Creating software for quantum computers to solve some of the most difficult computational problems known.
It doesn’t take long sitting around with the team from Zapata to feel like something magic is going on. The company, which uses the new field of quantum computing to solve difficult–or even otherwise impossible–mathematical problems, faces the unenviable task of explaining exactly what quantum computing is. “With classical computing, a bit is either 0 or 1,” says CEO Chris Savoie, talking in a conference room at Harvard’s chemistry lab. “In a quantum computation, quantum bits can be on and off at the same time.” Peter Johnson chimes in, explaining that what a quantum bit (or qubit) ends up being depends on how you interact with it. On the other side of the table, Jonny Olson tries an analogy. “Imagine you have someone voting between two candidates,” he says. “If you bring forth a candidate that agrees exactly with her philosophy, she is always going to vote for that person, but if you start deviating from that, there is some probability she might vote for the other candidate. But once you force her to vote, she has to choose.”
If it all sounds very confusing, it is, Zapata’s founders allow. “We’re trying to analogize stuff that is not analogizable,” Savoie says. It’s much easier, however, to understand what quantum computing can do. In classical computing, you write a program, and the computer tick-tick-ticks through all of the computation it needs to solve the problem. “When you have a bunch of qubits working in this way, it allows you take a shortcut,” Savoie says. “Instead of clicking through all the possible solutions, it compresses a chunk of those tick-tick processes and therefore speeds up the calculations.” That can allow the quantum computer to reach an answer much more quickly than a classical computer ever could—and even solve equations so complicated that even today’s most powerful supercomputers can’t solve them. That, in turn, could lead to radical new breakthroughs in our understanding of chemistry and other disciplines.
Zapata originated in the Harvard chemistry lab of Alán Aspuru-Guzik, a pioneer in applying quantum computing to chemical calculations. He had gathered an impressive group of postdoctoral researchers, who came together in 2016. Physicist Peter Johnson was focusing on mathematically understanding quantum entanglement–the way qubits interact with each other–and developing a new technique to correct errors in quantum computers from environmental interference. Yudong Cao, a computer scientist, was focusing on integrating quantum computing and artificial intelligence, developing methods which use quantum computers for machine learning tasks. Jonny Olson, an expert in quantum optics, developed a machine learning method for discovering properties of molecules only quantum systems could learn—a so-called quantum autoencoder. Rounding out the crew was a fourth-year graduate student Jhonathan Romero Fontalvo, a chemist from Barranquilla, Colombia, who has used quantum computing to analyze the interactions of electrons in molecules.
The group worked so well together, that rather than watching them leave for the likes of Google and Intel, which are developing their own quantum computers, Aspuru-Guzik proposed they create their own quantum computing company. Instead of focusing on making the computers, however, they would focus on creating the applications for them to use. He named it after Emiliano Zapata Salazar, a leader of the Mexican Revolution, to express just how revolutionary the company could be to traditional computer science. To run it, he tapped Savoie, a longtime entrepreneur who bridged the gap between machine learning, biology and chemistry. Years ago, a company Savoie founded was part of the original DARPA project that developed the technology for Siri. More recently, Savoie had been running a photonics company that used quantum chemical calculations, and licensed AI technology developed by the Aspuru-Guzik group. As luck would have it, Savoie was leaving his current company at the same time Aspuru-Guzik was forming Zapata, and he leapt at the chance to focus on quantum computing directly.
Quantum computers are still in their infancy, with a number of different designs currently under development. For example, some use aluminum-based superconductors cooled close to absolute zero Kelvin to enable quantum states, others use lasers to entangle trapped ions suspended in a vacuum. Rather than work on aspects of quantum computing hardware, Zapata focuses on what might be called quantum computing software, developing applications that could apply quantum computing to some of the world’s most intractable problems. For example, chemists have long struggled to predict how electrons and other particles inside atoms interact with each other. “We’ve known the set of equations we need to find the answer since the 1920’s or 30’s, but solving these equations is insanely hard,” says Romero. “Even for a small number of electrons, a traditional computer quickly runs out of computational power.” In fact, adds Savoie, there literally isn’t enough computational power to solve them, “even if you used every bit of silicon in the universe.”
Instead, chemists have had to use algorithms to approximate answers, then explore reactions experimentally. Using quantum computing, however, scientists could calculate the exact values to describe interactions between molecules, such as binding energy between atoms. They can use this information to custom-design new catalysts to speed up reactions and develop new drugs at a much more rapid rate. “It’s going to fundamentally change and disrupt a lot of things in chemistry in ways that will contribute to society,” Savoie says. “Being able to develop drugs using quantum computation could take years off the development of life-saving drugs.”
That is just one example of the possibilities quantum computing holds. Quantum resources could also jumpstart the next generation of neural networks and AI—an endeavor that has been driven largely by founder Yudong Cao. In addition, the team is developing algorithms to solve problems such as simulating financial markets and optimizing routes for a company like FedEx. “There are many kinds of optimization or linear algebra problems we can focus our energies on,” Savoie says. Ultimately, quantum computing is so new that Zapata will be on the forefront of discovering all of the potential it may have. “Our hope is that by actually finding uses for quantum computers, we’ll gain a new understanding of what quantum computing is,” Johnson says. If they are successful, then one day people may have as clear a grasp of quantum computing as they do of classical computing today.