STEM Considered Harmful
Recently a few of us were comparing notes about working in high-tech after having studied liberal arts in college. The discussion led me to reflect back on my more than 30-year career. For the most part it’s been satisfying and successful. I have fond memories of many of my colleagues. But my memories also have a certain edge to them.
Part of what I remember is a kind of hyper-competitive machismo. We were always competing to see who could be the fastest to come up with the best solution to an optimization problem. Being the slowest one to figure out a given problem exposed you to ridicule: sometimes, but not always gentle. Occasionally it led to ostracism.
I suspect this attitude reflects the nature of the computer science educational system. The history of computing is one of always trying to optimize: faster CPU clock speeds, better caching, lower-latency memory and networks, more efficient algorithms and programming languages, etc.
There is also a attitude of one-upmanship that comes with the so-called “elite” colleges from which Silicon Valley engineers are typically recruited. Is Stanford better than MIT? Is MIT better than Cal Tech? In hindsight, if Facebook is the best thing Harvard can produce, should we really be recruiting from elite colleges at all?
I remember interviewing at Microsoft in the late 80’s. The entire day consisted of tests of my cleverness and the quickness with which I could figure out how to make an algorithm smaller and faster. In an era when you only had 64k or even 640k of memory available, optimization was important. On the other hand, when you build your entire world view around the ability to save time and space by assuming you’ll never have more than 255 of something, you start to lose perspective. (For what it’s worth, they offered me a job, which I turned down because I didn’t like the weather.)
Over the years I’ve participated in innumerable formal and informal meetings where we intellectually boxed with each other. I’ve watched technology startups spend money on terrible ideas because they assumed they were the smartest people in the room, if not on the planet, and could fix the rest of us with the sheer size and power of their engineering brains.
Even DevOps, which is supposedly about empathy, suffers from a certain amount of condescension and smugness: “if you have a DevOps team you’re doing it wrong”; ”survival is optional”; “if you’re not learning you’re losing to someone who is”. (I will freely confess to having been guilty of this behavior.)
We talk about psychological safety and radical candor and feeling comfortable giving each other feedback, but what kind of feedback do we mean? Corrective, or amplifying? In art critique we were carefully taught not to judge but rather to encourage. We learned to say “how could you take it further” or
“where could you take it next” instead of “it doesn’t go far enough”.
Everyone talks about the importance of STEM. I’ve started to wonder if that’s exactly what we don’t need more of. The humanities and arts emphasize exploration and interpretation. There is no right explanation for the meaning of Madame Bovary. Students are prodded towards originality rather than correctness.
Artistic breakthroughs come from accidents: spilled paint, or a broken vase. Matisse became who he was by forcing himself to view the world like a child. Picasso entered his groundbreaking cubist period after immersing himself in art from a completely different culture.
Sometimes breakthroughs come from desperation. Radiohead made its stylistic breakthrough after Thom Yorke had a nervous breakdown and considered quitting the band. The lyric “every day I waking up sucking on a lemon” expressed how he felt on tour.
My college art professor challenged us every day to push ourselves further out of our own comfort zone. Success lay on the other side of feeling dumb, and mute, and lonely, and inexpressive. My experience of the software industry has often been the opposite. We cling to comfort by trying to prove we’re smarter and faster and more confident than each other.
According to our industry, software has eaten the world. I think it’s actually the other way around. Software is no longer a thing unto itself. It’s been fully absorbed into how we live. Even pacemakers inside people’s bodies are connected to data centers that analyze and tune their functioning.
As a result, over-confidence, isolation, and self-satisfaction no longer work (if they ever did). What we need is more curiosity. Curiosity presumes humility. It proceeds through not-knowing, wondering, and wandering.
Diversity and inclusion need to include thinking and learning processes. Psychological safety is much more than feeling like it’s ok to challenge each other. It also needs to be ok to feel mute, or confused, or lost. If the liberal arts teach us anything, it’s the non-linearity of the path to genuine insight and value.
Some people have started extending STEM into STEAM, where the A stands for “arts”. Just inserting some art into a fundamentally left-brain approach, though, isn’t sufficient in my opinion. If we’re to become true system thinkers, and build things that are genuinely helpful to the society we’ve made reliant on software, we need to encourage and reward openness, exploration, curiosity, and vulnerability as the core of what we want from ourselves and each other.
P.S this topic has a personal twist for me. My autistic brain makes spontaneous and extremely non-linear connections between things. Sometimes they lead to insight and synthesis, but sometime’s they’re dead ends. The point is that one doesn’t happen without the other. It’s also not possible to be sure in advance which connections will be fruitful. Without embracing what we think are “bad” ideas, we can’t hope to find ones that are truly “good”.