04.03.2025 In News By rogler

Interview with Jens Eisert on Quantum Computing, Google Collaboration, and Outreach

Google Quantencomputer | © Google Quantum AI

In an interview with MATH+, Jens Eisert, Professor for Theoretical Physics (quantum theory) at FU Berlin, shares how his passion for quantum technology began, discusses groundbreaking research in quantum machine learning, and reflects on the future of quantum computing. From collaborations with Google to inspiring young minds, he emphasizes the importance of rigorous science, public outreach, and long-term research commitment.

When did your interest in quantum technology begin, and what is your current focus?

My interest in quantum technology began when I was searching for an interesting topic for my PhD after earning a master’s degree in mathematics and a diploma in physics. Around 2000, the emerging field of quantum information fascinated and inspired me because it combined ideas of physics, computer science, mathematics, as well as the notion of computation, turning them upside down while also challenging conventional ideas about computation. I was so captivated that I read all the major papers available in the field within five weeks, which was still possible at that time.

At the time, quantum technology was a theoretical concept with little practical application. However, in the last five or six years, advances in building quantum devices have made the field more tangible and quantum computers might soon become a reality.

My current focus is on rigorously applying mathematical ideas to shape future information technologies while exploring their implications for physics—without getting caught up in the hype.

What major successes have you achieved in recent years, and are there any connections to MATH+?

We have made significant contributions across multiple areas, particularly through collaboration guided by ideas within MATH+. One key success involves the intersection of quantum algorithms and machine learning. While quantum algorithms are known to solve some structured problems more efficiently than classical ones, the question remains to what extent they can provide an advantage in machine learning tasks. Together with Klaus-Robert Müller (TU Berlin), we achieved the first affirmative results on this topic through a MATH+ project on Quantum Machine Learning.

Another notable achievement involves our work on optimization, which is central to combinatorial optimization problems in scheduling and routing—tasks that are ubiquitous in industrial applications. In collaboration with Jean-Pierre Seifert (TU Berlin), and now also with Sebastian Pokutta (TU Berlin/ZIB), we have been exploring how quantum algorithms can optimize complex tasks more efficiently.

Is there a unique project or result you are particularly proud of?

I am incredibly proud of our work on making quantum algorithms more practical through a data-driven approach. In a MATH+ collaboration with Christof Schütte (FU Berlin/ZIB) and others, we advanced methods to extract meaningful conclusions from quantum data. I have also pursued this data-driven approach in quantum technologies and quantum physics to push the field further.

Regarding machine learning, I was especially pleased with a paper co-authored with Wojciech Samek (Fraunhofer HHI) and Thomas Wiegand (TU Berlin) on explainable quantum machine learning, which connects the fields of explainable AI and quantum notions. Explainable machine learning asks, “What is it that the machine learns?” This question is crucial in classical machine learning, and we successfully merged the fields of quantum algorithms and explainable machine learning, also within MATH+.

A particular highlight was our January publication—a joint effort with researchers from Freie Universität Berlin, Zuse Institute Berlin, and WIAS, including Michael Hintermüller. We developed new tensorial data structures to model open quantum systems more efficiently, enabling large-scale simulations that outperform previous approaches. This interdisciplinary collaboration, which also involved a long-standing, trusted partnership with a team from Munich, combined ideas from quantum information, applied mathematics, and computer science. It showcased the strength of MATH+ in fostering cross-disciplinary innovation and collaboration.

Overall, MATH+ provides in our field a unique environment for applied mathematics where mathematics, computer science, theoretical physics, and experimental physics converge—making it an exciting space for excellent research and collaborative exploration.

The United Nations declared 2025 the “International Year of Quantum Science and Technology.” What does this mean for you and your work?

It means a lot! It highlights quantum science and increases public awareness. Quantum theory has profoundly changed the world—modern technologies like superconductors and lasers rely on it. This remarkable scientific revolution began 100 years ago, with foundational work in the 1920s that still shapes our understanding today.

Within three years, a completely new and almost indescribable picture of the world emerged. There was no internet at the time, and hardly any flights for collaborative work, yet researchers discovered a new theory of nature along with its mathematical underpinning. A remarkable book by John von Neumann, written from 1928 onwards, and published in 1932, provides a profound mathematical foundation for quantum theory. That book is still more modern than most textbooks available today and fundamentally changed the world.

For decades, quantum technology remained a theoretical and abstract concept with little hope for practical applications. However, recent advances—particularly in quantum error correction and device development—have made real-world applications possible. This special year celebrates those achievements, but it is important to avoid. unrealistic hype and expectations.

Personally, this year means engaging in more outreach activities. I recently gave a public talk on quantum computing at the Marburger Hochschultage and participated in a radio interview on quantum science for German public radio. As publicly funded researchers, I believe it is our duty—and very important—to share our work with society.

You collaborated with Google Quantum AI to precisely calibrate the Hamiltonians of the Sycamore quantum chip, achieving unprecedented accuracy. Can you describe the significance of this study and collaboration?

The collaboration began unexpectedly—I was in Brazil when I received an urgent call from a friend at Google AI. They were struggling to calibrate their quantum chip because their data model wasn’t working. This was the famous Google Sycamore chip, which first demonstrated a presumed quantum advantage—where quantum algorithms outperform classical supercomputers.

Together with two exceptionally talented PhD students, we identified the problem: their spectral recovery lacked the necessary precision due to resolution limits, requiring us to develop new super-resolution techniques. However, it took us three more years to understand how to address the noise in the data. This realization led us to develop entirely new methods to resolve the issue.

Four years later, in January 2025, we published a paper in Nature Communications introducing this groundbreaking approach to learning the Hamiltonian—the operator describing a system’s evolution—from data with exceptional accuracy. This precision is crucial for making reliable predictions about the chip’s behavior—particularly for the Sycamore chip, which achieved Google’s breakthrough quantum advantage.

Our study tackled a critical technical challenge and highlighted the role of academia in advancing quantum technology. Despite Google’s resources, the breakthrough came from an academic team at Freie Universität Berlin in Berlin Germany. This is a true MATH+ success story—combining advanced mathematical techniques like super-resolution and manifold optimization to solve a real-world problem that even a tech giant couldn’t crack.

Quantum computers are supposed to solve problems in many fields, from logistics, medicine, and drug discovery to material science. Where do you see the most significant potential for future quantum computers?

Quantum computing is exciting—but only if we stick to what is actually possible. Some applications are overhyped, while others are underestimated. Our job is to rigorously investigate where quantum computers can truly make an impact.

All the mentioned fields hold promise where quantum computers could make a big impact. For example, logistics involves complex optimization problems—a research area within MATH+—where quantum computers could help solve such problems more efficiently.

Fields like logistics and optimization are usually NP-hard, meaning that neither classical nor quantum computers can solve all instances efficiently—in what is called polynomial time. The key question is: In which cases can quantum computers offer an advantage? Some claims, like the recent Handelsblatt headline suggesting quantum computers accelerate all solutions, are simply false. Identifying their real strengths is our mission.

One area with strong theoretical backing is quantum simulation, particularly for quantum chemistry and material science. Around a third of modern supercomputing power is already dedicated to these fields, so the potential for quantum computers here is substantial. The challenge now is to translate these theoretical advantages into practical applications, such as drug discovery.

In fact, we recently published a paper within MATH+ demonstrating a quantum advantage in a machine learning task—specifically in generative modeling with unstructured real-world data. This is an exciting step toward bridging theoretical insights with practical applications.

Some risks associated with quantum computers, such as the potential to break encryptions, can pose security threats. How do you view these concerns?

I’m not overly worried. Quantum computers could, in theory, break RSA encryption, which secures online banking, messaging apps like WhatsApp, and much more. However, breaking encryption is an extremely complex task and won’t happen overnight. Meanwhile, researchers are already developing post-quantum cryptography—encryption methods resistant to quantum attacks.

That said, once powerful quantum computers exist, they could also reveal past secrets, which may have historical and geopolitical implications. However, the benefits of quantum computing far outweigh the risks. Every transformative technology, including AI, comes with challenges. For instance, AI affects truth perception and work culture in unpredictable ways. Compared to AI, quantum computing poses fewer immediate societal risks if properly managed.

What is the connection between artificial intelligence and your research?

Machine learning, a key part of AI, connects to quantum research in two major ways:

  1. Using AI to advance quantum technologies: AI techniques can assist in designing quantum devices, solve problems in quantum chemistry, and improve quantum error correction. Within MATH+, we apply modern AI methods to address challenges in quantum chemistry and enhance quantum devices.
  2. Using quantum computing to advance AI: This direction is equally fascinating. Much of our MATH+ work has focused on using quantum computing to advance AI. We were among the first to prove that quantum computers can accelerate specific AI tasks. One of our most-cited papers, co-authored with Jean-Pierre Seifert, a cryptographer from TU Berlin, demonstrated a quantum advantage in AI—a key milestone in the field.

Both directions are incredibly exciting, and we continue to explore how quantum computing and AI can complement each other.

What are your plans and ideas for the future?

I plan to continue developing rigorous mathematical ideas and bringing them into practical applications. My goal is to make a real-world impact—helping to build devices, advance the field, and drive technological progress.

I have a long list of concrete research topics centered around this vision: to think rigorously and use science to improve the world. I remain as fascinated by science as I was as a PhD student—so much so that I still get up at 4 a.m. every day to work on new ideas and unresolved questions.

Structurally, I also wish for a greater appreciation of science, both from the public and from policymakers. While initiatives like the “International Year of Quantum Science and Technology” are promising, science isn’t always valued as much as it should be—especially when it comes to research budgets in regions like Berlin. Scientists must do a better job of explaining how their work drives societal, technological, and economic progress. This is especially important in areas like Berlin, which urgently need innovation and investment.

When I advise the German government, I emphasize one key point: sustainability. Instead of short bursts of funding, we need long-term commitments to support research and innovation. Sudden changes in funding can do more harm than good. Maintaining leadership in fields like quantum science requires consistency and a stable research environment. I hope that Germany—and notably Berlin—will continue to invest in and value the scientific progress we’ve achieved.

You are very committed and involved in explaining your field to the public, even to elementary students. What do you think is essential in outreach, and how can we inspire more young people—especially girls—to engage with mathematics?

Outreach is extremely important, though I sometimes feel a bit alone in prioritizing it. It’s our responsibility to explain to the public—who fund our research through taxes—why our work matters. It’s not their job to understand us; it’s our job to communicate clearly and effectively. I try to do my part by giving public talks at venues like the Urania in Berlin, the Deutsches Museum in Munich, and through media like public radio.

Beyond public understanding, outreach is vital for inspiring young people to engage with science and mathematics. Our future depends on cultivating curiosity and encouraging more students—especially girls—to explore STEM fields (known as MINT in Germany). There is a genuine economic need for more future experts in mathematics, physics, and engineering.

We must show young people that mathematics isn’t dry or dull—it’s creative, exciting, and fun. Every Wednesday, my wife and I teach a mathematics course at an elementary school in Berlin, where 25 children eagerly participate. We cover advanced topics like topology, fractals, and polytopes in playful and accessible ways. For instance, after a breakthrough in Penrose tiling in 2023, we 3D-printed tiles, made dough, and baked cookies shaped like the tiles. The kids loved it. Our course takes place before regular school hours at 7:30 a.m., and it’s encouraging to see children—mostly girls—enthusiastic enough to insist on attending. We aim to break stereotypes by showing that math is for everyone. Whether exploring geometric series with older students or examining fractals in broccoli with younger ones, we focus on making abstract concepts tangible and fun.

I also encourage my team to participate in outreach. Recently, five of my students gave fantastic presentations at Freie Universität’s Science Day, speaking to high school students considering mathematics or physics. It’s crucial to motivate young scientists to engage with the public early on and share the excitement of their work.

Outreach takes effort, and convincing others—especially senior colleagues—to participate can be difficult. But I firmly believe it’s a core part of our job and essential for the future of science.

Thank you very much for sharing your insights into your research and activities!

 

© private

Jens Eisert is a professor for theoretical physics (quantum theory) at Freie Universität Berlin, where he is based at the Dahlem Center for Complex Quantum Systems. Since 2019, he has also been affiliated with the Helmholtz-Zentrum Berlin and the Fraunhofer Heinrich Hertz Institute.

His research focuses on developing new technological applications from quantum mechanics and exploring the complexity of quantum systems. Combining theoretical physics and mathematics, his work addresses practical problems in the field. Eisert has received prestigious awards, including two ERC grants, a EURYI award, and a Google NISQ award, recognizing his contributions to quantum science and quantum computing.

LINKS: