Rahko: The future of pharmaceuticals is Quantum machine learning
It’s not an unfamiliar story: a small team of academics working in something quantum-related spinning out from their university to capitalise on their combined knowledge. Rahko, however, isn’t the culmination of decades of researching, lecturing and professorships in academia. A product of University College London’s ‘Conception X’ accelerator programme - a “deep tech” initiative that helps PhD students to turn their research topics into businesses - Rahko began with two PhD students and today is, amongst other accolades, one of the select few partners of the Amazon Quantum Solutions Lab. In the accelerator programme’s promotional video, both Rahko-branded promotional materials and Rahko’s CEO feature prominently under the voiceover. Though the company isn’t named explicitly, it’s one heck of an endorsement.
“I moved to UCL and met my co-founder Ed [Grant during the PhD,” says Leonard Wossnig, Rahko co-founder and CEO. “We were both in the same research group. We were working on quantum codes and quantum algorithms: I was on the theoretical side, Ed was more on the practical side, so we had both the ‘theorist’ and the ‘practitioner’ in there. Ultimately, we built a framework to enable our quantum machine learning research. At some point after that we engaged with a large tech company. Based on their feedback, we decided there was commercial value in what we’d built and decided to found a company.”
When pressed, it’s apparent that Wossnig is being somewhat diplomatic about this interaction. The framework that he and Grant built during the course of their PhDs didn’t just attract casual curiosity or interest: the big tech company made an offer to buy their work. The founders declined the offer, recruited two more co-founders, and Rahko was the result - a company that Wossnig describes as “one of the world’s most advanced teams in quantum machine learning.”
Rahko’s focus of research is on material simulation - particularly in pharmaceuticals. Material simulation at the quantum level has game-changing potential for everything from batteries to building materials, but pharmaceuticals - notoriously expensive, notoriously risky - are arguably the best candidate for seismic change. While the team waits for error-resistant quantum hardware to arrive, Rahko can, in the meantime, deliver its clients results on classical computers.
“The beautiful thing about quantum machine learning is that you can run many of these algorithms on classical hardware and already push the state of the art in quantum chemistry for our customers,” says Wossnig. “Once we get really useful quantum devices, the methods we have just improve.”
Like many quantum CEOs, Wossnig is acutely aware of the balance between selling a product and overpromising on a technology already buoyed by hype, the media and techno-evangelists. Quantum is not, he clarifies, a technological panacea. Rather, quantum will become part of a computational ecosystem, leveraged to perform tasks beyond the capabilities of classical computers without necessarily ever relegating them completely to history.
“Our perspective is that quantum computing will never entirely replace classical computing, even in these sectors like chemistry,” says Wossnig. “So it will always be important for us to span the whole space. We integrate what we do with classical computers into quantum computing, because quantum computers will be efficient for certain tasks - but there are other tasks where classical computation will always be cheaper and faster.”
Wossnig addressed quantum’s hype problem with pleasingly blunt concision in 2019, saying (in a statement announcing a £1.3 million seed investment from Balderton Capital) that the majority of people still found quantum computing “mysterious” and had been left in a confused, head-scratching state over whether quantum computers were going “to save or break the world.” Even within the quantum community, the messages can be mixed: companies and teams dedicated to hardware and simulation skew positive on quantum’s shockwave potential, while security companies report the vulnerability of encrypted data and the death of privacy. Quantum computing is a broad church - Wossnig says that one of the challenges in the early days of Rahko was zeroing in on a specific area with such an abundance of tempting alternatives - but sees quantum overall as a saviour technology, in material design at least.
“The potential for quantum computing is huge. There are a lot of promising areas in which one could work,” he says. “But there’s still a lot of uncertainty. People are wondering which of the two [the saving or ending of the world] is actually going to happen, but I think depending on which area you look at it’s still unclear and on which timescale. I think it’s going to save the world, personally, particularly with regard to quantum chemistry where we’ve got very clear evidence that there are a lot of use-cases. For example, we know that once we have fault-tolerant quantum computers we can simulate chemical systems that are totally out of reach today. And I think that while this is the case for quantum chemistry - materials design, pharmaceuticals, all these areas - it is unclear to me whether that also holds true in other areas like security or finance.”
In fact, in whichever industry quantum first makes itself commercially useful, Wossnig isn’t convinced the average person will notice - at least not at first. While early computers arguably went from business machines and academic curios to coveted consumer devices with the advent of e-mail, quantum computers won’t have that same obviously transformational effect. No monoliths of whirring beige wedged under the desk; no ‘one-more-thing’ iDevice that idly plays with molecules when it’s bored. Instead, existing computers - monolithic or portable - will over time outsource more and more of their processing to unseen quantum workhorses. The only visible change will be over the longer term: a sense that technology development is somehow moving faster than before.
For those people outside of the quantum bubble, Wossnig has a convincing present-day analogy in the form of HPC.
“I think for the average person it’s not entirely clear what the advantages of quantum computing will be, because quantum computing will play a similar role to High-Performance Computing,” he says. “High-Performance Computing is extremely important to industry today, however the average person will only recognise the impact it has by looking, for example, at the progress of technology. So, the everyday person will see an advantage, but it will be a ‘behind the scenes’ advantage. I’m not sure there will be something like the classical computing ‘E-mail Moment’."
Instead, Wossnig suggests that quantum’s impact may seep through in the rapid advancement of technologies responding to market demands.
“Looking at the pharmaceutical industry at the moment, there is a growing demand for productivity,” says Wossnig. “Because of the increasingly aging population, we need more plentiful, cheaper drugs. One immediate application - if pharmaceutical companies adopt these methods and are able to really have this step-change in productivity - it would of course enable the production of a huge number of new drugs - with massively reduced development times.”
Pharmecuticals are an obvious target market: more medicines, cheaper medicines, better medicines - each candidate molecule assessed at a pre-preclinical stage for a fraction of the cost on quantum hardware.
Quantum machine learning - quantum computing in general - likely won’t be as well suited to grand unveilings on a stage a la Bill Gates or Steve Jobs. But facilitating a sea-change in the pharmaceutical industry? Surely that’s more impactful than the iPod.
For more information, email Amit Das directly on firstname.lastname@example.org
To visit the Rahko website: https://rahko.ai/