina: Thank goodness that as people develop new AI technologies, they usually develop ways to detect whether they've been used. And that's one of the things we do in education.

But one thing I would like to add to what Bernie's been saying that really struck me as I was looking over the list is that the power of data science, AI and computing is really throughout the entire list. We singled out a couple where the AI advances were particularly noteworthy, but that power of using computation and data to take a systems approach runs through the entire list.

Bernie Meyerson: And by the way, there is a very interesting emergent field, interestingly, in defeating AI. In fact, there is a very clever bit of work going on which might make the list next year, which is how do you detect AI's use in places it is not meant to be used.

As one example, human beings have many, many different ways of saying the same thing in words and they randomly select words that would give you the same outcome. Turns out you can actually set it up so that a system, such as AI, to identify itself always chooses among, let's say, ten different word options, it always chooses one. And, therefore, if you look at a response where it actually used all words that you know are actually AI affiliated, even though they could be randomly used, and you compare it to what should occur if all the words that I would basically know to identify yourself appear in an answer that you're given, you know, it was produced by AI, as opposed to a student handing it in.

it's inevitable that there is a plus and a minus. The good news is the world is beginning to deal with it proactively. So, poor Mariette will not need to sweat about whether somebody has used their brain or their fingers.

Robin Pomeroy: Generative AI. It might be people listening to this are saying, What do you mean emerging? It's emerged - because in the last six months, everyone, it seems, is using it. However, in those applications in health care or in medicine, we're still really in the foothills I think of what this technology can bring us. Do you want to go on to the next one Greta?

Greta Keenan: Absolutely. We just heard about generative AI and AI-facilitated health care, which made the top ten list. There are, of course, other technologies on the list that are related to health and health care. Perhaps, Mariette, you want to start us off with spatial omics. What is that and why is it on the list this year?

Mariette DiChristina: Thinking about health care and about what we can see and know, I think we've all seen a picture of a cell that looks pretty or maybe you've seen pictures of brains where there's an area of the brain that's active.

The problem with those previous pictures is that they're quite static and your brain, your body, all your metabolism is moving constantly with more than 37 trillion cells in our bodies right now, all actively working away.

To get a sense of what they're doing in between the one snapshot or in the case of brain imaging, things that are done at a fraction of a second, we are going to soon be able to turn to spatial omics. This uses advanced imaging technologies and specificity and resolution of DNA sequencing combined.

As I mentioned a minute ago, it has a lot of computing power behind it. And what can we see? We can see processes actually unfolding and the ability to see those processes unfolding, whether they're part of your metabolism or how your brain is working or how other areas are working gives you two things.

First, critically, a better understanding of actually what's going on. There are many things, like if I have a headache, that headache could be I just got a concussion. It could be I have a tumour, it could be something else. Sometimes things present in the same way and only if you can really see the mechanisms underneath them, can you understand what's going on.

And second, and critically, once you understand in detail, you can then develop therapeutics to approach challenges. Like if I had a cancer tumour, this technology could be used to exactly see what was unravelling in my cells and come up with approaches to tailor cancer treatments to that tumour or tumours in the body. It's a kind of an amazing technology.

In the next several years, financial folks forecast that it will more than triple its growth and impact. I think we've been watching this one in the committee for years and waiting to see when we can say, all right, within three to five years we're going to see a real impact here. And that is spatial omics.

Robin Pomeroy: Spatial omics. That's a term most of us won't have heard before. You've been discussing it for years in the committee. What is the word? What is an omic? Where does this word come from?

Mariette DiChristina: Omics often apply to different areas of study in biology, where we're understanding the processes of something, like metabolomics or genomics. Get used to the word omics, it's often applied to the study of different areas and spatial here, understanding the spaces and how they move along in the body.

Robin Pomeroy: Greta, on to the next.

Greta Keenan: So, we've heard about spatial omics. Another one of the medical-themed technologies on this year's list is designer phages. Again, probably something no one's ever heard of. Bernie, do you want to talk us through that one?

Bernie Meyerson: It's interesting. Basically, a phage is a delivery vehicle of information. In this particular case, you are taking a virus and one of the things viruses do to replicate themselves is they deliver information into your DNA that essentially causes it to replicate themselves when it's an infectious agent.

However, designer phages are a very different approach.

What you do then is, imagine you have a tub of a given bacteria and you want to manufacture something and, as an example, literally maybe manufacture insulin. Believe it or not, you can actually inject into the actual mix a phage which has been designed so that its DNA, or the fraction of its DNA that is injected into the bacteria it attacks, turns that bacteria into a factory. And that's a fascinating thing.

It's a very clean sort of thing where, if you design this phage correctly, it will actually insert DNA information into the bacteria that are present to turn them into factories for highly desirable things in vast quantities, whether it be a drug, whether it be insulin, whether it be any other chemical that you actually need in a biologic process. It's a remarkable thing.

In terms of manufacturing, if you think about it, it's very natural. It's not like you're going out and getting some magic ingredients and going and mixing them up and hoping for the best outcome. This literally is exactly tailored and it's kind of an offshoot of something we pointed out years ago, CRISPR/Cas9, where you actually do gene editing.

It's gene editing, but it's done with remarkable simplicity, literally using a designer phage, a virus that has been tailored to do the editing for you at scale.

So, it is something that is really revolutionary in terms of, oddly as it might sound, manufacturing. Literally, it's a biological manufacturing system that is making use of a virus to do your work for you. That's damn impressive and will become important because if you want to talk about low-cost drugs and the cost of drugs, one way of really driving that down is being able to do a very natural process that involves creating the drugs through some sort of fairly straightforward biologic process that then scales enormously.

Robin Pomeroy: I have heard the word phage before and, correct me if I'm wrong, isn't there some relation to the fight against anti-microbial resistance?

Bernie Meyerson: Yes. The antibiotics that are in use today, The challenge you have with them, is let's say they kill off 99.99% of the bacteria. You have to really worry about that .001% that survived and how they manage that. And it may be that they have some disruptive genetic information that prevents them from being attacked by this antibiotic that you've applied.

Unfortunately, over the course of many years, as you kill off bacteria that are susceptible to your drug, the only ones left are the ones that are resistant to it. So, you better find other ways of attacking them.

One of which is, again, using these viruses that are capable of genetic alteration such that they can actually make the necessary alterations in this biologic agent you are using for attacking these microbes that render them once again essentially susceptible to attack or attack them directly, literally by inserting some fatal genetic disorder into the microbe itself, using the phage as an actual weapon.

So, you've made a very good point that this addresses yet another one of the challenges we're facing due to, oddly enough, the success of modern medicine.

Greta Keenan: So phages selectively target specific bacterium and let's say you're taking antibiotics, they don't knock out everything the good and the bad bacteria, but selectively focus on one. So if you can use CRISPR-Cas9 or other genome editing technologies to modify that phage to target the exact bacterium that you want, then you're going to have a much more precise outcome. I think some of those examples within drug design and also in preventing antimicrobial resistance, are really showing the impact of this technology. Thanks, Bernie.

Robin Pomeroy: Next on my list, we're still on health, but on mental health. Mariette, how can the metaverse help our mental health?

Mariette DiChristina: So, although we've all heard of the metaverse and maybe we haven't really engaged with it so much yet, one thing that scaling and computing can do for us is help address a chronic but really underappreciated problem of all of humanity, and that is mental health.

I remember some years ago, while I was still a science journalist, I read about a decade-old mental health study that was done in the US every decade. The thing that really struck me was about half the people in the US at some point in their lives would suffer some form of mental illness and that 20% or more would suffer clinical depression at some point in their lives.

These are huge numbers, but everybody gets similarly a cold now and again, everybody gets other kinds of illnesses. So, why don't we fully appreciate and grapple with what the US surgeon general has called, "the health issue of our time."

The metaverse for health stands poised to help us with that, just as we are all right now sitting in different places around the world and engaging with each other, you can harness these technologies, but in a more kind of amped-up version to help with mental health diagnosis and treatment. That's what the metaverse stands to do.

We already have seen some signals of these coming technologies. For example, gaming companies have for some years now started to offer some mental health benefits. For instance, Deepwell Therapeutics has created video games that help address problems of depression and anxiety and there are others and there will be more coming along.

What's going to bring it all together is our new metaverse, our health and our computing platforms. Eventually, I hope, when it's hard to find a mental health professional to help support you in your care, we'll be able to find that more reliably and more readily than we would have in the past.

Robin Pomeroy: I was actually doing an interview this morning with someone senior in a metaverse company. It'll be on another episode of Radio Davos. And she uses the metaverse. She does mindfulness every morning and she does it sitting on top of an iceberg in the North Pole. So that's the environment she surrounds herself with. Presumably it's less cold than that, but there's this calming place for her to do her her mindfulness, I guess that's an everyday application that could be used to help mental health. Bernie, you were going to come in.

Bernie Meyerson: There is a fascinating example that I've seen in Australia where they have a company that does human-like avatars. What they found was in mental health treatments, people on the autism spectrum, across a broad range, actively worked well with this avatar, better with an avatar than with a real human. They felt they could tell that it was clearly an avatar and I think with intent the avatar was made to be clearly evident as non-human, but nonetheless very human-like.

They found that those who called in for help worked very well over Zoom and other technologies with the avatars, as opposed to humans. It is a very powerful example that we should not assume that, well, you need the human touch. There's a certain disconnect in some ends of the spectrum in autism where there's a level of comfort found using the metaverse as opposed to real people. So there's a really incredibly broad range of applications we are yet to explore.

Greta Keenan: Thanks for walking us through some of the medical-related technologies on this year's list. But obviously for any technology to have those real-world applications in medicine, they often require similar advances in engineering. Bernie, tell us a bit about flexible neural electronics. What are they and what advances have we seen on the engineering side of things to earn this technology a spot in this year's list?

Bernie Meyerson: It really is an enormous technical challenge and several of the things that we'll discuss in a moment, we're really talking about just enormous materials of challenges.

Think about this. You're actually putting something into the human body that has to maintain its location, that has to interact with neighbouring cells without damage, and has to extract some sort of data from it and transmit to the outside world without further damaging the location it's at. That is an enormous challenge in medicine, particularly with implants. And really the reason it earned this is we're finally seeing the ability to take electronics and first of all render them flexible. That's not trivial if you've ever taken a silicon wafer and tried flexing that, you must enjoy getting a little chip, some silicon. I don't mean good chips. I mean the broken pieces! There ain't a lot of flex there.

The remarkable thing now is there are new classes of electronics that are actually flexible, physically flexible. You can bend them without damage. More importantly, they now have biologic coatings over them that enable you to implant them in incredibly sensitive areas, like the brain, and extract direct signals.

In the Holy Grail of this work, you want to take somebody who, because of, for instance, a stroke or a spinal cord injury, you want to be able to take the necessary information from the brain and bypass the damage back to a region where you can then drive motor function, for instance, that was otherwise lost.

To do that, you have to have incredibly precise connectivity to where that information is coming from, which is the human brain. Previous sensors were either external to the brain, but, because of the locomotion literally of the brain, you constantly had to retune them; and/or inflexible electronic elements placed on the brain, which had a tendency to slip around and also were not remotely as friendly to the surrounding tissue because they're not flexible.

So, the fact that these implants are now being used and are proving to be both biologically compatible and good at extracting data is the very beginning of a revolution that hopefully will resolve many of the challenges we currently have with enabling those with an impediment due to some sort of either brain injury and/or physical injury and loss of connectivity through various parts of the body.

It would really help resolve those things in a remarkable way. It's really the Holy Grail.

It's not necessarily the scale that it actually is in some widespread use at the moment, but the impact would be almost unimaginable to people who are essentially quadriplegic in the longer term. It's something that has a really good future and it is now finally at least becoming a reality in terms of execution.

Robin Pomeroy: That's flexible neural electronics. So we've talked about implants or sensors in the brain, but another one on the list is sensors in a plant. I was watching a World Economic Forum social video on this just yesterday and I had no idea this is possible. Mariette, tell us about wearable plant sensors.

Mariette DiChristina: It is amazing technology.

One of the things that really strikes me when I was looking over the report, the UN has said that by 2050 we'll need to increase food production by 70% to feed the world's population. At the same time, we're struggling with various issues around climate and how different growing conditions are changing. We also, a little earl ier today, talked about systems in the body and how in the past we might have been able to take a picture or a series of pictures, but being able to look steadily provides us with revelations on what's actually going on at the ground.

In the same way, we can do this in a human body, we could do this for plants, with wearable plant sensors. The idea is you get a non-invasive sensor that can monitor resistance in the plant and by doing so and tell folks who are watching it, does it have adequate nutrition, how is its water supply and is it getting enough sun. And the difference here is going from being able to take very low-quality satellite pictures of a field down to exactly what's happening, one plant to the next. This will enable us to help feed the world.

Robin Pomeroy: When you say resistance, you mean electrical resistance. I had no idea that there was any kind of electrical signals passing through a plant. But, in fact, that is what this is measuring, right?

Mariette DiChristina: All of our cells, our cells, the plant cells, anything alive, uses teeny tiny electrical signals to communicate and transmit information. We are, I suppose, different in some ways, in many ways, from electronics, but we do contain our own little voltages going on in our bodies, and the plants do, too.

By measuring those electrical changes in the plant, the sensors will be able to give a lot of precision to what's going on in the field and enable us to get the maximum we can out of every acre in the world.

Greta Keenan: What do you think are going to be the biggest challenges in terms of reaching scale for this kind of technology, Mariette? I mean, it seems amazing that we could have that level of precision for each individual plant within a field or many, many fields in agriculture. But how are we going to get there?

Mariette DiChristina: Obviously, any new technology has a bit of work to really be emergent.

I mean, the sensors themselves have a cost to them. It's a cost that's now not yet applied to agriculture, and to bring that cost down scaling will need to happen and that's always a bit of a chicken and an egg event.

These sensors have to be applied in some way. We do have to put the seeds and plants in the ground and they will need to get some efficient way to connect them. Then, of course, we're going to need to manage all of the computing associated with hundreds of thousands of plants.

Bernie Meyerson: There is progress in an area, which has a bizarre name called smart dust. But remember that the miniaturisation now possible with chips literally means you could build a chip and you look at it and you would say, good lord, that's a speck of dust. That could be an entire sensor with communication capability where, in theory at least, you could literally fly a drone over a field and just sprinkle them, and enough of them would land on the targeted plants to get the data. So, technology is moving in a direction that ultimately will become pervasive. It's a question really of 'how fast'? It's not 'if'.

Greta Keenan: You've told us a bit about flexible neural electronics and wearable plant sensors. Naturally, for any technology or electronic to be flexible, it's going to have to be powered by something that is equally flexible, which brings us to one of the technologies on this year's list. Bernie, do you want to tell us a bit about flexible batteries?

Bernie Meyerson: If you think about what you want to do, for instance, where you want to make medical care, essentially a medical knowledge pervasive, one of the best things you could do as an example is take a shirt like mine and embed sensors in it, but you have to transmit the data from those sensors or store that data so it can be transmitted later and you have to take the data.

To do that you need power. People have really come a long way in basically making flexible lithium-ion or other types of batteries, which are remarkably resilient, like you throw your shirt in the washing machine and you don't kill the batteries. It's not a little problem. This world is coming to the level where we need material scientists to dig in hard because the capability of essentially extracting power in a way that you can make these kinds of sensing systems pervasive is crucial.

There are many, many companies out there now that are building sensors that are wearable, sensors that detect blood glucose levels for those who are diabetics. Imagine how unbelievable a breakthrough it would be if you had a Type One diabetic, a child who's born as a diabetic, where instead of having to either prick their fingers or inject them daily, you have the ability to literally just put a sensor on pants or on a shirt or on an undershirt and it literally could sense the glucose level in the blood and transmit that in real time to regulate the glucose level at a fairly constant point. Because the problem with a diabetic is when you have these big spikes in values of glucose levels, that's what does damage to the body.

To do that, you need power. So that is why we believe there's both a tremendous demand for these kinds of batteries and there's been enormous progress lately in achieving this. And it's not just a battery. You also have flexible electronics that literally use the motion of your body to generate power. I mean, there are a lot of approaches people are taking here, but because there's this tremendous need that actually is part of what drives emergence and demand for something and the demand is there.

So, the technology and the material scientists are working with pretty good success right now in generating stuff where you could build it into a piece of clothing and throw it in the laundry and it would still work. We're well along the way. And again, the three to five-year horizon seems reasonable.

Robin Pomeroy: One of the things I think about when I hear about new technological advances, is oh that's wonderful, but where's the downside?

And the last two on the list here we're going to talk about are also related to energy, but we're looking at sustainable aviation. Weren't aeroplanes a marvellous invention, aren't they still? But how do we make them sustainable?

And the other one is -- and it comes back to maybe the first -- thing we were talking about at the top of this list, sustainable computing. Generative AI we've learned uses a huge amount of power and a huge amount of compute. Mariette, maybe you can get green for us and tell us about sustainable aviation fuel and about sustainable computing?

Mariette DiChristina: You're right, some of these wonders of the modern age come with consequences that we need to grapple with at a systems level and again in a kind of combined way with the benefits of computing.

Talking about sustainable aviation fuel. Aviation is a significant piece, it's not the biggest piece of carbon use, but it's quite a significant one at about 2 or 3% of all carbon dioxide or more annually. But to get aviation carbon neutral, we're going to have to make tremendous changes in that fuel.

The problem is, where it's relatively straightforward to get cars to be electric because we can have charging stations and you don't need to have huge amounts onboard, although there are a lot of mechanical and electrical problems to be solved, aviation currently requires extremely energy-dense fuel because you can only carry so much on that aeroplane and get it off the ground. So the challenges there are a lot.

Sustainable aviation fuel reminds me of when you see those biologic fuels that you pump at the gas pumps, where a portion of the fuel, the normal aviation fuel, is replaced by fuel that is either produced from biologic ingredients or using non-biologic where you treat the fuels in a different way to make them less energy impactful on the environment.

Today, I think that about 1 or 2% of all aeroplanes use sustainable aviation fuel, but we're going to have to increase that by up to about 15% for airlines to become on the carbon neutral path by 2040ish on the way to 2050. The good news is that there's a lot of energy because, as Bernie just pointed out, there's a lot of demand and a lot of challenge here. So, there's a lot of energy in trying to produce additional types of sustainable aviation fuel used in different circumstances and to get those more available to the airlines that need them.

I wanted to touch on sustainable computing. Data centres, as you rightly point out, use a tremendous amount of energy, something like 1% of all the energy produced in the world today, 1% of anything is a lot. When you have to take a systems approach, you've got to look at all of the different pieces.

You can manage heat management issues to try to bring energy costs down with liquid cooling systems. AI again coming back to the beginning, is being used to analyse and optimise the power systems for sustainable net-zero energy centres for sustainable computing, and technology infrastructures supporting this are also being developed, being more modular and demand based so that is actually only responding as needed, rather than burning the same amount of energy at all times. So going back almost to the beginning, we do need to take care of reducing our energy use to make computing more sustainable as well.

Bernie Meyerson: That's actually a delightful side effect of things like AI and generative AI training, because these things do not require latency. It's one thing when you type a search into something like Google and it spits out an answer in milliseconds. That's the latency, the delay you're willing to tolerate. But when you've got something like training Chat GPT, which literally can run for ten, 20 hours, even days, then what you can do is you can basically transport that computation to a region where energy is essentially carbon neutral or better.

As an example, you know, you have people, for instance, in Iceland who are using geothermal power to do work in crypto currency, same sort of thing. Latency is that crucial. You have people who have now moved their data centres to where there's proximity to hydropower. Again, it may be far away from where they're physically located, but because latency is not a crucial thing, you are able to dramatically reduce the carbon footprint because the source of your energy is hydro.

People are finally beginning to parse the problem where latency is crucial. You don't have to burn down a forest to get the thing to work. That is a very positive trend. It's exactly what you've been talking about Mariette, which is, you know, these are just knobs you can turn to dial down the energy demand for this technology.

Robin Pomeroy: We've gone through the ten. Thank you very much.

What amazes me, is - this is the third time I've done this with you on Radio Davos - there are ten totally different technologies every year and you're seeing dozens and dozens more that don't end up in this report. So, it's a very impressive report. Congratulations to you. It's called the Top Ten Emerging Technologies of 2023. You can read the full report at weforum.org/reports .

Thanks to Greta for co-hosting and to Bernie and Mariette, who are co-chairs of the steering group, for walking us through the top ten of 2023. Thanks to both of you.

Bernie Meyerson: Thank you, guys.

Mariette DiChristina: Thanks for having us.

Robin Pomeroy: You can download the Top 10 emerging technologies of 2023 report on the World Economic Forum’s website. And please do check out our coverage of previous reports, still available on your Radio Davos feed or at wef.ch/podcasts.

Please subscribe to Radio Davos wherever you get your podcasts and please leave us a rating or review. And join the conversation on the World Economic Forum Podcast club -- look for that on Facebook.

This episode of Radio Davos was presented by me, Robin Pomeroy, with Greta Keenan. Editing was by Jere Johansson. Studio production was by Gareth Nolan.

We will be back next week, but for now thanks to you for listening and goodbye.