This post is a transcript from the Exponential Investor podcast, and has three main parts. The first is with Eoin Treacy, the investment director of Exponential Investor, then part two is an interview our own Farah Nayeri did with Kaspar Althoefer, head of the Centre for Robotics Research at King’s College London. Kaspar discusses a state-of-the-art flexible surgical robot that takes its inspiration from the octopus. Part three is a round up the week’s unmissable tech news with Nick O’Connor, publisher of Exponential Investor.
MF Mischa Frankl-Duval
ET Eoin Treacy
FN Farah Nayeri
KA Kaspar Althoefer
NO Nick O’Connor
MF Welcome to Exponential Investor, the MoneyWeek podcast dedicated to next-generation technology and the investment opportunities it provides.
Up first it’s Eoin talking about next-generation energy, moon shots, and why tech giants remind him of medieval city-states.
MF Eoin, tell me a bit about the technology sectors you’re looking to invest in.
ET Generally speaking you can look at three particular sectors that really have the potential to have this massive growth spurt. One is in the energy sector – the crash that we’ve had in oil prices is really only the beginning of that process – then in the healthcare sector because that is where even a small innovation can have really transformative effects on the way people live their lives, and then in the high-technology sector such as in artificial intelligence, and then of course there is what we’re going to consider moon shots; so, things that are, you know, way out there that really have high aspiration and we might not be able to say with a high degree of confidence that this is a slam dunk but that if it comes off it could be, you know, the next Apple.
MF What kind of sector is a moon shot company most likely to appear in, then?
ET Well, if you think about it, it can only be healthcare, energy, or high technology because when you think about, say, for example, a new consumer goods company like, you know, a McDonald’s, that is already being done. There is no capacity really for a new consumer goods company to really come out and set the world alight in the same way that you could get from energy, healthcare, or high technology. Personally I think that it’s going to be somewhere in the cross-section of technology and energy. It’s going to be the one that’s going to really come through first and then we’re going to have something big happen in the healthcare sector within the next decade, but what we’re now starting to see… in both the design of solar cells and in batteries you’re getting polymer batteries and you’re getting polymer solar cells.
Everyone knows what polymers are. They’re plastics and it was really only about 15 years ago that people figured out that you can get electricity to flow through a plastic and that really has set off just incredible innovation. So, what we’re now starting to see is that polymer batteries are incredibly more efficient. They are upwards of ten times more efficient than the types of batteries that you’re seeing in a Tesla or an electric car today and a lot of people are focusing on what that’s going to mean for the electric car industry because it is very visible, but that’s really only scratching the surface.
Something that I always try to keep in mind when you consider the price of a commodity is the Jevons paradox. Very simply, what it means is that when the price of something comes down, demand for it increases so that the total demand continues on a very long upward trajectory and it means that if we are going to have incredibly efficient solar cells, which is looking increasingly likely, and if we are going to have incredibly efficient batteries, then that opens up the potential that we’re going to have an awful lot of very cheap energy and that means that our demand for energy is going to rise and it’s going to rise considerably, but right now it’s hard to imagine that because we’ve gone through a 15-year period where oil prices have been really very high and what people have been really worried about is being efficient with energy. We’re in an efficiency cycle in terms of energy, but we could be moving into a period of time where energy prices come down and stay down and availability of that energy becomes increasingly more abundant.
I think where it’s going to be absolutely transformative is in robotics because right now you’ve got a lot of money moving into the robotics sector, but one of the things that they’re really inhibited by is battery life. It’s all well and good to have a robot that can walk around and pick things up and pick itself up, but it can’t do that for very long. If they’re going to walk out of the lab and into your house, then battery life is going to have a very meaningful impact on that. So, when it gets down to the robots that are really going to start to just do away with the need for household chores altogether, then that’s going to have a big impact.
MF Presumably, then, considering the potential of this new energy technology and also of robotics as a sector, do you expect to see some very big deals struck between companies doing those two things?
ET Yes. I mean, we’re already seeing that. One of the issues that we are seeing is that right now very large companies have a great deal of cash and part of the reason for that is because we’re in this enormously accommodative situation with central banks. There’s an awful lot of money sloshing around. So, smaller companies that are coming through are getting bought up and they’re… in many respects – particularly in robotics – Google is buying everything. So, that’s a different type of situation than we would have had, you know, at the beginning of the tech boom in the late eighties. There is so much additional money floating around right now, that smaller companies are staying private longer. I see that for example with a battery company I’ve been reading about where they’ve moved into production already for this incredibly innovative battery without ever having had to come to the market to seek equity capital.
So, yes, we’re definitely getting periods of time right now where companies are performing more like countries. They’re like semi-autonomous states, so I call them autonomies. You know, this is sort of like in the early modern period where you had a city-state like Florence or Venice where they performed just like a little independent country. Because these are types of companies that have benefited enormously from capitalism and capitalism trends towards consolidation, the bigger companies will tend to get bigger. The improvements that we’re getting in terms of technology and healthcare and energy are enormously beneficial for these large companies. There are two ways to invest in any kind of big boom. You can invest in the companies that are developing the new technology or you can invest in the companies that are benefiting from that technology and it is the very large companies who are going to be enormous beneficiaries of what’s coming forward and that’s where the partnerships will come through because they can see that it’s in their interests to enhance the technology that’s coming through because they’re going to be enormously benefiting from it.
MF Do you see these technology giants buying… for example, if we stay on the subject of batteries, do you see them keeping battery production in-house through acquisition or do you think that the battery companies and the energy providers will be capable of holding out and becoming giant companies in their own right?
ET That’s often really down to the management of those individual companies because it’s a decision whether you want to sell your company to a larger company, whether you want to build a… and Facebook is a very good example. Mark Zuckerberg could have sold his company but decided not to. He wanted to build it up into an entity in its own right. So, then it’s really only a question of the incentive that people starting out really have to… and the confidence that they have that they can actually get the job done on their own without having to sell out to a much larger company.
It’s all about scale and you have to also consider that, you know, larger companies are investing billions in development, but technology is an ethereal thing and it’s not necessary just by spending a huge amount of money that you’re actually going to get the incredible breakthrough that you’re looking for and that means they always have to be out there looking for ways to invest. We’re at a point where interest rates have been down for a very long time and central bank money has been around for quite a long time and these companies have a lot of cash, but that isn’t always going to be the case and as we progress through a monetary transformation – and the US is only at the very beginning of that right now – then the potential for more of these companies to come onto the stock market to lift – I think – is probably improved.
MF We’ll hear more from Eoin a bit later, but first I’ll hand over to Farah Nayeri, who has just interviewed a leading roboticist about the future of surgery.
FN So, Kaspar Althoefer, you’re a professor of robotics and the head of the Centre for Robotics Research at King’s College London and recently you were in the news because basically a team of roboticists, engineers, and surgeons from King’s have for the first time operated on a human body using a soft surgical robot as part of keyhole surgery. As my first question I thought I’d get you to speak a little bit about this extraordinary breakthrough.
KA So, these results came about through the work we have done over the last four years in the framework of a new project called STIFF-FLOP and that project – as the name implies – is really about creating new robots that can change their stiffness. So, they can be very stiff if required and then change to another state where they are more floppy. So, really the main idea is to create these new robotic devices which are very different from what we know from traditional robotic systems with very rigid and hard links, with the aim to create something that is – when required – very soft and can interact in a very natural way with the human body as it is required for minimally invasive surgery.
So, our robotic devices… they are made up of silicon. They have inside chambers that can be pneumatically actuated and through this pneumatic actuation we are able to move these robots into different directions. So, very much like the octopus is moving about its tentacles, we move about our robot. So, we take really inspiration from biology, from the octopus – that really was an important part of this project – and try to recreate that in an artificial system and with the long-term aim of course to create these tools that can be used in the surgical environment.
FN Yes. And can you talk a little bit about the experiment that you did the other day? If I understood correctly, you basically used the robot to operate on human cadavers and medical mannequins. Can you explain exactly what the procedure involved?
KA So, really bringing together this work that we carried out over the last four years, we then managed to create a new prototype – so, another of these soft robotic systems – and this particular prototype was small enough to be used on a human body. So, in Dundee at the institute there, which is called IMSaT, we were then able to carry out these experiments where we used our robotic devices, our soft robot arms – and inserted them or introduced them to human beings. These were human cadavers and we were supported there by clinicians who were part of the project – in particular Professor Alberto Arezzo from University of Turin – and we were able to conduct an operation like a standard operation using our robotic device as part of that.
FN Can you describe the operation that you did on this human cadaver?
KA So, here the focus was on colorectal surgery and the idea was to recreate or mimic a standard colorectal surgery as it is currently conducted on humans; especially humans who have cancer in their colon. It’s a very complicated surgical task often carried out actually as an open surgery – so, the patient is cut open – but of course the idea is to move forward towards minimally invasive surgery. So, some surgeons attempt there to use laparoscopic tools to carry it out and in this context we have developed our robot arm to support such minimally invasive surgical procedures and so we created this robotic device and this particular one was equipped with a camera, and we could show through our experiments in Dundee on the cadavers that our system… our flexible soft robot with its camera attached to it was superior to a standard straight-line laparoscopic camera. We could show that this flexible system could move deeper inside various cavities inside the human deli so to speak and the surgeons could get better images from the operating side. This was really the first time that a soft robot has been used inside of a human body in the context of minimally invasive surgery.
NF But of course this soft robot was aided by human surgeons. I mean, it was not operating on its own, was it?
KA That is correct. So, the system we have developed was remote-controlled by the surgeon and what we observed here was that it was very intuitive to, you know, move the robot around and the system could be used over many hours without any problem. Now, of course, if there is interest in moving forward from that and to automate that system. So, at the moment we are still at a point where human interaction is very important, but as part of this project STIFF-FLOP we have also looked into new ways to make these surgical robotic devices more autonomous and so we have developed a number of algorithms there so that these systems become gradually more independent and capable of carrying out certain tasks by themselves.
NF So, obviously you have not as yet used your instrument in a living human being.
KA That is correct. Yes. We have not used it in a living human being. That would require more research and also more development towards creating a real product at the end that can be used also in a human; in a living human.
NF I mean, is there anywhere in the world where surgical robotics are being used on real human beings?
KA Yes. That is very much the case. There is one very successful company called Intuitive Surgical and they have developed a number of robotic systems and they come under the name of da Vinci Surgical System and these are already used routinely in surgical procedures.
NF Right, but with your instrument you’re not really copying what they do? You’re trying to do something different?
KA That is definitely the case. So, what the main issue we can see with those existing robots is that they are very much based on the traditional type of robot – the kind of robots we can see in the manufacturing environment for example used in the car industry – and so very rigid links used there and potentially dangerous for the patient because if a wrong movement is done, these things are extremely robust and very strong as well, so tissue could be damaged. So, here we want to really draw a line and move radically away from that and that is really our vision to create these soft robots which are inherently safe when in use on a human being.
NF I see. You have said in various interviews and lectures that one day these procedures will become entirely automated. Now, isn’t that some… a kind of scary prospect? I mean, if someone tells me that I’m going into an operating room and that a robot will be performing the surgery, I would be quite petrified. What are your thoughts on the level of, you know, danger, risk, fear?
KA Yes. I mean, I can of course understand that and there is probably a certain reluctance there from the public to embrace these new technologies, but I would think the main reason is that this approach is something new, something that hasn’t been tried and tested yet. So, we are very familiar, very used to humans operating on humans. That’s what has been happening for centuries effectively. So, suddenly coming up with this new idea could be kind of a daunting thing. I suppose once these new systems are introduced, once this is happening, once people understand that it’s safe – possibly even safer than having a human in the loop – I think the public will embrace that as well.
NF How many more years until we get to that point, do you think?
KA Right. That’s always the difficult question. Let me put it like this: there will be more and more robotic systems used in the operating theatre, but they will remain tailor-operated for quite some time or at least with a human in the loop to oversee certain aspects of that operation. Now, moving to a complete autonomous system… I think we have to wait quite some time. The intelligence needed for these totally autonomous systems is not available yet.
NF I see. So, when you say quite some time you mean…
KA 20 years.
NF Yes. I mean, what you’ve explained in previous talks is that a lot of issues have to be resolved, that human intelligence is required to manage operations, and it’s just… not just about moving joysticks; that you have to make a decision where to cut, where to move the instrument to… is that right?
KA That is a very important point. So, what is usually done nowadays is to use camera images to guide the surgeon whether it’s a laparoscopic surgeon or a robot-based robot assistant surgeon; to guide the surgeon to the areas of interest and to inform the surgeon where to cut, etc., where to carry out the operation. So, just interpreting these images is extremely difficult. I am personally not a surgeon, so I look at these images… I don’t understand anything. However the surgeons through year sort of training can interpret those images and carry out the right tasks. So, to automate that process by itself is a challenge.
NF Well, you need an artificial intelligence at a very sophisticated level.
KA Exactly. Yes.
MF That was Farah Nayeri speaking to Kaspar Althoefer about the future of surgery. The second part of my interview with Eoin actually took quite a similar tack. We discussed virtual reality, one of the high-technology sectors we’ll be writing about. Eoin told me about how virtual reality can be used to treat chronic pain and how it could reduce the cost of medical treatment to zero. Remember: if you enjoy the show, please do share this podcast and follow us on Twitter.
Right. So, speaking again about big companies in a sense, I was reading something Nick forwarded me that you’d written about virtual reality and when we were talking about big companies and new technologies I was thinking about Facebook and Oculus and I was wondering whether you think Oculus Rift will be a breakthrough or – like every previous virtual reality attempt – whether it will fall flat?
ET Well, I think that it’s already a breakthrough because the technology that’s out there right now has in many respects solved the seasickness problem and that’s what has really held back virtual reality over the last 30 years because I remember 20 years ago there was movies and a great deal of excitement about virtual reality and then we never got to see any and right now T-Mobile in the US is doing a two-for-one deal on Samsung’s latest phone and because you take that phone you get Samsung’s headset with it. So, companies are doing an enormous push into virtual reality and it’s because the experience has been enhanced beyond recognition from what we saw just 20 years ago.
So, I think that there are so many different strands to where virtual reality is going. There are short-term, medium-term, and long-term considerations to consider with it because, you know, short-term it’s compelling and it’s enormous fun and it’s a breakthrough in the gaming experience. So, that’s going to drive growth in the short term, but medium-term what we’re already seeing evidence of is that it’s being used to treat chronic pain and that is something that could be incredibly transformative because if you go to a hospital and you are complaining of pain, the chances are that you’re going to be prescribed with an opioid. That has been our response to managing pain for nearly 5,000 years. It hasn’t developed at all in that period of time.
There really aren’t that many areas of our lives where we’re using some tool that the tool hasn’t improved in 5,000 years and yet that’s exactly what we have in pain management and what we’ve already got evidence of is that people that experience chronic pain – and there’s 100 million of them in the US alone – spend upwards of $600 billion a year on managing that pain. We have evidence and there’s a lot of research now getting poured into this that… people who experience chronic pain, people that have pain all the time are seeing relief when they get onto specific programmes with virtual reality that… it distracts the brain enough that they don’t feel the pain anymore and that’s incredible. That’s something totally new and it’s non-pharmacological.
So, virtual reality in the medium term has got incredible potential for hospitals and then long-term the potential that it has for learning, for leisure, for conferencing, for working… I live in Los Angeles, you live in London, and we could sit down by the fire and have this interview in a virtual setting and people could come along and visit us. We’re moving into this period of time when you can have social interaction without ever leaving your house and the potential that has for shopping and really enhancing the online shopping experience is very compelling because what people like about shopping is that they get to walk around. They get to go for coffee. They get to have the whole shopping experience. You get the prices and you get convenience by doing it online, but you don’t get the experience. Virtual reality could give you that experience without you ever leaving your house and that’s another one of the things that it could be very powerful for.
MF Many very interesting uses clearly… I was hoping to bring you back for a second to medical use. It’s something I find quite interesting that if you look at a company like GW Pharmaceuticals who obviously have had a very good couple of weeks, whose share price has rocketed on the news that their marijuana-based medicine has been proven to be very effective for certain types of pain relief and other medical treatment, the challenge they face primarily has been using the medicinal effects of marijuana and trying to rid patients of the psychoactive element, you know, to kind of get all the benefits of the healing power of marijuana without the negative kind of narcotic effects, but in a sense this VR treatment is almost the opposite, isn’t it? If you create a VR environment in which people are so distracted from their pain that they no longer feel it, it’s almost kind of the complete opposite end of that spectrum, isn’t it?
ET Yes. It is and there’s certainly an argument that marijuana is going to be legalised and that’s kind of a short-term trend. There are companies who are going to benefit enormously from that, but, you know, this is the disruptive power of technology; is that something that people have been fighting for for a long time, something that, well, an increasingly vocal minority have been fighting for for a very long time… they finally get it and then there’s the potential that it gets simply swept away by something new that comes along. There’s incredible interest right now in marijuana.
I was in the departures lounge in San Francisco just two weeks ago and the guy sitting next to me… we were sitting down waiting for our plane to leave and he works on one of these marijuana farms. He said that him and three other guys… they were able to make $4 million in ten weeks by growing marijuana and that, you know, was a record. So, the governor of California came out to interview them and there were stories written about them in the San Francisco Chronicle. So, certainly marijuana is in a boom market right now and that’s because there’s limited supply and abundant demand and this treatment and the medical uses for marijuana will certainly help it, but, you know, I think that, you know, marijuana right now is more like a craft brewer or, you know, moving into the recreation is where the big business is there whereas, you know, the medical uses of marijuana – I think – is a less important part of that trend.
MF That’s interesting. It’s particularly interesting; what you say that after so many years of campaigning, virtual reality or some kind of technological treatment could certainly make it much less appealing to medical professionals. Is there a possibility in the future that virtual reality environments because open source, their price plummets, they even become free and treatments for things like chronic pain become cheap enough that drug gougers and high prices no longer become a barrier to treatment?
ET I think that’s inevitable. You think about… what does ibuprofen cost today relative to what it cost when it first came out? Yes. I think that’s absolutely true and then there’ll be different business models for those virtual reality-type environments, so… yes. They’re already giving away the hardware. We already see… the way that companies see that they’re going to make money out of virtual reality is in service… software as a service or very much in the service-oriented way of doing it, so it’s going to be about, you know… you might even get the environment for free, but what you do inside that environment might cost you money and then the amount of money that it’s going to cost you will be progressively lower and that will encourage more people and, you know, that’s the same way we get with just about all technology; is that the initial price is high and then it trends lower until you get to a point where it gets to, just pile them high and sell them cheap.
MF Have you tried a virtual reality headset yet?
ET Yes. I have and unfortunately it was way too short because there was a very long queue and it was everything that I’d hoped for. It was absolutely awesome.
MF So, I’m here with Nick, the publisher of Exponential Investor, and we’re going to talk through a little bit of tech news, what’s really happening that’s important this week, the kind of things we write about, and the kind of things you should hear about. So, Nick, what’s your first story this week?
NO First story from me this week is about something that we’ve written about in the past maybe two or three weeks ago. It’s the field of optogenetics. Now, if you’ve never heard of optogenetics, essentially what that is… it’s a very complicated area of research in medicine, but to boil it down to its simplest terms, it’s editing the genetics of certain parts of your body so that they’re responsive to light. So, there’s lots and lots of implications for this in the field of neuroscience and stuff like that. Basically it means you can use light to trigger a response within your brain or… no; within a cell. It could be within your brain, could be on your skin, could be within your eye, which is what I wanted to talk about today, which is that the very first person to have optogenetic therapy has had that in Texas. They… doctors still don’t know whether or not it has worked.
So, literally it has just happened, but that’s why I think this is particularly interesting; because we’re right at the very start of something that could have extraordinary potential. So, I brought a few notes about it. If you’re interested… if you go on the Exponential Investor website, which is exponentialinvestor.com, just type optogenetics in the search field. You should be able to find it because I actually wrote about this study before it happened. What it is is… a woman who has a degenerative disease in her eye called retinis [sic] pigmentosa and basically what happens… in the disease, light-sensitive cells in the retina gradually die off. What optogenetics does is… it… to kind of simplify it slightly, it uses the DNA from algae that are responsive to light, puts them into your eye, and then uses the effect of that, which means that the cells in your eye are then more responsive to light; a kind of physiological trigger. Light becomes a physiological trigger. So, when light shines into your eye, it’s supposed to then trigger something within the nerves, which hopefully helps you see again.
MF I see. So, how exactly is that implanted? Is that a question of… is that something done through a technology like CRISPR or how does it work?
NO No. I don’t think it’s… I’ve… in all the stuff I’ve read about this, I’ve never read about CRISPR in relation, so I think genetic editing is probably a slightly different thing. It’s more… so, the way it’s described in the press is, the eye was injected with viruses carrying DNA from light-sensitive algae. So, it’s kind of trying… it’s kind of in the field of genetic editing, but it’s taking a kind of characteristic of something else and transplanting it into the human body so that that is… so that that part of your body is then responsive to light.
MF That’s really interesting and what kind of other modifications would be possible with Optogenetics?
NO So, the main one or the big field that they think would work or could work is neuroscience. So, the thinking there being… if you can make certain parts of the brain responsive and have a kind of physiological reaction based on light, that gives you a much more precise way of effecting changes within the brain. So, stage one would be making that part of the brain responsive to light and then rather than then having to… I think generally now you would have to use electricity or electromagnetics; something like that; something slightly less precise to then effect a change within the brain. What you would do here is… you know, you’d make that change first so that the part of the brain is responsive to light. Then you would use light in a targeted way to try and trigger a response within the brain. So, it basically… it turns light into a tool we can use.
MF I see.
NO Which is way more precise… I think they discovered it by accident years ago actually. It was connected… it’s kind of two things if you know what I mean. It’s taking characteristics from something else and transplanting them into your body and then using that characteristic to effect a change within the body. So, it’s very early-stage. I think perhaps in the next two or three weeks there may be results out from this trial in Texas; this first person to have the therapy. So, possibly we would know more then.
MF Yes, and I suppose we’ll write about it at the time if the results are interesting.
NO Yes, and just the final thing just from an investment point of view… the company behind it is called RetroSense Therapeutics, but it’s a private-listed company as with so many of these things; so, quite a hard thing to invest in at the moment, but worth keeping in mind because, you know, with the situation with CRISPR now… if you wind the clock back maybe two or three years or maybe four years, that’s probably where we are now with optogenetics; right at the start of research, trying to figure out whether or not it has potential.
MF And those companies… some of the companies involved have raised huge amounts of capital even though there are ongoing disputes about the kind of… the rights behind CRISPR.
NO Yes. So, no one disputes that that… that CRISPR works, right? So, they don’t quite know whether or not optogenetics works yet. They think it could. By they I mean experts; not me. They think it has enormous potential, but it’s too early to say definitively whether or not it works, but this trial that’s taking place at the moment… that’s the first step on the road to seeing if it actually works and it actually is viable. So, it’s worth keeping an eye on.
MF Interesting and definitely worth keeping an eye on… my first story is similarly about something quite exciting that is still in prototypical stages. It’s – as far as I can tell – the world’s first amphibious drone. So, we’re talking robotics now really. There are obviously unmanned aerial vehicles – UAVs – which are the drones most people would think of as drones and there are a few unmanned underwater vehicles or UUVs, which do exist. They don’t work particularly well and they mostly have military applications, but what makes this new one developed by the Applied Physics Lab at Johns Hopkins so interesting is that it can station itself underwater for a period of months, completely resist the effects of pressure and of seawater, and then suddenly launch itself from the water and assist people in the area.
NO Wow. So, it’s a way of rescuing people.
MF Well, hopefully. I mean, like a lot of this technology…
NO That’s the theory though, is it?
MF It is. I mean, that might not be the most profitable use, but it’s possible that this could be used… I mean, of course it could be used in military ways, you know, in nefarious ways. It could be a mobile landmine or a mobile kind of assassination bot, but hopefully it could also be used for good and hopefully it will be used for good as you say like stationing itself underwater on shipwreck routes and – when ordered to – popping up and dispensing supplies.
NO It’s like an autonomous lifeboat.
MF Well, that’s exactly it and it’s interesting because… I’ll put a link to this in… underneath the podcast because it really does astonish you when you watch it. It works fantastically well. It’s a bit like watching Jaws if Jaws could fly. I mean, it’s really that level of kind of… it’s completely undetectable and suddenly just pops up and is aerial.
NO And is that response… when you say pops up, is that intuitive? Is it controlled remotely from somewhere else or does it do… does it just do that in…
MF Well, it is controlled remotely for now. I can imagine… perhaps I’m being ambitious and perhaps I don’t understand, but I can imagine a time where – due to some kind of outside stimulus – it would wake so to speak.
NO Well, just because you think, you know… to connect it to a different story, which is obviously much more fully developed, but… driverless cars… so, if you have driverless cars plus drones plus underwater drones, you know, it’s not hard to foresee a situation where, you know, people might be much more coMFortable having an autonomous drone under the water than they would parked on the drive next to them. So, you could see that something like that could – in theory – be developed.
MF Definitely and what’s quite interesting about this from a more practical point of view is that it’s 3D-printed using commercially available materials. So, it’s not cheap to make yet, but soon it will be cheap and mass-producible and that will be really exciting when these things in the not-too-distant future could very feasibly exist in numbers in the seas.
NO Do you know who’s developing that?
MF It’s… so, it’s currently a university research project to my knowledge, but the best part about it is probably its name, which is CRACUNS, which stands for Corruption-Resistant Aerial Covert Unmanned Nautical System, which is a bit… it’s a bit mealy-mouthed.
NO They wanted it to be Kraken, didn’t they? I think that’s the wrong thing to call it. It’s like… drone is intrinsically sounds kind of dangerous even though they’re neutral. It depends how they’re used, right? Calling it a Kraken… you may as well call it the giant squid. Anyway, shall we move on?
NO So, I have another story which is about something entirely different again, which is obviously the point of this segment. So, this is looking at the next step on from Moore’s Law. So, everyone knows what Moore’s Law is, which is that computer processing power doubles every 18 months or so, and that’s based on the fact that we’re able to cram more circuit boards onto a computer chip, but they think there’s a theoretical limit to how many circuit boards you can get onto a chip and possibly we’re approaching that limit although it’s worth noting that people have predicted the end of this… of Moore’s Law several times in the past years and years ago, saying, well, we can’t go any further than that.
It has always been broken, but it’s reasonable to assume that at some point… I mean, it’s called Moore’s Law, but it’s not a… it’s not like gravity, you know? There’s a point at which potentially it wouldn’t work, but I read a story this week which is… it’s not about increasing the power of a computer chip, but it dramatically increases the energy efficiency. So, this is findings published on 11th of March in a peer-reviewed journal, Science Advances. So, essentially it reduces the amount of energy needed to… down to as little as one millionth of the amount of energy used per transistor in modern computers.
So, it takes the energy that… so, if you think of a huge – I don’t know – bank of servers somewhere or… presumably there are places on the, you know… on the planet where there are just sort of warehouses full of computer processors. You know, cloud computing and things like that… it takes up huge amounts of space and all that takes up huge amounts of energy. So, if you could reduce the amount of energy needed to power that same size of computer down to one millionth of what it is now, that would be remarkable breakthrough because it doesn’t increase the processing power, but it does mean it’s a lot cheaper to do it and a lot more efficient.
So, if you think about the amount of power it takes to keep your phone on for instance at the low level… I’m going to say my phone lasts 48 hours. Well, imagine you could reduce the power needed to power those same transistors by… to one millionth of what it was. How long could your phone last then? You know, I’m not going to do that calculation live on air, but it’s a long time. That’s on the micro level, but if you look up at the top end of the scale where you have massive computers that take up huge amounts of power, then the effects are even more radical in lots of ways because that power can either not be used by them or it can be used much more efficiently.
So, that’s quite a significant breakthrough. Again, it’s quite theoretical in the same way as the optogenetics thing. It’s not the sort of thing that you’re going to see rolled out in the next year or so, but once something is theoretically possible I always think, well… or if it has been proved on a small scale, how long can it be before that becomes sort of widespread?
MF Well, it is a very interesting breakthrough insofar as computers at any level of sophistication often work well below their kind of stated spec. They’re a bit like cars in that sense that, you know, a car might say it has 600 horsepower, but if you test it, it might be closer to 400 because it’s not on a perfect road the same way a computer might have a theoretical processor speed and be much slower because it’s not under perfect conditions and it’s not perfectly efficient. I can think of one use for sure and it’s a topic we cover that’s not particularly glitzy, which is Bitcoin and… a lot of Bitcoin mining is not profitable because the energy required to process the calculations necessary to mine Bitcoin is so energy-intensive. So, I know that there’s a start-up – I spoke to somebody the other day about it – that houses huge warehouses as you were saying in the Arctic Circle; in Finland, I think.
NO Is that because it’s… they give off so much heat that it’s just cooler to have them out in the Arctic Circle?
MF Well, that’s exactly it. It’s cheaper to cool them; on top of which the cost of energy there… the cost of the grid is so much cheaper that at that level Bitcoin becomes incredibly efficient and… as you say, this kind of… were this prototype to go into use, it would make that even more profitable still and – as you can imagine and as you said – there are all kinds of uses.
NO Actually, before we move on, here’s an interesting one for you just to connect that to something that’s happening slightly more in the real world. So, we saw the machine beat the human at the Chinese board game Go. I wonder if someone – and by that I mean you – can find out how much energy that machine uses up to play a game of Go or per 24 hours or something because it would be interesting to see how much energy it takes to do that level of calculation because obviously a huge amount of processing has to go on, which would lead you to believe that there’s a large amount of energy needed, which basically makes it unfeasible to have that level of processing power available to everyone. It’s just too expensive to run, but if you could reduce that… well, that then makes… it drops the cost of computing, doesn’t it? So, that’s your challenge for next week or the week after. I’ll give you the Easter break. See if you find out how much energy that uses. There has got to be… someone has got to know. Someone’s paying the bill for that thing.
MF Yes. I mean, I’d ask Demis Hassabis if he ever said yes to an interview, but I think it’s off the cards until he has solved intelligence.
NO Yes. We’ll just let him do that first and then we’ll try and get him on.