My Struggle with Impostor Syndrome
I sat in the meeting room filled with other engineers and stared at the projection screen. Like a newborn child, I couldn’t understand most of what I was looking at or hearing. I recognized almost none of the acronyms and comprehended little of the discussion that was happening. An intern, taking a break from his Ph.D. program to spend a few months at our company, was presenting the results of a summer’s work. He had done something impressive with deep learning (and branch of machine learning and artificial intelligence), building some incomprehensibly effective solution for some mind-bogglingly challenging problem. Though I could somehow tell that he was competent and effective, the details of what he had done were almost completely lost on me. My job title was Senior Deep Learning Computer Architect. Technically, I was supposed to know what was going on.
I found myself in that predicament partly because I took a ten-year sabbatical from my employer, a company that I had busted my balls to make successful. Ten years of sabbatical would leave anyone a little rusty. Sure, I spent some of those ten years practicing various forms of engineering, but I was mostly focused on getting a Ph.D. in clinical psychology, which involved a lot of talking to people about their mothers (and fathers). As my doctoral research was coming to an end, I began to study deep learning. Then, once I had defended my dissertation, the money I had made in the early years at NVIDIA had almost run out, and I had to look for work.
The easiest way for me to make money was to do engineering work, and the easiest way for me to make a lot of money was to join the right start-up, as I had when I joined NVIDIA in May of 1998. From that, I made millions of dollars in just a few years of intense work.
There were two start-ups that I was interested in. One was in Canada, and they understandably wanted me to move to Canada to be physically present with the small, tightly-knit research team. I didn’t want to move to Canada. The other start-up was in Palo Alto. Even though the CEO of this company seemed very interested in hiring me, and I met with him many times, their first software engineer wanted me to work unpaid for three months, which I couldn’t afford to do. Finally, one night, unable to sleep from worry, I emailed Jensen Huang, the CEO of NVIDIA and asked, “If I came back to NVIDIA, could I work on deep learning?”
When I’m not in a self-doubting mental state, I think that I must be somewhat capable. I recount to myself some of the ways that I’ve succeeded: I have a master’s degree in electrical engineering from Stanford, with a good GPA (3.6/4.3). I designed a very advanced adaptive noise cancellation system during my undergraduate studies, a solution that made millions of pounds for the company that apparently stole the invention from me. I flushed many bugs out of NVIDIA’s first successful graphics accelerator chip (the Riva 128), and I designed large sections of various graphics processing units (GPUs), including the first GPU. By anyone’s objective standards, I believe that I am a relatively good engineer.
On the other hand, when I’m tired or overwhelmed, or when I’ve drunk too much coffee, or when I have had more than a little alcohol the previous day, or when I haven’t meditated or exercised consistently enough, then I start to question everything. At those times, I begin to doubt myself and my capabilities. At those times, I seem to switch from a growth mindset, a mindset with an attitude that I can learn and grow in any way I need to, into a fixed mindset, a mindset in which I’m not intelligent enough, or in which I don’t know enough, or in which I’m just not capable.
One day a couple of weeks ago, after meeting for lunch with an accomplished and well-respected engineering leader at work, I sent him a message on Slack telling him about how I was feeling anxious because I didn’t feel that I was making enough progress on a new project that I was working on. He revealed to me that he sometimes struggles with impostor syndrome and this led me to realize the extent to which I do as well. As I sat in the semi-private phone booth at work and wrote back to him that “I wish we could see ourselves the way that others see us,” tears started to run down my cheeks.
It’s not fun when I’m beating myself up. It’s worse than not fun. It’s like a form of hell. There’s nothing more that I can do to make something go faster apart from harming myself by losing balance in my life, by sleeping less, by spending less time with my wife, by forfeiting this therapeutic writing practice, or by allowing all kinds of stressful personal tasks to accumulate. I also know that none of the ways of trying to force a difficult project to move faster actually make it move faster. Even though I know that those sacrificial approaches only slow things down, I often feel compelled to keep trying them.
I’m certain that all of my success in life has come from being playful, and from being engaged in an effortless, curious flow. I know that worrying about how well I’m performing, or putting pressure on myself to make something magically move faster, has only ever slowed things down, yet I keep finding myself worrying and pressuring myself. I keep finding myself in these mental spaces where I feel trapped and on the verge of panicking. What if someone discovers that I’m really crap at engineering, that I’m really stupid, that I don’t know what I’m doing? What if they realize that I incapable and that I have no value to give them?
There are many reasons that I left my highly-paid engineering job back in 2007 to get a Ph.D. in clinical psychology including that I was isolated, with very few social connections. I was working from home in the UK, recently separated and then divorced from my first wife, and unable to spend much time with my young son. In my mind, I had failed as a husband and as a father. Having moved back to the UK in an attempt to keep my family intact, my role at work had gradually shifted from being a manager and leader of a large new architectural re-design back to that of an individual contributor. I had been through an experience of de-motivational management practices, which I now suspect my body orchestrated its own escape from by breaking an ankle. Those broken bones had left me incapacitated, and unable to work, for months.
These external circumstances nurtured a slightly different flavor of my impostor syndrome. I was specifying low-level tests for a large and complex unit that I had birthed, a unit that a team of dozens of people had then built. Because of my mental state, and because of my circumstances, I felt disinterested in what I was doing, finding it hard to gain the traction of curiosity. I felt that I wasn’t adding as much value as I could. During my annual reviews, I had always been told that I was a “top performer.” In those days, employees were rated on a scale of one to four, and I was a “one,” meaning that I was in some very high percentile of performers, maybe even the top 5%. (I now don’t remember exactly what being ranked “one” meant.) Even during this period of final exhaustion, I was still, amazingly, rated a “top performer.” Writing that now, I feel sick. There’s something that I hate about being raised above others. Why should I be ranked against the very people that I’m supposed to collaborate with. These are my friends. Why does it have to be a competition? I’m crying now as I write this. Yeah, I’m a fucking “winner,” and I don’t want to be a winner if it means that it’s just me standing alone, with my team considered beneath me or somehow less than me.
Some people are more conscientious than others. Some people work longer hours. Some people have a lot of insight. Some are very creative at problem-solving. Some people are great at big-picture thinking. Some people go through great surges of productivity followed by troughs of hibernation. Some inspire others. Some help others. I’m tired of Silicon Valley’s bullshit story of the “genius” or “brilliant” engineer who outshines everyone else. We build massively complex systems, and we do it as a team. As a team, we can pull together and support each other.
Because my personality style is enneagram type three (the achiever), my gift (and also my curse) is my indomitable work ethic and persistence. Without self-awareness of my own personality, it would be easy for me to proclaim that “adaptable, efficient, conscientious high-performers are the best!” But I don’t think that any functional organization can be built only with threes. Any sustainably successful organization must nurture a flourishing ecosystem of diversity, including diversity in personality fixation (both intensity and style).
When I left my position of great competence at work and started formally studying psychology for the first time, it was a true experience of being an incompetent beginner. I was no longer the leader that I had been, but just a student, sitting in a class, listening to a professor talk. Amazingly, for the first time in my life, I had to learn how to think critically (in the academic sense), to construct engaging arguments that artfully wove together evidence into a compelling narrative. Up to that point, almost all of my academic engineering submissions had been in the form of technical writing, or code, or equations, or complete functioning systems. Interestingly, at this and other times of true incompetence, I never had the anxiety that I would be found out. I was openly incompetent, and reveling in it.
Since returning to work, I have had the experience of joining new teams and projects twice. In both cases, I have revealed to others my frustration with what feels like slow initial progress and my feelings of incompetence, and, in the process, I have discovered that others feel the same way. It seems to often be the biggest elephant in the room. I suspect that impostor syndrome is imposing itself everywhere, not just where I work now, but in all high-performance environments. We’re all sitting there at our computers, hacking away, trying to solve seemingly ridiculously trivial problems, struggling to make the progress we believe that we should be making, hoping that nobody notices how crap we are. Meanwhile, when we’re able to slow down for a moment and notice what’s actually happening inside of us, we find that the biggest thing that’s occurring is this gnawing anxiety, which when left unchecked by self-awareness actually eats away at the curiosity, the creativity, the persistence, and the flow that truly contributes to outstanding performance.
Many of us are infected with a disease that we can’t talk about and that we think nobody else suffers from. Even writing this, I worry that you are going to read this and think, “What’s wrong with you!? I don’t suffer from this!” or “You think too much!” I’m worried that you’re going to think that I really am an impostor. Now my dirty little secret is out of the bag.
Let’s roll with that then. Okay, I’m a shitty engineer. I’m not a robot. I’m not highly-performant all the time. My energy and interest ebbs and flows. Sometimes I find things really boring and then it’s hard to keep working on them. I often struggle to find the right balance between just getting something done (checking a box) and “polishing” it. By the way, I think that “polishing” is often an unnecessarily derogatory term for paying-down technical debt and ensuring that boxes can get checked more quickly in the future and that work doesn’t have to be repeatedly re-visited. When I’m upset, I can’t work. When I don’t trust my manager or supervisor then I also can’t work. Amazingly, writing this right now is helping me to see myself in a more balanced way. It’s not that I’m a shitty engineer, it’s just that I’m a human. I’m actually a human being. I’m a human being first and an engineer second.
We are all human beings, even the most autistic, Spock-like among us. We’re all just humans. Engineering, or whatever it is we produce (e.g. scientific research, or writing, or parenting, or governance, or athletic performance), is just something we do for fun and profit. First of all, we’re just these big sacks of flesh and blood and bones controlled by brains that were adapted for hunting and cuddling and fucking, and for making the odd tool here and there. I need to keep remembering to give myself a break, to give us all a break. There’s actually no problem here.
I remember walking up the high-street in the Clifton area of Bristol in the UK, early in my engineering career. It was seven or eight in the evening, and I was returning home from a long day at work. As the sun set, I walked past a group of marketing people that I recognized from my work. They were in a bar, still wearing their formal work clothes, having drinks and laughing and talking. I remember envying them. They were able to live such human lives, I thought. It seems ridiculous to me now, but I thought that I could never have a life like that. A few years ago, I visited the UK and walked around in the streets of central London. There, again, I saw workers, dressed for business, drinking and laughing. I liked that, and I wanted that for myself. But what exactly was it that I wanted? I don’t particularly want to drink beer, and I definitely don’t want to have to wear a suit. I want to spend time with other people. I want community and friendship.
This points to another of the reasons that I started the Ph.D. program: to be part of a tightly-knit cohort, a group of humans traveling through a carefully-orchestrated, intentional, transformational-learning ordeal.
There is now a bar at work, and people spend time there with each other, and not necessarily drinking alcohol. I now also intentionally meet with people socially at work. I meet with a different person every day for lunch. All of my lunch meetings, for the next month or two, are scheduled.
Even though it’s usually me that invites others to lunch, yesterday I met with an intern who had reached out to me. He saw my profile on LinkedIn and was intrigued by my blend of interests, from engineering through psychology. We sat together and talked about our lives, and our feelings, and our families. He told me about his father’s struggle with bipolar disorder, and that his father had recently killed himself. As I listened intently to him and tried to understand all of what he had been through, I felt tears pressing out of the corners of my eyes. After this hour of deep connection, I returned to my desk and focused intensely and effortlessly on debugging and coding.
I’ve been through many periods of true incompetence, times when I have been a complete beginner. Those are experiences that I have actively sought out and reveled in. There have also been times when I have lost balance, not had my human needs met, and reverted to old and non-adaptive habits of self-criticism and self-doubt. At those times, I have forgotten my humanity, and I have treated myself not as a human being but as a machine. In sharing all of this with you, I have deepened my understanding of myself as a human being with human needs and that my ability to be productive in the world is a by-product of taking care of myself. This has led me to increase my commitment to building organizations and systems that put people first, with the intention of not only maximizing profits but also of having a shit-ton of fun.