A Future Filled With Possibilities

Two 2025 Nobel Prize in Physics Laureates joined the 75^th Lindau Nobel Laureate Meeting, but for one it was not his first time in Lindau. Michel Devoret attended as a Young Scientist in 1976, an experience he descried as “very inspirational”, particularly when he met his scientific hero, Paul Dirac (1933 Nobel Prize in Physics).

Fifty years later, Devoret – the third Lindau Young Scientist to return as a Nobel Laureate, after Bert Sakmann (Nobel Prize in Physiology/Medicine 1991) and Morten Meldal (Nobel Prize in Chemistry 2022) – took to the stage on Monday 29 June. In his Lecture, he not only described his Nobel Prize-winning research, he also shared his experiences and wisdom with the next generation of researchers (just like Dirac did on many occasions).

Building on the Josephson Effect

“I was able, during my postdoc, [in the mid-1980s] to do an experiment that not only used quantum mechanics but tested its foundation,” he began. “The purpose of the experiment was to test whether emergent collective variables like currents and voltages in a circuit obey quantum mechanics.”

More simplistically, the experiment aimed to show whether the weird properties of the microscopic quantum world could be upscaled and seen in useful macroscopic objects, specifically electric circuits, that we can see with our eyes and which normally obey classical physics rules.

Critical to achieving this were two properties. “In order to see quantum effects, you need nonlinearity but that’s not the only element, […] it is very, very important to suppress dissipation as much as possible,” outlined Devoret. “Up to that stage, there was only one element which had no dissipation and enough nonlinearity: the Josephson tunnel element.”

Named after Brian Josephson (1973 Nobel Prize in Physics) – who first predicted the Josephson effect that the junction demonstrates and was in attendance for the Lecture – a Josephson junction (or tunnel element) consists of two superconductors separated by an extremely thin insulating barrier.

With this element in place, the three 2025 Nobel Prize in Physics recipients – group lead John Clarke, Devoret and doctoral student John Martinis – built a superconducting electric circuit and showed that an electrical current could ‘tunnel’ right through the Josephson junction, even when it should not have had enough energy to cross the gap, at least according to classical physics. Moreover, they proved that their circuit could only hold energy at specific energy levels that would only increase in a stair-like pattern, an effect known as energy quantisation.

Tunnelling and energy quantisation are hallmarks of a quantum system: the trio of researchers had built a macroscopic quantum circuit. “What we accomplished was that we made artificial atoms, systems that behave like atoms,” Devoret enthused. “We were pioneers in the field of quantum superconducting circuits”

Quantum Supremacy

It was fascinating to hear the same story from a different perspective when Martinis delivered his Lecture later in the week, on Thursday, 2 July. However, knowing Devoret had already described their joint work in detail, Martinis did not want to dwell too much on this. Instead, he focused on how those 1980s experiments led to today’s research into quantum computing.

“Quantum mechanics is not just the physics of the small, […] it’s actually bigger than that,” shared Martinis. “It can include manufactured systems with a chip, […] so that now our quantum playground is not just the periodic table of elements, it’s also new devices that we can manufacture in an integrated circuit fabrication line.”

Martinis is of course referring to the holy grail of quantum engineering: a working, useful general-purpose quantum computer. Quantum computers’ promise famously lies in the qubit; the quantum version of the traditional bit. Where a common bit can represent either a 0 or a 1, a qubit takes advantage of the quantum phenomenon of superposition to be able to represent a 0, a 1 or a state where it is any proportion of both 0 and 1 simultaneously – opening the door to a whole new world of possibilities. When you have more than one qubit, other quantum mechanical phenomena, particularly interference and entanglement, come into play too.

In 2019, while leading Google’s Quantum AI team, Martinis took a huge step toward this holy grail, claiming quantum supremacy (i.e. performing a calculation that is impossible for a standard supercomputer) with the 53-qubit Sycamore chip. Google’s announcement stated that Sycamore had performed a computation in 200 seconds that would have taken a state-of-the-art supercomputer 10,000 years to compute. “This is not a useful algorithm, per se, but it’s an algorithm where you can check how well the quantum computer works,” explains Martinis. “It really showed that […] we could move forward to build a quantum computer.”

After leaving Google in 2020, Martinis co-founded private company QoLab in 2022.  The purpose of QoLab, explained Martinis, is to advance semiconductor chips for quantum computing. “What really enabled the computer revolution was the mass-production of the microprocessor in the 1970s,” he said. “So that’s what we decided to do in our startup company, put all the pieces together to make a big quantum computer.” Work continues towards this aim.

Evolution or Revolution?

Although Martinis and others have made significant progress since Google’s 2019 quantum supremacy announcement, useful general-purpose quantum computers remain aspirational rather than an impending reality. In a Panel Discussion featuring Devoret and Martinis, as well as William Phillips (1997 Nobel Prize in Physics), Heidelberg Laureate Forum Alumna and #LINO75 Young Scientist Victoria Sánchez Muñoz (Université Libre de Bruxelles, Belgium) and Young Scientist Christophe Valahu (University of Sydney, Australia), panellists discussed if and when such quantum computers will be made, and what kind of impact they could have.

Though there is less pessimism in the science community than a decade ago regarding whether a useful quantum computer is even possible, Phillips represents a popular viewpoint: that useful quantum computers are not just around the corner. “I don’t believe there’s anything in physical law that says you can’t make a quantum computer,” he explained, with a twinkle in his eye. “But at the same time, there’s nothing in physical law that says you can’t make a fusion reactor – and people have been claiming that they can make fusion reactors since I was a little boy.”

Martinis agreed, and explained a core reason why we will not be seeing useful quantum computers any time soon: “The basic problem with the quantum bit, as compared to the classical bit, is that […] you can’t make a real quantum computation without at least a small error,” he said. “This is fundamental in most of the qubits we make, and because of that, it’s just a lot more difficult to make a quantum computer.”

However, others were more optimistic, including Devoret and Valahu. The latter described how quantum engineers typically quantify the performance of quantum computers by the number of ‘gates’ (instructions or operations) it executes before making an error on average. “In 2019, we were sitting at a bit better than one in 100, and I had a healthy level of skepticism, [but] then you fast forward to today, and we’re at one in 10,000,” Valahu explained. “That forced me […] to realize that there is no fundamental engineering challenge.” Supporting this view, Devoret added that he thinks “we need on average an improvement by a factor of 10 in all metrics”, but that should not “take that many years”.

Cybersecurity Threat

Assuming useful quantum computing becomes a reality within young researchers’ lifetimes, panellists were asked what impact these powerful machines might have on science and society. Everyone agreed that the biggest will be in cybersecurity, and in fact this impact is already being felt.

The standard RSA algorithm (named after computer scientists Ron Rivest, Adi Shamir and Leonard Adleman) used across industry and society to keep information secure relies on the difficulty of finding which two prime numbers were used to create a specific huge number. It would likely take current supercomputers trillions of years to crack a single 2048-bit standard RSA key. But given a quantum computer with enough stable, error-corrected qubits, Shor’s algorithm (named after computer scientist Peter Shor) could crack RSA and other modern encryption schemes rapidly.

This is why society needs to switch to post-quantum encryption (encryption schemes even quantum computers would find hard to crack) immediately, warned Devoret. “Even though we do not have quantum computers that can break RSA yet, it is dangerous for people to encode with RSA because their adversaries can record their messages and decrypt later,” he said. “We have to change a lot of hardware, and so this will take time, and it has to be done now.”

Outside cybersecurity, the panellists were keen to play down the potential capabilities of future quantum computers. As mentioned in the associated preview blog post, optimists suggest quantum computers could deliver improved fertilizers, discover new drugs, design better batteries and harness fusion power, among many other claims. This enthusiasm needs to be tempered, the panellists agreed. “I know there’s a lot of applications for quantum technologies, but to me they are like the extras in your vehicle,” said Sánchez Muñoz. “I can still drive without them.”

But it was Martinis who perhaps summed up the potential of quantum computing best. “The biggest and most obvious misconception is you can take any classical algorithm and speed it up with a quantum computer,” he said. ”It’s only certain problems that you can speed up, but the nice thing is for certain problems you can speed it up greatly.”