Q&A

Question 1: Learning Quantum

Where can I find the platforms that teach me how to develop applications for Quantum Computing?

You can find these at almost every major vendor in the IP business because just as they are busy offering tools for artificial intelligence, they are starting to get busy offering tools for Quantum Computing as well. One of the busiest probably is Amazon which is creating an API, an Application Programmer Interface for almost anything. As a consequence, you can also find an Amazon Quantum Center And Solution Lab and so on and so forth, to start playing with these software developer kits. In particular, this is based in Python, so you can start experimenting and having an AWS account is very easy.

There are free tiers for a lot of things that you can start using, and there are some very generous promotions as well, so that you can really build and learn without spending money. Another one is Microsoft. They’re calling it the QDK, the Quantum Development Kit, called Azure like everything in the cloud for Microsoft. They have this very, I would say convenient, graphic or chart of what is the relationship of the various layers. The hardware layer is both in the quantum implementation and the classical implementation, then software, tools and services on top, and at the highest level there are the applications themselves: simulation, machine learning, optimization. It is interesting that Microsoft is explicit about the fact that they will be building Quantum Computers, and their particular research is based on the so-called Topological Quantum Computers which don’t exist yet, so their answer is “they’ll be available in the future”. They also have others that are available and can be accessed.

Another thing that I would like to highlight is that part of the software layer is 1Qubit, which is a Canadian company founded by former students of mine at Singularity University. Amazon and Microsoft are fierce competitors. Obviously, so are the others. But sometimes there will have to be some collaboration, I believe, especially because developers will want some degree of interoperability, certainly at the level of what you learn, but code portability will have to be possible as well as in some way.

IBM is another important player, of course, and you can sign up for an IBM ID account, and then start what they call the “IBM Quantum Experience” which really enables you to look at the Python code integrated in their Quantum Development Kit, and then start building Quantum Circuits, all of this for free. One that is surprisingly inaccessible, at least today, is Google that proposed a Quantum Software library for Quantum Circuits, called CIRQ, but this was launched two years ago, and the GitHub repository is not very lively. We’ll see how this will develop. And if it doesn’t, it is not representative of what will happen in the future, but it looks like that for AI developers are favoring PyTorch, which has been proposed and supported by Facebook, as opposed to TensorFlow, which has been proposed and supported by Google.

So what is going to happen in Quantum? Who is going to win the hearts of developers? Is it going to be TikTok? Just to give a crazy name, why not. We don’t know. In the meantime, all of this is available from IBM, Microsoft, Amazon, and there are some other players as well.

The most not famous but the first original Quantum player is D-Wave. It is important that their computer is not a universal Quantum computer. It is built for very specific classes of problems, but it is the only commercially available Quantum computer. If you have $30-$40 million you can buy one. They also have programming resources so that you can access the D-Wave Quantum computer remotely.

 

Question 2: Quantum in the near future

Which sector or industry do you think Quantum Computing will impact the most in the near future?

The tricky part there is “the near future”. It depends how we define the near future. In the course of the next 2-3 years most of the industries are not going to be disrupted through Quantum computing except, you know, academic research, PhD thesis publishing reviewing, editing, Quantum computing research and academic papers, organizing conferences. This is already happening of course, and this has been happening, and there is an entire supply chain that is being formed. For Quantum computers to become even more viable and exciting then before. These ecosystems are absolutely necessary.

This is an image from a wonderful presentation that the chief architect of Intel made at a very specialized conference on the future of electronics manufacturing, just a couple of days ago. And to prove you that Intel understands our paradigm of Jolting technologies, rather than just improving 8 times in 5 years, they are already planning to improve their overall capacity of producing electronics a thousandfold, and they are not doing it alone. They are doing it together with all these partners: foundries, design services, universities, materials, packaging and so on. The same thing is happening in Quantum computing as well, and the industries that participate in this effort are going to be impacted and potentially disrupted most rapidly.

Another way of looking at the question is: “What industries can use Quantum computing the soonest? Are the companies within that industry going to receive a substantial competitive advantage by using Quantum computers?”. I believe that materials, science, simulations, modeling, molecular biology, chemistry are going to be the first to be impacted, and then immediately after the ability to design Quantum computers with Quantum computers, and then immediately after designing quantum AI systems. What will it mean to have machine learning done on top of a quantum computer, rather than being done on a traditional computer? And that will potentially make this number even larger and Intel showed the chart that I disagree with. This chart shows two lines, rather than the curve of increasing acceleration. When Quantum computers are going to design AI systems, there will be an additional phase, and, at the time, things are really going to accelerate.

 

Question 3: Quantum availability

When do you expect the huge price drop from the commercially available quantum computers?

Not soon because that basically would mean that rather than improving features, we can start optimizing industrialization. We are still in the process of improving features which is actually related to this question as well. Today’s quantum computers are very exotic as you have seen in the course. Temperatures that are lower than in any part of the known universe, crazy characteristics. Also, our ability to interconnect the Qubits is still relatively low, we want a dense computer. Our ability to correct for errors is low, we want to improve that as well. I think that the prices of buying a Quantum computer will stay very high for maybe the next 10 years. That doesn’t mean that the ability to use quantum computers is going to be expensive too, but it will be like the mainframe era when nobody would have thought to buy a mainframe to the point that even IBM underestimated the need for mainframes, they thought a half a dozen would be enough for the entire planet. As of right now very few individual companies will buy a Quantum computer, but a very rapidly increasing number will use a Quantum computer.

 

Question 4: Hot Qubit

What about this recent announcement of a high fidelity “hot” Qubit that Intel announced together with QuTech?

The announcement that the question refers to is just from a few months ago. And “hot” is in quotes but rather than single quotes they should have used 3-4 quotes because hot actually means -272 degrees, rather than minus 273. A relative term, -273 means 0 Kelvin, so rather than very very close to 0 Kelvin these hot Qubits function at 1 Kelvin, and everyone is: “Wow, amazing, incredible!” Still crazy cold and not hot in normal terms. But of course, it is an important step because similarly to how we are looking for room temperature, superconductors are also looking for the ability to use these room temperature superconducting circuits, potentially in Quantum computers. Maybe we will get there; designing these kinds of materials will be a fantastic application for quantum computers themselves.

The high fidelity part is referring to the error correction that I mentioned. The circuits today are extremely fragile. They can fail to produce the expected calculation very easily. This fragility comes from all kinds of sources of coherence; one of the fundamental features of the Quantum system is that, as they interconnect, they almost become a single Quantum system as long as the external world is not stimulating and as soon as some kind of stimulation arrives. The Quantum coherence decays, so the current systems are extremely fragile in this point of view, and that is why they are talking about quantum practicality. What is the level of reasonable expectations of the Quantum computer being reliable? Without the calculation having to be executed multiple times and so on. What do I think about it? It is a necessary step. I am not worried about the temperatures too much because we are becoming quiet too good at managing those temperatures. Of course, there are certain thresholds of what kind of cooling is required; already at, let’s say at 10 Kelvin, we can start employing less exotic types of cooling so much more convenient, much less expensive. I am much more interested in the high fidelity aspect, the kinds of error corrections that are available.

 

Claude Shannon was the mathematician that in the 40s published a fundamental treaty called “A Mathematical Theory Of Communication”. This was extremely important because before his studies it wasn’t clear if we could have ever reliable digital communications. It was expected that errors would destroy our ability to reliably communicate with digital circuits. After Shannon’s studies, this was established as possible, and that is the birth of our digital era. The ability to use error correction on Quantum circuits that have high fidelity is the equivalent of what Shannon originally achieved for digital circuits, and as a consequence these steps are very, very important.

 

Question 5: IBM Network

What do you think about the community research called “IBM Q Network”?

The IBM Network is not like Facebook: it doesn’t have billions of people on it and it is not free. It is a very interesting and important opportunity to collaborate and to share all kinds of scientific results in an environment of collaboration. However, it is also kind of a club, so you have to ask to be a member and your membership can be denied. You will not find Chinese members in the IBM Q network for example, because the guys here at Los Alamos will say: “No, we don’t want the Chinese”. And as such, it may also renounce some ideas that could be interesting and important because the Chinese Quantum results are one of the best in the world, and not letting them into the club is a decision that has its price. It is a very, very interesting and important initiative open to to people who have the resources, not only financial but also dedicated talent to take advantage of what is possible.

 

Question 6: Quantum military

What are the military applications of Quantum technology, and specifically Quantum communications?

Here for example they talk about simulations because today, even when you do the Traveling Salesman Problem for something like a dozen nodes. The Traveling Salesman Problem is optimizing for the shortest path for the traveling salesman that has to visit a given number of cities, even with a dozen cities, or so. The problem is intractable with a classical computer; it takes more than the life of the universe to find the solution for such a small set of points. Imagine if you have an optimization problem for larger sets of variables. You know that you cannot use classical computers, so absolutely, large scale simulations for the military is an important application, and I spoke about material science. For sure the military will use Quantum computers to design better materials for offense or defense.

As far as the communication aspects, this is twofold: the military, the signals intelligence agencies, like the NSA. During a jobs fair, I picked up a little metallic button from the NSA, where they had their little desk. And if you wanted to work for the NSA you could apply to be a spy. So the NSA has been a client of D-Wave since the beginning, but not publicly. So the NSA already has Quantum computers, and the reason they are already using them is because they are storing encrypted communications today. And then they go back and try to crack them tomorrow. Quantum computers are expected to be able to crack a certain degree of encryption. And this threshold is going to increase year after year. What is the length of your encryption key that is needed to be pretty sure that a brute force attack is not going to be able to solve it. Well, if you have a four or six letter password already today, in a few minutes without Quantum computers your encryption will be broken. If you have a 10 or 15 letter password. That is not the case. Are Quantum computers already able to correct that? It is not sure, but the NSA is already thinking along those lines. The opposite is what we also covered in the course “Quantum Encryption”, because Quantum cryptography or Quantum encryption is not only going to be such that Quantum computers cannot decrypt it, but Quantum cryptography, coupled with Quantum communications, has the feature of alerting you if someone tries to intercept it what you are saying. Someone using Quantum communication channels would know if the NSA is trying to store the message for future decryption.

That is why the Chinese Quantum satellite is so interesting. Doesn’t matter what is the use today: the NSA is already worried because if this kind of Quantum communication is widely deployed, not only it cannot be decrypted, they cannot even be intercepted because the interception itself will be detected. So you cannot plan ahead and say: “Well, I cannot decrypt it today but maybe I can decrypt it next year so let me in the meantime store the message”. You cannot achieve that theater. What does this mean? Does this tip the balance? Well, working out all the consequences is a very fun strategic exercise. There is a worldwide push for weakening encryption, and I think it is extremely misguided. “If this attempt passes weakening encryption, the world is going to become a less secure place rather than a more secure place”. The proponents of this are myopic, in my opinion, and I find it rather paradoxical that it feels that the US is the first to push for this, and persuading other countries like the European Union to adopt this kind of weakening, where for example the goal is add end to end encryption. I think that it’s a horrible goal.

The Electronic Frontier Foundation protects internet freedom, so it is not surprising that they are so dramatic about this. Actually, this has been published very recently. There are, they say, a lot of proposals from the FBI and the Department of Justice who want to stop end to end encryption. Isn’t it paradoxical that it looks like China is ahead in assuring end to end encryption? We will see who is going to win in this battle, that is not only a technological one, but also one for a kind of society that we want in the future.

Question 7: Space exploration

How far out are we from using Quantum for space exploration?

In one sense, we are already using Quantum technologies in space, because the examples of the global positioning system are such that they wouldn’t work without us understanding Quantum technologies. It also wouldn’t work if we didn’t understand and employ Quantum technologies. Possibly the question is related to Quantum propulsion, if there is such a thing.

One of the future units is going to be about energy, and one of the things that we are going to talk about is fusion energy in our current nuclear reactors that are all based on fission. We break atoms and that frees up energy. But for the past 30 years there have been experiments for creating fusion reactors that are exactly like the sun: to push atoms together, and get energy from pushing them together. The first are big atoms like Uranium that breaks up and frees energy. The second fusion is for light atoms like Hydrogen and Deuterium that when you push them together, you create energy. We are hopefully close to being able to create fusion reactors. When we are able to create fusion reactors, we will also be able to create fusion rockets. And certainly, these are based on Quantum technologies. They are impossible to even imagine, without understanding quantum technologies.

There are of course farther out concepts, and even when we develop a theory of Quantum gravity we may also develop different kinds of propulsion that are based on our ability to manipulate space time. And at that point, we could be in a Star Trek or Star Wars situation of being able to say that we have a Hyperdrive! I actually don’t believe that it’s possible, and I believe that it should not be possible because the implications for me are that we live in a simulation actually, already. The simulation argument is something that we may discuss as well in in one of the sessions, but the ability to explore space further than our solar system will require a completely new technology, and whether that is fusion drive or the ability to do different things, even though we won’t go far faster than the speed of light. If we want to become an interstellar species going beyond what Elon Musk is already building to become an interplanetary species, then certainly this will be also thanks to Quantum technologies.