Quantum computing represents the next leap forward for computing, opening the door to powerful machines that can answer questions beyond the capabilities of today’s computers. Researchers, academics, and governments are already using AI systems to solve challenging problems across sectors including healthcare, energy, and financial services, so what impact will quantum computing have on AI and which sectors might benefit the most from their marriage? Experts debated this question during a panel discussion hosted by the Center for Data Innovation.
Denise Ruffner, chief business officer at Cambridge Quantum Computing, explained that quantum computing is an emerging technology that takes advantage of the laws of physics to solve computational problems. This means quantum computing is best suited to solving problems that resemble those found in nature—exponential problems. However, Joseph D. Lykken, deputy director for research at Fermi National Accelerator Laboratory and member of the U.S. National Quantum Initiative Advisory Committee, stressed it is too early to know exactly which industrial and academic problems this maps to, noting that further research in this area is needed.
Markus Pflitsch, chairman and founder of Terra Quantum, pointed to the growth of venture capital investment as a good indicator that major advancements can be expected in the short term. Freeke Heijman, special advisor on quantum technologies to the Dutch Minister of Economic Affairs and Climate Policy and director of strategic development at QuTech Delft agreed, adding that the growth in investment, startups, talent, and qubits, which are the quantum version of the binary bit, over the last seven years have already exceeded initial expectations.
On the tangible benefits of quantum-enabled AI, panelists agreed quantum computing can improve AI models by reducing the amount of data needed for training AI systems. Quantum-enabled AI can also enhance AI by handling larger datasets with many more strongly correlated variables. Unlike classical computers, quantum devices can compute huge numbers of potential outcomes simultaneously.
To explore the use cases where these characteristics are best exploited, Heijman and Hodan Omaar, policy analyst at the Center for Data Innovation discussed two examples that the Dutch government is investigating with industry. First, researchers are exploring how quantum algorithms could improve and accelerate existing methods used in genetic prediction for the Netherlands’ seed trading industry. Current prediction techniques rely on small datasets that are highly correlated, and existing quantum computers with a limited number of qubits can already run applications using small datasets, which means there is potential for quantum to improve this process. Second, researchers are exploring how quantum algorithms can optimize traffic flows in the automotive industry.
On the challenges to widespread adoption quantum-enabled AI faces, Eline Chivot, senior policy analyst at the Center for Data Innovation recalled that many problems that will impede these devices scaling up, such as errors and cooling costs, are yet to be solved. However, Ruffner argued quantum computers are, in general, more energy-friendly than traditional high-performance computing systems. She noted that doubling the processing power of Summit, the United States’ most powerful supercomputer, would be as energy-intensive as building an entire power plant, whereas exponentially increasing the capabilities of a quantum computer requires incremental additions in qubits, which would have minimal additional impact on energy.
Engaging in international collaboration in this field will be essential to make rapid progress and be competitive. Quantum is a deep tech field, meaning that it is a technology based on scientific discoveries and engineering innovation that requires significant investment in basic science R&D to drive innovation in the field. Countries that identify common, high-priority research opportunities and share their ecosystems of top talent, investors, and resources will not only foster quantum innovation, but cement their own collective leadership. In turn, having multiple global leaders in quantum and AI will stimulate the development and adoption of these emerging technologies by putting pressure on firms to innovate.
To ensure the responsible development of quantum technologies, including by developing research agendas, promoting workforce development, and spurring innovation and commercialization, panelists concurred that there should be international governance frameworks for quantum AI similar to those that already exist for AI—for instance by expanding the Global Partnership for AI, a G7 group established to provide cooperation between nations on AI, or adopting the Tokyo Statement on Quantum Cooperation, which the United States and Japan signed in December 2019.
The EU has already rolled out programs creating opportunities for joint collaboration, such as the European Flagship initiative to foster R&D, the Quantum Communication Infrastructure (EuroQCI), and the EuroHPC Joint Undertaking (EuroHPC JU) that develops hybrid infrastructure for HPC and quantum computing. Similarly, the U.S. National Quantum Initiative aims to coordinate and accelerate U.S. quantum R&D, but also to connect the U.S. scientific community with others in Italy, the United Kingdom, and Canada.
On the role of quantum tech startups in advancing commercialization, Pflitsch and Ruffner noted startups are essential early adopters.The quantum startup ecosystem has grown from 10 companies in 2018 to 360 today but this sustained growth requires more access to capital and financing. To this end, connecting the various communities of researchers working on IT, AI, and cybersecurity, will be essential to advancing the field.
Finally, panelists agreed that to take advantage of the many potential applications of quantum computing, countries need to invest not only in hardware and software, but also in workforce development to foster the expertise needed to develop, deploy, and adopt quantum-enabled AI. To scale this technology and to be competitive in this field will require training the entire chain of talent to work with the technology at all levels and addressing the talent shortage.
In the foreseeable future, and for nearer-term applications, AI will take advantage of quantum computing in certain cases, but hardware capabilities for quantum-enabled AI do not currently exceed classical computing capabilities. It is already possible to run applications on quantum computers that currently have a limited number of qubits, before these devices scale up in the future.