Quantum mechanics seem to be beyond the limits of possible understanding for most of us. But there are a few people who get drawn into it and thrive. One of them is Leonard Wossnig, co-founder of the quantum machine learning start-up Rahko. In our interview, Leonard gives us exciting insight into his career, his field of research, and what he thinks the future will bring.
Quantum machine learning. Many people cannot really make sense out of these three words. Leonard Wossnig is certainly not one of them. He is the founder of Rahko, a quantum machine learning start-up that is on its way to great success. By exploiting the latest advances in quantum machine learning, the company aims at developing models and software to predict chemical properties at high speed and low cost. With their work, Leonard and his team have already attracted important pharmaceutical and technology companies worldwide as collaborators. But is the general public ready to understand and integrate quantum mechanics into their lives? Maybe that might not even be necessary. In the following interview, Leonard tells us about his career path, the challenges about his field, understanding in the general public, and where he sees the impact of his company and quantum machine learning in the future.
Q: You co-founded a business in the field of quantum machine learning. How did you end up in this field?
A: I originally started with a bachelor’s degree in Physics, but then moved already during my Masters into computational sciences and particularly into machine learning (ML). Soon afterwards I discovered that ML, and computer science generally offer an entirely different way to understand and predict the world, and I ultimately decided to do a PhD in this area. It was during my Masters and PhD that I also discovered the exciting area of research called quantum machine learning. When I then met my co-founder, Ed Grant, during the PhD, I found someone to share this enthusiasm about the combination of science (quantum) and computer science (machine learning) with, which led to many joint research projects. It was also during these times at University College London (UCL), where we did our PhDs, that we realized the commercial relevance of the technology, and decided to found our company Rahko.
Q: How would you explain the scope of your company to an audience without a physics background? How is your company connected to chemistry and pharmaceuticals?
A: Today, when we build houses or planes, it is standard to use computers to predict whether these houses are stable, or whether the planes fly. We do this through a process called simulation, in which the computer calculates all the forces which act on a house or a plane. From these, it predicts their behavior, i.e., what will happen in reality. Houses and planes are very large objects, and for these objects, the rules and forces are very simple. This allows us to calculate and predict their behavior reasonably fast. In these cases, the simulation works very well, because we can mathematically describe nature. This is what makes it indeed a standard process in engineering today. However, if we look at objects of the size of atoms or molecules, the situation is completely different. At these length-scales, objects behave very differently because they follow the laws of quantum mechanics. The problem with quantum mechanics is that it is so complicated that it becomes very expensive and slow to use computers to predict the behavior of small objects. This inability to simulate, and hence predict quantum mechanics, has made progress in many areas very slow. For example, in pharmaceuticals or materials design, the development of new drugs or better batteries is very slow and expensive. As a consequence of this difficulty in simulating and predicting behaviour, researchers need to manufacture hundreds of thousands of new drugs and batteries in the lab, which is much slower and much more expensive than doing so on a computer.
Rahko, our company, is developing new tools based on machine learning and quantum computing, which allow our customers to do these predictions on a computer with an incomparable speed and low cost. We hope to enable the development of novel drugs, better batteries and chemicals across many industries in the next years.
Q: Do you face challenges because the majority of the population – including scientists of other fields – might not understand what you are doing? Do you have the feeling that it is more difficult to get funding for projects like yours for the same reason?
A: Generally, it is true that in deep-tech areas it is much harder to raise money. This is true because most investors do not necessarily understand the technology and the timelines of such businesses. Indeed, deep-tech companies typically require a much higher initial capital to allow sufficient progress to be made for the research and development of their product, but ultimately also have a much higher growth once this initial phase is overcome. These differences in the capital requirements of such companies are typically not easy to understand for traditional investors. However, over the last decade, an increasing number of investors have appeared that focus or even specialize in these types of investments. Our own experience was extremely positive, in particular with Balderton Capital, which is one of our early investors. Regarding the question of comprehensibility of quantum mechanics: I believe that any topic, however complicated, can be brought down to an easily comprehensible level. In my eyes, it is hence the obligation of the founders to do so and to communicate the company’s ambitions in an easy fashion.
Q: Do you think bigger companies should invest more in enhancing the public visibility and understanding of quantum physics?
A: I think that many large companies have started doing this over the last few years. I believe that it is not necessary to understand the details of quantum mechanics for the general public, but what is important is to communicate the relevance of quantum mechanics in everything that we are using today. Perhaps this is something even schools should teach more, and I am hence not sure whether the obligation is here on the side of the companies.
Q: What are your future plans with Rahko? Where do you see yourself in 10 years, and where could one expect to see contributions from your company?
A: We have a bold company vision: ‘Solve Chemistry’. Although at the moment this is of course far out of reach, it is something we are aiming for. We want to solve some of the hardest problems of today using quantum machine learning. In particular, we want to become the world-leader of quantum machine learning for quantum chemistry, enabling the design of novel drugs, better batteries and materials, and chemicals. Our team is in very good shape, and we have some of the very best academics as our advisors. I believe that this combination of great people will ultimately help us to get there.
In our interview, Leonard Wossnig managed to explain the scope of his company in simple terms. He believes that any topic can be explained to everyone, however complicated it might be – and people who are professionally involved should be able to do so. Nevertheless, he also mentions that it might not even be necessary for the general public to understand the details of how quantum mechanics work, but it is crucial that people understand its importance and the power of its potential applications. Leonard’s goal is to enable the development of novel drugs, better batteries, and chemicals within the next few years – he wants to solve chemistry. It sounds like a big goal, but he does seem to be confident of reaching it. We are excited to see what the future will bring!
Interview: Anna K. Stelling-Germani and Leonard Wossnig
