When I was starting my first job a little over 3 years ago, someone gave me advice that I keep going back to over and over again. It’s one that has shaped much of my perspective on the emerging quantum tech industry, and helped me add nuance to the conversation about quantum hype. Before I tell you what that advice was, let me tell you the context of the question that I asked to receive that advice. As always, this is my own view and not that of any employer, past or present.
One of the common (somewhat misguided, as I explain later) questions that came up in the early years of grad school, especially as I was socializing with other grad students, was whether I was interested in pursuing a career in academia or industry after graduation. This question was nonchalantly thrown around in the early years, but eventually became the way that people sorted themselves near the final years of graduate school.
As a result of this mentality, three years ago, when I was looking for my first job out of grad school, it wasn’t clear to me how I could combine what I loved (teaching) in the field that I enjoyed (quantum computing). Ideally, I also wanted to do this at a global scale. My logic was that teaching would simultaneously help me keep up-to-date with the field while getting more people excited about the potential of quantum computers. Global scale was important to me because it meant I could reach places where quantum research didn’t have a significant presence (like Ethiopia, where I was born and raised).
So, what was the advice?
I asked a university professor who had also founded a quantum computing startup if they had any advice that might help me in navigating the choice between academia and industry. Given this person’s experience in both worlds, I was hoping to hear about something that would help me achieve (what I love + in a field that I enjoy + global scale) using one of {industry or academia}. Instead of answering my question directly, they taught me how to think through my own biases. Let me explain.
The biases of academia and industry
Grad school taught me to think of ideas with a view toward quickly getting to results, by demonstrating that they are possible. Often, that meant quickly weighing a likelihood of success, and picking the path that leads to the fewest possible pathways to failure in the least amount of time. Because of this, I learned how to quickly think of the ways that an experiment could fail. My favorite conversations were always the ones where someone would come up with a cool idea, and we’d take turns looking for ways that it could be made to work by thinking of all the possible ways that it could fail, and eliminating each one until a small, well-defined experiment could be done that proves the viability of that idea. “If I made the device that does this, and I measured that signal, then we know that the idea works. Paper!” At conferences, I’d find myself listening to talks and thinking of ways that the ideas could possibly fail, and asking the speakers how they plan to handle these.
As a result, academia taught me (and many of my peers) to bias toward demonstrating that an idea is viable by thinking through and eliminating failure pathways. Without this bias, you couldn’t write and publish research papers quickly enough. Without papers, you don’t have a dissertation. Without a dissertation, you don’t graduate with your PhD. Of course this isn’t always the case (I know plenty of people who have graduated without papers and had successful careers), but the incentives were certainly there.
Industry taught me to think of a problem in another way.
Instead of biasing toward demonstrating that an idea is viable, I think about solutions to problems and how I can maximize the impact of these solutions. Instead of thinking, “I want to show that I can teach someone how to do quantum computing by explaining it this way,” I now think, “How can I assemble a team and the resources needed to educate a large number of people who are interested in quantum careers?” Now, I’m immediately thinking of the most impactful way to do something and looking for resources, instead of my academic bias which is to quickly demonstrate the possibility of that impact. That’s because I’m not limited by the ~6 years of a PhD or limited funding. I have access to new tools (like the ability to think beyond the time horizon of a PhD, the ability to hire a team, and to obtain long-term budgets) that I didn’t have as a grad student. The incentives of industry are very much aligned with large-scale impact.
Quantum computing hype
Every time I see an argument about what quantum computing projection is hype vs realistic, I am reminded about biases in academia and industry. I wonder how much these biases contribute implicitly to the conversation.
On one hand, there is a bias to be extremely careful about what projections we make because there could be many failure pathways lurking in the corner with so much new science to be done. On the other hand, there is a bias to think about a world where we have functional quantum computers and imagine what they might look like and what we might need to make and use them at scale.
The best way to see this is to compare academic review papers projecting how specific qubits will scale versus roadmaps published by the current industry players. The former anticipates challenges and lists how they will overcome these, while the latter envision stages in the development of a quantum computer based on where we are today and work toward making these stages happen. Given infinite resources and time, both would probably get to similar results, but likely differ in how they would get there in the short term.
Of course there is also plenty of quantum hype out there that is borderline bullshit. There are specific types of problems that we know quantum computers will not speed up significantly. For each one, you can probably find someone out there claiming otherwise.
Back to my own journey
Armed with this insight about biases from academia and industry, I took a big risk in February 2019. At a time when quantum education roles within industry were non-existent and my own biases favored doing what’s familiar — “quantum research”, I defined a role that allowed me to spend my time doing both. That risk paid off, and led to some amazing teamwork and products like the Qiskit Textbook and the first couple of Qiskit Global Summer Schools. Even though I’m no longer with that team today, I am thankful for the opportunity that allowed me to launch my career by combining what I love with a field that I enjoy at a global scale. I find it gratifying to see so many “quantum education” and “quantum workforce development” roles in industry today (and even a non-profit dedicated to quantum education!), and remain optimistic about the promise of quantum computing — its ability to eventually help us understand more about nature by engineering (something that looks like) nature.
How biases inform our understanding of quantum computing hype
Biases are everywhere. Going back to my years as a grad student wondering whether I should end up in academia or industry, I now realize that biases were built into my impression of each of those two choices. The advice that I received from that professor allowed me to look past those biases, and to focus on where I could achieve my goals in the best way possible.
Similarly, I hypothesize that biases are likely hidden in our conversations about quantum computing hype. I made the fancy, extremely quantitative picture below to show how I imagine this bias at play.
The picture shows statements about quantum computing that are on two sides of reality, and the number of people who believe/say these statements. The left-most and right-most statements are objectively incorrect as of today. Focusing on the two closer to the middle, left side errs on the conservative side by using words like “may” and “some” to highlight possibility of failures, while right side errs on the optimistic side by using words like “will” and “industry-scale” to highlight impact. The reality is that we know quantum computers will, in fact, speed up some problems. What’s not clear yet is how these speedups can truly be applied at an industrial scale, although one can imagine ways of doing so (with quantum error correction, for example). It’s going to take significant work, and there are no silver bullets that can solve the challenging problems we have in front of us.
The key takeaway is this: depending on a person’s biases, their understanding of quantum computing hype (and the way they communicate it) can be quite different from others. The truth, as always, is likely somewhere in the middle.
— Thanks to Nick Farina, Clarice Aiello and Dave Bacon (ordered in no particular way other than in such a way that their last name initials spell FAB) for reading earlier versions and providing thoughtful edits. Thank you!
Leave a Reply