Magic Markets Ep190: Is Quantum Computing the Next AI?

Listen to the show here:

Description:

This podcast is for informational purposes only and is not financial or investment advice. Please speak to your personal financial advisor.

Full transcript:

The Finance Ghost: Welcome to episode 190 of Magic Markets, and it’s going to be a goodie. We’re going to expose you to some stuff that you might not have heard too much about before, being quantum computing. And we have a great guest with us today to do that. Over to you, Moe. Lead us into quantum computing.

Mohammed Nalla: Indeed, Ghost. Very excited to be doing this episode this week. It’s episode 190 and we’ve got a guest that you might have actually caught a few weeks ago. We’re going to have Dimitri Zabelin on the show, let’s call it every six weeks or so. Dimitri is a geopolitical strategist. He’s looking at a lot of the bigger picture stuff that we just don’t get into in a lot of detail. This week we’re looking at quantum computing, and quantum computing is really exciting because it’s one of those things that really bend my mind. Dimitri is going to help us unpack not just what quantum computing is – we’re going to touch on that – but specifically, what are some of the applications? Why are we talking about quantum computing when everyone’s talking about AI? Where does quantum computing actually fit into the investment structure as a longer term megatrend? We’re going to unpack those applications, some of the risks, and also some of the investment vehicles in terms of how you could look at getting exposure to this trend, if it is something that you like.

With that, let me welcome Dimitri Zabelin back onto Magic Markets. Dimitri, welcome. We’re keen to hear your views.

Dimitri Zabelin: Hey, Moe great to be here. Thanks for having me.

Mohammed Nalla: I think kicking off, Dimitri, as we’ve indicated, quantum computing is one of those amorphous new topics. Every time I read something on quantum computing, it literally bends my mind. It’s really out there on the fringes of what’s quite understandable. And so it’s quite important, I think, to maybe just start off with what is quantum computing? Why are we talking about this when everyone’s just talking about AI? That’s the big rage. That’s what’s getting all the investment attention. Let’s maybe start off there. And if you could give us a very quick primer as well in terms of just some of the key concepts that make quantum computing important and relevant to our discussion today.

Dimitri Zabelin: Yeah, definitely. With quantum computing, in short, right now, the reason why I view it as an interesting thematic investment vehicle is because relative to AI, it’s still developing. And in my view, a lot of the assets that have exposure to quantum computing, specifically, are still pretty underpriced relative to its potential. And we can get into some of the more specific investment vehicles later, but really, the difference between quantum computing and classical computing, one of the biggest ones is the ability to vastly increase your computational power.

And to give you an idea of the scale of that, Google’s Sycamore computer, I think it was just a couple of years ago, announced what it called quantum supremacy, which is just another way of saying a quantum computer completed a task faster than a regular computer, a classical one. A classical one, it would have taken this task 10,000 years to finish. This quantum computer did it in 200 seconds. That’s to give you an idea of the scale of computational power that quantum computing has over classical computing.

And there are some parallels, right? A classical computer uses zeros and ones. It uses a binary system, and we have these transistors that switch on and off that correspond to a zero or one, off being zero, on being one. Fairly simple. With quantum computing, we don’t have bits of information, but what are called qubits or quantum bits, and this is where things get really, really weird. I’ll try to keep this as simple as I can, and I want to disclose I’m no quantum physicist expert, so these are just the basic principles. But what makes a quantum computer so powerful is the ability to leverage what’s called superposition, which is this quantum phenomenon where something can be in multiple states simultaneously. And instead of being just a zero or a one, it’s in many combinations of a zero or a one. If you look at an electron or an atom, it has its own magnetic field. And this corresponds to an orientation that’s either spin up or spin down. You can see the binary there, too.

In the same way you have transistor on and off, you have a magnetic field that also has a spin up or spin down. You can correspond it to a zero or a one. The problem, and this is where things get really weird, is that – or not a problem, it’s a phenomenon, and a great one – is that superposition allows these atoms to exist in these states in various probabilities of zero or one. This is called superposition. That’s just the basic concept of it. And coherence is the stability of this quantum state, because that’s where you have all of these calculations. And there’s another phenomenon called parallelism, which is because you are in a quantum state, that allows you to calculate in various probabilities between zero or one. It allows you then to vastly increase your computational power. And to give you an example of how this works in a more practical way, imagine you have a maze. A classical computer would go through every single possibility once. It would go through, okay, is this route working? No. Is this working? No. A quantum computer could calculate all of the possibilities simultaneously.

And this obviously has huge implications and applications in various economic, financial, and geopolitical domains. But that’s the basic concept. In summary, superposition, something can exist between various states or zero and one; coherence, which is the stability of the state, where you can have these calculations; and then something called entanglement, which is when these qubits, these quantum bits of information, are related to each other. A change in one corresponds to a change in the other. And that can also vastly increase your computational power. Those are the basic ideas. And as we’re seeing with classical computers that rely on digital transistors, what we’re seeing is that at a certain point, if you compact these transistors too closely, without getting too much into the weeds of it, what may happen is you encounter this phenomenon called tunneling, which is when electrons pass through barriers that they normally shouldn’t. This causes the circuits to potentially be shorted. Not melt; I think that’s a bit of an exaggerated term, or heat up and malfunction. This means silicon appears to have an expiration date to a certain degree. Quantum computers, on the other hand, don’t seem to have this problem. Those are essentially the basic principles of it.

The Finance Ghost: I think that was a pretty good explanation, because I know dangerously little about computers, and I managed to follow that. Thank you, Dimitri, for taking something that is so confusing and making it something that is understandable. It’s some serious Star Trek stuff here. This really does sound very, very cool, and I can understand why it’s so much faster than if you have to test every single thing once, if you can do it all at the same time, rather than waiting for a result from the first one. It does make sense why that’s so fast. I’m glad I’m not playing a game of chess against a quantum computer, because I think we know the outcome. But other than absolutely spanking people at chess and sending spaceships one day and inspiring movies, what are some of the applications for quantum computing now and in the very near future that is driving this investment thesis, potentially?

Dimitri Zabelin: First off, there are many different kinds of quantum computers. You’re seeing, essentially, this testing and development occur. You have, and I’m not going to get into the details, but you have D wave, topological, photonic, ion trap chamber, all these really fancy-sounding words that not many people understand, which is actually part of the risks that I’ll get into later. But some of the economic applications are really quite exciting. One  – and a big one, this is a thematic investment vehicle I’ve been looking at – is precision medicine. We are approaching a point where we will be able to customize medication, not just to, let’s say, your genomic group, but to your DNA specifically. The problem with that is that classical computers can’t. They don’t have the processing power, and we don’t have the time to input all of these factors to see how it could affect the drug’s, let’s say, efficacy vis-a-vis the patient.

To give you an idea, a classical computer still can’t compute the different kinds of permutations and outcomes a protein has and how it folds without a vast amount of time and computing power. We’re talking about tens of thousands of years here. A quantum computer would be able to simulate the simultaneous calculations of how this drug and the protein and how it folds may affect the patient. Again, that goes back to these principles we discussed earlier of superposition and parallelism. It leverages these quantum phenomenon to increase its computational power. And as we know, when it comes to pharmaceuticals, it costs upwards of a billion dollars to create one super drug or a miracle drug. With quantum computers, they would be able to accelerate that process by simulating all of this in a digital context, rather than doing it physically, like in a petri dish and other sort of physical laboratory work that is very capital intensive. That’s one of the applications of quantum computing.

Another one is agriculture. We spend – I think the figure is 2% of global GDP, I could be wrong – but it’s a ridiculously high number on creating ammonia and fertilizer. The problem is that for creating ammonia, it is an incredibly capital intensive and very expensive process. But what’s interesting is that in nature, this is found to occur, well, quite naturally. In other words, in the roots of legumes and beans, there is an enzyme called nitrogenase. And what it does is that, in short, it separates the nitrogen atoms from the air and then creates ammonia by binding these hydrogen atoms to it. What’s amazing is that nature does this with very unstable conditions, right. It’s not done in a laboratory with very precise medicine. Done in nature, there’s a lot of different kinds of interference. Yet it takes us quite amount of time and effort and money to be able to create this, specifically ammonia. We still haven’t mapped out the nitrogenous enzyme. It’s too complex, and we don’t even fully understand it yet. The quantum computer would enable us to understand it and vastly increase, let’s say, agriculture output. That’s a very long winded example of how you could apply it to agriculture.

Another one is logistics. How do you optimize supply chains when so many factors are taken into consideration? What if there is a trade disruption through a war or a conflict? What if there are sanctions? What if there’s a bad weather event? What if it’s just a malfunction? What if there’s a spike in demand? As you add in more variables, let’s say not just one ship transporting goods to one place, but multiple ships transporting goods to multiple places under various conditions, that is a lot to take in, and that overwhelms a classical computer. A quantum computer, again, leveraging the properties of parallelism and superposition, meaning this ability to vastly increase your computing power by calculating multiple possibilities at once, would help essentially lay out the most optimized route. That’s an application with agriculture and logistics. Now, the weird one, and this is where we really, like you said earlier, get into some Star Trek territory, is the ability to create new materials. Not new elements, but new materials, meaning blending various alloys and metals and minerals to see how you could create, let’s say, a superconducting material, a material that’s, let’s say, very light, but very durable.

Those are some of the applications that we can see. And from the geopolitical side, you can, for instance, again, optimize, let’s say, supply chains for getting your military hardware or your troops into a certain location, taking in all sorts of factors, weather conditions, risk probabilities, etcetera, all of these different elements that would, again, overwhelm a classical computer, but is well within the wheelhouse of a quantum computer. Another one is cryptography. The processing power of quantum computers leaves a lot of modern encryption guardrails, so to speak, in jeopardy of being undermined and being broken. And that obviously has huge implications economically and politically. There’s an emerging race, and we can get into that a little later, towards developing quantum computing not only as a function of economic security, but as a function of having a political security advantage as well.

Mohammed Nalla: Yeah, Dimitri, I think lots of interesting applications to unpack there. The obvious ones with benefits to mankind as a whole, agriculture, medicinal uses, and so forth. The cryptography one is an interesting one to segue from this discussion on practical application into a discussion on not just risks and obstacles, but also some of the geopolitical tailwinds. At this point in time, my read, my understanding, is that quantum computing exists alongside AI as a trend, in that it’s the raw computing power that might actually allow AI to just be utilized a lot more effectively in time. But at the same time, the cryptography raises lots of cybersecurity questions.

We’ve recently seen some major cybersecurity outages, and this becomes quite complex, because my understanding, again, is that quantum computers are probably where conventional computers were back in, let’s call it the 60s or so. These are massive machines that require billions and billions of dollars worth of investment to actually set up. And what I’ve read is that China is actually leaps and bounds ahead of the United States and other global players in terms of building out some of these quantum computers. That’s going to have some serious ramifications just in terms of this global balance of power. Is it the new space race? Is that what we’re seeing right now? Can you unpack what some of those risks and obstacles are, and kind of bucket that alongside a geopolitical discussion before we get into a discussion in terms of the investment case and the investment vehicles.

Dimitri Zabelin: Definitely, and you’re right, there are – that seems to be the common denominator in the more secular geopolitical trends we’re seeing, is targeted industrial policies for frontier technologies. China has been doing this really since 2016, and it’s what’s called more vertical versus horizontal. But that’s a discussion for another time. In short, some of the tailwinds are, like you said, because it has so many applications and implications, economically and politically, you are seeing a lot of countries and regional sovereign units like the European Union develop their own quantum computing initiatives and putting in hundreds of millions, if not billions of dollars into developing it. As far as I’m aware, though, you are right, China is developing its quantum computing capabilities quite rapidly, and they’re building it out. But something we’ve learned, particularly vis-a-vis China and the US, is quantity does not equal quality. China may have more, but that doesn’t mean necessarily that it’s better. I think the most advanced quantum computers that memory serves are Google’s Sycamore computer and IBM. They have one called Condor and then another one, I think it’s called Eagle. Its name is escaping me. And their respective computing powers, I think are much higher than what China has.

But we’re seeing how countries, like I said, the EU, US, China, I think Australia, the UK, all of them are developing their own quantum computing initiatives. And what’s interesting is that you mentioned AI. Nvidia is working closely with an Australian laboratory to develop chips that are specifically optimized for quantum computing, which touches on the point you said earlier about this merger between quantum computing and AI, sort of between hardware and software, between hard power and soft power, if we want to use the geopolitical terminology here, because quantum computers will essentially provide the processing power and AI will have the analytical power to analyze all of this data and all of its outcomes. They’re complementary, they’re not adversarial. And I would say those are some of the tailwinds – the risks and the opportunities that every country has in getting a first mover advantage.

And like you said earlier, I think that’s a great parallel you drew about computers, quantum computers now and computers in the 1960s, how big and bulky they were. To give you an idea in terms of some of the risks and obstacles about why they’re so big and bulky, one of the risks is cost. Let’s say, Google’s Sycamore computer, which is a superconducting quantum computer. It requires for its qubits to be calculated in something called near absolute zero. Absolute zero is a theoretical temperature that, when atoms physically stop moving, it’s the coldest theoretical temperature known. I don’t think it’s ever been actually reached in a laboratory. To give you an idea, it’s colder than outer space. That takes a tremendous amount of power, this cryogenic freezing, and that amounts to between $100,000 to $500,000 a year. That’s just for the refrigeration aspect of it. That has nothing to do yet with all the other hardware and software that it takes to build it, all of the talent shortages of which there is an immense amount, because, as you can see, this is a very specialized field that very few people, relatively speaking, are in.

Those are some of the obstacles, is the cost. Another one is deployment. If you’re in an environment with high interest rates, obviously your appetite for frontier technologies with a long time horizon – we’re talking a decade or two – that’s not really compelling if you have shareholders to be accountable to or just investors that want a return much sooner. That time horizon is also contributing to, let’s say, some of the risks and obstacles in accelerating its development. But we have seen a recent uptick in interest in investment for quantum computing because they’re beginning to see how it could be applied to AI and how it can merge. At this current moment in time though, I would say it’s wildly underpriced relative to AI.

The Finance Ghost: The AI valuation uplift has been insane, right? Everyone’s basically just jumping onto that thing like crazy. Personally, in my portfolio, I’ve always taken the route of saying I definitely am not clever enough to understand all these underlying IT trends and get them right. But I do hold a lot of the big tech names because I feel like if anyone’s going to get it right, it’s going to be those sort of companies. Is that still the case in something like quantum computing? Is my kind of catch-all of Microsoft likely at some point to catch this thing and some of the other big names? Or is the technology at a stage now where if you believe in the story, you do actually need to go and buy something that you’ve possibly never heard of, or is at least very much off the beaten track in terms of tech exposure?

Dimitri Zabelin: That’s a really great question. Certainly some of the leaders like – okay, let me rephrase that – you can get indirect exposure to let’s say, quantum computing by, let’s say, if you purchase shares of Google, they have their own quantum computer that I mentioned. Microsoft, IBM. I think Honeywell is working on some of these things, so you can get indirect exposure that way, because these companies are well-established, they have a big market cap, their R&D budget is considerably higher. They can certainly magnetize both capital and talent, which makes them, I guess you would consider, compelling candidates. If you want more direct exposure, one ETF that I’ve looked at, and full disclosure, I’ve invested in, I don’t want to lead anybody astray and be completely transparent, is $QTUM. It’s a quantum computing ETF issued by Defiance. The reason why I like this one is that unlike most other investment vehicles that have exposure to other assets, and quantum computing is kind of diluted, for me, this seems like the one that is the most focused on quantum computing without being one company, but a whole conglomerate of them.

Furthermore, if you look at the weightings of each of all of these holdings, I don’t think any single weighting ever exceeds 2% or 3% of the entire portfolio. It’s pretty diverse in that way. It’s not like it’s 50% Nvidia and then just 50% all these other ones where 50% of your portfolio is determined by oscillations in that one stock alone, which we’ve seen over the past few months has shown incredible volatility in both directions. It’s very diversified in that way, and it’s very targeted, which is what I like. In my view, the reason why I think it has bigger implications for the market is we saw how Nvidia pushed the overall stock market much higher because of all the euphoria that was accompanied with Nvidia’s earnings and the potential for AI and the releases of ChatGPT and all its iterations. I think quantum computing would have a similar effect in that way. You can get, again, direct exposure through, let’s say, this &QTUM ETF, or indirect, through buying, let’s say, tech-heavy indices or companies that specialize, not specialize, but are putting in R&D into developing their quantum computing initiatives.

Mohammed Nalla: Dimitri, that’s fantastic. There’s so much to unpack there, and as usual, with topics as deep as this, we’re always just kind of scratching the surface. I’ve read some of your work. You’ve written up on quantum computing before. For listeners that are interested in exploring this further, where can they find some of the stuff that you’ve written, maybe some of the interesting books that you maybe have read. Where can they go and learn more about quantum computing? Because it’s certainly a fascinating theme that I’d certainly like to explore in a lot more detail.

Dimitri Zabelin: Definitely. Every week I publish through my Substack a weekly insight, and you can find it at @pantheoninsights. And there I’ve published research not just about quantum computing, but quantum computing and AI, the energy transition, how critical minerals necessary for creating the hardware and software for all of this come into play, and how geopolitical forces are turbocharging it, hampering it, or neutralizing it. If you’re interested in that kind of macroeconomic, macro financial research and the analysis of those secular trends, you can find that at Pantheon Insights. And I release a weekly insight every week completely free of charge.

The Finance Ghost: Brilliant. Dimitri, it’s always so great to have you on the show. Thank you for coming back. We certainly hope to get you here again because I think you do bring a deep voice and deep insights – that’s basically your strategy, and you do it very well! But on a serious note, you bring us a lot of stuff that I’m not sure that we are necessarily looking at otherwise. Thank you, and it’s great to have you. We look forward to it again. To our listeners, go and follow Dimitri. Dimitri, where is the best place? Is it a LinkedIn vibe? Is it more of an X situation? Where are you more active?

Dimitri Zabelin: It’s really a combination. LinkedIn, Twitter and Substack. I post frequent updates on there. Give those a follow and you’ll see some of my thoughts I get throughout the week.

The Finance Ghost: Yeah, go and check out Pantheon Insights. We’ll include a link in the show notes. And Dimitri, thanks. We look forward to having you back.

Dimitri Zabelin: Thanks for having me on.

Mohammed Nalla: Thanks. Cheers.

This podcast is for informational purposes only and is not financial or investment advice. Please speak to your personal financial advisor.