This morning, as I was prepping for my class on Quantum Integer Programming, I received this email:
Dear Dr. Tayur,
I’m a reporter with business news site Fastinform, and I’m working on a story about the state of quantum computing and its finance applications. Given your involvement in researching the technology at Carnegie Mellon, I would appreciate if you could give me a few-sentence comment with your thoughts on the field over email.
Specifically, how would you describe the pace of research in both quantum computing and its use in finance in the last 5-10 years? And what role do you expect the technology to play in finance in the near (and/or far) future?
I had not heard about this news site before. Their About Us:
Founded in New York in 2020, Fastinform aims to become the leading source of business intelligence for professionals worldwide. Our goal is to develop industry expertise to help our subscribers understand emerging risks and opportunities. Anchored in time-honored practices such as independence, fairness and reliability, we seek to innovate journalism by developing new technologies for news gathering and distribution.
I have a soft corner for entrepreneurial ventures 😏. So, I responded:
I believe the past few years, we have seen Quantum Computing v1.0 being explored towards Finance applications. Not surprisingly, many first generation Industry pilots were executed on previously available hardware with naive quantum algorithms, and consequently have not led to any commercial breakthroughs.
We are entering a new era, in 2020, that of Quantum Computing v2.0. The second generation algorithms we have now, such as GAMA, are vastly more capable; we have superior hardware choices; we better understand how to actualize quantum speedup, and so are able to tackle important problems in Finance that have defied classical methods and technologies.
At CMU, we recently engaged with Principal Financial Group, for example. Additionally, a post-doc alumnus from our quantum computing group will be joining IBM in their finance and quantum group. We should expect to see good commercial applications in the intersection of quantum and finance in the coming years.
Within a few minutes, I received:
Thank you very much for your response, this is really helpful! Could you just explain what you see as the defining difference between quantum computing v1.0 and v2.0?
I thought this was a good opportunity to lay out what I feel is going on in quantum computing 1981-present, dividing the horizon into 1981-2017 (QC v1.0) and 2017-present (QC v2.0):
We can compare and contrast in three dimensions:
1. Hardware capabilities. 2. Computing model. 3. Quantum Algorithms.
In QC v1.0:
1. Hardware: low number of qubits, limited connectivity among qubits, low fidelity and range of the coupling strength between qubits.
2. Computing model: Narrow focus on Gate/Circuit model of quantum computing.
3. Algorithms: Academic papers on theoretical analysis of worst-case computational complexity in Gate/Circuit computational model; Naive brute force industry pilots.
In QC v2.0:
1. Improved hardware in number of qubits, connectivity.
2. Expand computing model to include Ising Model, of collective state computing, not sequential state computing (which is a not too great leap from 1936 Turing machine conception, popularized in 1960s in classical computing), that was the main work horse of 1980s/1990s academic quantum computing, called Quantum Complexity Theory (QCT).
3. New Algorithms tailored for practical problems, hybrid quantum-classical approach, taking advantage of best of both worlds, through Quantum Integer Programming (QuIP).
I have summarized it in a table form (at the top of this post). His response:
Thank you, I appreciate you sharing these details — this clarifies the transition you described very well. I believe that’s all the questions I have, so I’ll get back to you once my story is finished.
It feels déjà vu (all over again😏).
In 2000, I communicated Infinitesimal Perturbation Analysis (IPA) to Fortune. This article was read by many CEOs, including that of Deere, who became one of SmartOps best customers!
In 2014, I communicated OrganJet to physician-writers who covered it in The Atlantic and NEJM.
Now, in 2020, I am communicating quantum computing to a business journalist. Hopefully, channeling Casablanca:
This is the beginning of another great friendship.