Analysis of “Quantum Computing” as a Product

BACKGROUND/PROBLEM PERSPECTIVE: After reviewing much of the available material on IBM’s Quantum and the general approach to Quantum computing, I felt somewhat conflicted. While I was able to watch the video with Dr. Talia Gershon, read about the 2019 Quantum Award Winners, and look through other featured Quantum news items, I was unable to reproduce a similar emotional state I had to only hearing about the SpaceX rocket landing backwards.

It’s not for lack of appreciation about the achievement of QC, but from the standpoint of a “Quantum as a Product”, the wow factor isn’t there… but it could and should be. For most people familiar with the “science fiction” of anything Quantum there’s an expectation of Star Trek-like teleportation, time travel, sending messages faster than or equivalent to the speed of light (via entanglement), “Quantum Leaps” through time and space, the show Dr. Who, and scenes from Guardians of The Galaxy or The Avengers.

I’m not saying it’s fair that is what Quantum Computing is compared to (consciously or not), but the cultural weight of anything associated with the word Quantum makes people think of the far-future. That the actual quantum machine made by IBM is compared to a chandelier (in the Gershon video) and the majority of use-cases are academic; it feels anticlimactic to the promise of the innovation.

REGARDING THE STATE OF QUANTUM COMPUTING, TODAY: [this section was added based on feedback that the following sections didn’t meet several reader’s expectations of what Quantum Computing is. For IBMers working on Quantum, this section may be unnecessary, but useful to understand my line of thinking in the remaining sections; even IBMers not working on Quantum may find it useful ]

I am aware of the “hype” surrounding cyber security issues, encryption, and the power of quantum in general. I say “hype” because after reading IBM’s Quantum Computing and Cyber Security Report, the report on exploring Quantum Uses Cases for Manufacturing, QC use cases for financial services, Getting Your Financial Institution Ready for Quantum Computing, and Building Your Quantum Capability I was struck by how speculative so much of the language used is, especially considering the above articles are written by those doing the actual engineering of Quantum Computers, to sell, as a product. The contrast between how QC is presented by a business that intends to capitalize on it is almost stark in comparison to the wild (and sensational) promises made by tech evangelists.

These articles use less-speculative language: QC Use Cases for Healthcare, QC Use Cases for Life Sciences, QC Use Cases for Chemicals and Petroleum. AND yet, while the outcomes suggested seem quite reasonable, it is worth noting that QC would only serve as a tool to make impossible solutions possible, not that it is the solution itself. To put it differently, while QC may make it possible to model nature, proteins, chemicals, DNA, there is no guarantee that what is modeled in a computer can be synthesized in a factory or a lab. Though we may discover a way to make industrial spider silk (which has already been done), does not guarantee there is a profitable way to manufacture it. To put it dramatically: even if a Quantum Computer could mathematically deconstruct a human and reassemble it without error, or even a hamburger, it is unlikely that it would be financially possible to do it in reality.

Perhaps I’m reading something that isn’t there, but my impression across all content is this: every article talking about the benefits of QC is also a forward-looking “call to action” for early adopters and first movers to transform theory into practice. Despite many probable “use cases” it seems that an actual “use case”, on par with the hype presented in WIRED or a Press Release, has yet to be achieved. Or, better to quote the following article (IBM’s Taking the Quantum Leap): “Despite claims to the contrary, no one has yet used quantum computing to solve a business problem that could not currently be solved with a classical computer. Only a few, relatively small quantum computers have been built.“

To reiterate what I’ve written above, the article Quantum Supremacy co-authored by Jay Gambetta, IBM Fellow and VP of IBM Quantum, seems to make the case that there is currently nothing a Quantum Computer can do that a Classical Computer cannot. To quote: “we urge the community to treat claims that, for the first time, a quantum computer did something that a classical computer cannot with a large dose of skepticism due to the complicated nature of benchmarking an appropriate metric”.

As an outsider looking in, there’s this feeling or impression that while QC may be “better” than supercomputing, it doesn’t seem to be “10x” better. 10x in the sense as popularized by Andy Grove (Intel) and John Doerr (Kleiner Perkins). Besides articles like “Major Quantum Computing Advance Made Obsolete By Teenager” we can look around the world we live in, consider the computing power necessary to run Google Stadia, or Facebook’s Oculus, or AR/XR mapping, or Tesla’s Self-Driving cars, Drone Swarms, Netflix’s Recommender System, The Human Genome Project, Watson winning Jeopardy, and whatever real time calculations are necessary for SpaceX rockets to re-enter the atmosphere and land backwards??? … we can SEE all that has been done WITHOUT quantum computing, and reasonably wonder how long before Quantum Computation can become profitable for its own sake, in the same way that “Augmented Intelligence/ML” is. At the moment, it seems like “quantum” may have a steep uphill battle for profitable adoption, given its main advantage seems to be the promise of solving complicated math problems, faster.

A WIDER PERSPECTIVE: So, to get a better understanding of how Quantum computing might be use-cased, I reached out to my friend Rishabh Uniyal, a Graduate Research Assistant at CERN. I thought if there were anyone I knew personally that would have a sense of the boon that quantum might bring, RIGHT NOW, it would be him. In speaking with Rishabh, I discovered there are some analyses they run at the lab that might take 10 days. A quantum computer that could cut the time of that analysis down to 1 day would indeed be something to look forward to, but in a larger sense, there would be the initial penalty/cost associated with learning a new way of doing things, and as an early adopter: having certainty that the computations were running correctly.

In talking, it seemed the main obstacle for QC to catch on is not whether or not the technology can be built, but whether or not a compelling use-case will be found that will force companies to embrace it, or not. Whether or not QC can become a tool, or a platform for tools that will give companies an unbeatable edge.

With that in mind, we also realized that QC seems yet to have a compelling WOW moment. While people can become intellectually excited (or claim to be) over complex calculations being performed faster, it’s hard for that excitement to become contagious when the reward is a mostly abstract time-saver. If mathematical prowess is the superpower of QC, then a dramatic display of something made possible (not just faster) might be key to an explosion of Quantum Computing. NOTE: while the allure of QC seems to be exactly that (making the impossible possible), from the standpoint of QC as a product — it, Quantum Computing, ought to be the hero, and not a supporting cast member. Circling back to the section above; Quantum Computing that makes it possible to model some impossible aspect of nature is, for lack of a better word, a “so what“, if whatever it models does not create some intrinsic competitive advantage. To put it differently: a pharmaceutical company that uses Quantum Computing to model “previously impossible” products that are ineffective, dangerous, or too expensive to manufacture is NOT a win.

A CRAZY IDEA: To the point/goal of quantum computing in a starring role, Rishabh and I started entertaining some “science fiction” ideas. Things that would likely be impossible without the calculative power of QC. The most intriguing idea we had was a “flying carpet”. Yes, you read that correctly. A carpet made from millions of electro-magnetic fibers/sensors controlled remotely by a quantum computer that could alter the state/polarity and cause it to fly. While Rishabh could not say with 100% confidence that he could design this, he felt that such a thing COULD exist, and that if it did, it would likely need quantum calculations to work.

From a “product” standpoint, the company that builds a flying carpet using quantum computing would be the one to usher in an age where technology truly became indistinguishable from magic.

The idea “behind the idea” is to take QC off its ivory pedestal, to unhang the chandelier and drop it under the hood of a Lamborghini, or Tesla. For example: if IBM were to work with Tesla, to create a first of its kind event, where self-driving cars competed in something like an Indy 500… it might not be the most “erudite” win for quantum computing, but if it were considered previously impossible, or unsafe, for unmanned vehicles to race each other in close quarters at speeds of up to 220mph/355km per hour, that is the kind of dramatic product win that quantum computing needs.

REASONABLE SOLUTIONS FOR WIDER ADOPTION: One might reframe the entire approach towards Quantum in the same way that we now consider A.I. to be “Augmented Intelligence” instead of “Artificial Intelligence.” Perhaps something like “Quantum Turbo” or “Quantum Nitro”, an offering focused on enhancing the speed and performance OF ANY pre-existing “process heavy”/”mission critical” application.

In the context of IBM’s vision for AI existing as a utility, quantum computing might exist in the same way: as something a company can plug into without having to understand how it works in order to reap the benefits.

For example: if IBM could go to Epic Games, or Riot, or Blizzard, or Facebook (now that they’ll soon be running titles with Oculus) and say: imagine if you could a host million players interacting with each other, in a single instance, just as if it were the real world, with ZERO lag.

Another example. I read that Pixar has one of the 25 largest supercomputers in the world; a “render farm” composed of 2000 machines, and 24,000 cores which it used to render the film Monster’s University in 2 years. Imagine going to Pixar, or any “Hollywood” Animation Studio that needs rendering and being able to do it 10x faster. Would a QC product offering that made it possible for animation studios to bring movies to market much faster than the competition be a compelling reason to “plug in” to “QC on Demand”. Also, how much money might they save from reducing production time, or how much more realistic might their final product be? Would a fully animated movie, with the realism of live action, done on a fraction of the budget that either would cost, be compelling?

Another example/thought: “real time fraud detection” that is not only defensive, but offensive, due to massive computing power. As it stands, it would seem when someone attempts identity theft they go uncaught and unpunished. However, what if: with the power of QC to process larger loads, the IP address from any single suspicious transaction could be cross referenced with the IP address from EVERY other suspicious transaction.

Or what about advertising technology? Could quantum computing make it possible to serve the same exact advertisements to groups of friends in real time? From what I understand, ad-tech companies like MediaMath can already do this for high profile clients, but it is extremely costly to do so. Perhaps it’s misguided to think this way: but it seems that anything financially costly in computing is also costly in terms of actual computation. Could IBM’s Quantum Computing reduce the costs associated with advertising, and finally allow small businesses to compete with the majors?

Could “Quantum Computing on Demand” be sold to financial companies not to improve analysis, but merely as a “guarantor” that trades will be processed instantaneously?

---

CONCLUSION: Quantum Computing IS the future, but if the power and essence of QC can be reduced/boiled down to “faster math”, there may not be any reason to hold it back behind a dam of lofty goals. If the quote “Despite claims to the contrary, no one has yet used quantum computing to solve a business problem that could not currently be solved with a classical computer.“ still holds true, does it make sense to treat Quantum Computing as a chandelier that may be out of reach/out of touch forever, OR might it be time for quantum computing to make a quantum leap downwards in terms of expectations and burn some rubber?

For, if QC cannot convey its value as a “money maker” in addition to being a “time saver”, then I believe it may struggle to be as impactful as it should be.

Also, the supply of talented QC experts in the open source community will be influenced by how much demand there is for people with the ability to program QC solutions. It’s worth noting that the type of person with the aptitude to do quality Quantum Computing might be, at this very moment, considering if mastering TensorFlow, becoming adept with Wolfram, or entering a Kaggle Competition are more (or less) profitable uses of their time. If QC were widely “use cased”, even if it only made use of a fraction of its ability, then would that not speed up its development due to the use?

Summing it up, because Quantum Computing is a true innovation, something groundbreaking, it cannot be a slightly better alternative, but something 10x better… OR, something that makes an existing solution 10x better.

For products and services that already exist and integrate with QC: QC will have to give them an edge so vast that they might seem to be entirely different products (as would be the case with a gaming shard capable of hosting an unfathomable number of players), OR, as with the movie studio rendering example, change the way movies are made to the extent that Netflix changed how they’re distributed. OR, make the impossible possible (such as catching identity thieves in real time), or making magic seem real as would be the case with a flying carpet.