Seven Things I Learned on the Way to Not Achieving My Career Goal
Harry Shum
Executive Vice President, Artificial Intelligence and Research
When I graduated with my PhD in robotics from Carnegie Mellon, I had a career goal—to become a computer science professor. I would help shape young minds and contribute world-class research to the field.
I aspired to emulate my professors, like Raj Reddy, who delivered some of the greatest lessons of my life. It was all set. I’d follow the path to tenured professor in about ten years.
But that didn’t happen.
I made a series of choices that took me away from that goal. I don’t think I realized it, but with hindsight and experience, I can see that I was driven by another force.
Years ago, the mathematician Richard Hamming gave a speech, You and Your Research, that stuck with me. He asked: “What are the most important problems of your field? If what you are doing is not important, and if you don’t think it is going to lead to something important, why are you even working on it?”
Many of us think about a specific career goal—to be a manager, to be a VP, and on and on. But instead of asking “What do I want?” perhaps we should ask, “What does my field need?”
If we do, Hamming’s suggestion to follow the most imperative problems could lead us to where we’re supposed to be.
Here are seven lessons I’ve learned in (perhaps unknowing) pursuit of this philosophy.
Lesson #1: You can’t do everything.
Straight out of school, I decided to become employee #4 at a start-up working on virtual reality, then a technology far ahead of its time.
At a start-up, you have to do everything and that’s still not enough. At the time, I had a newborn. I quickly realized that two things are mutually exclusive: having a baby and doing a startup. I chose the baby!
For the first time, I realized that my time and energy weren’t endless. You can’t do everything at once.
Lesson #2: Before you go broad, go deep.
I joined Microsoft Research (MSR) when it was just getting started. Somehow, I knew it would be a place that would break new ground.
I met so many extraordinary people like Rick Szeliski, who taught me the importance of really digging into fundamental problems like motion estimation in computer vision. I learned when you take something on, own it, write about it in a compelling way that changes minds, and do it really, really well, it will lead to something bigger. He was right.
Together with Rick, I wrote a lot, including one influential paper in 1997, titled “Creating full view panoramic image mosaics and environment maps.” Today when you take a panorama with your cell phone, you’re probably using our algorithms!
The more you seek imperative problems and solve the tough challenges, the more you put yourself on the path to leadership. Become an expert in something and really make your mark—then branch out.
Lesson #3: Storytelling matters—even for engineers!
In research, business, and life, how you communicate your ideas may be even more important than the work itself.
I learned this from SIGGRAPH—the TED of the computer graphics and interactive techniques field. Over a decade, SIGGRAPH taught me new standards of quality through the high bar set for presentations.
Even as engineers giving technical presentations, you need stories to explain your ideas to peers, to inspire people to contribute and advance your work. The best work is nothing if people don’t believe in it.
Lesson #4: You get what you measure.
I had decided to take the position as director of the new MSR lab in Beijing—and over four years, I truly found out what it means to be #1, the person in charge.
When we set out, we didn’t know what success would look like for an industrial lab for a multinational company in China—we were a first! We developed three goals: (1) advance the field of computer science, (2) contribute technology to Microsoft’s products, and (3) benefit Chinese academia and local industry.
And we worked tirelessly to achieve those goals. Defining success metrics early on really put the lab on the map. My colleagues in China would turn MSR Asia into one of the leading labs in the world. Define your goals wisely.
Lesson #5: Control the controllable, observe the observable, and leave the rest alone.
I was asked to return to the U.S. and join Bing, a new effort for Microsoft at the time, as VP of Product Development, although I had little engineering experience in program management, testing or development. I had to relearn the basics: how to survive, learn quickly and add value.
I figured out that solving the most important problems in Bing required deep research knowledge: machine learning for search quality and distributed systems for search infrastructure, to just a couple! So I went back to MSR to recruit over 50 people.
There was tremendous pressure on our inexperienced team to compete with Google. We had to persevere through the hardest of times, and we disagreed a lot. During this time, I developed a saying: “Control the controllable, observe the observable, and leave the rest alone.” People get agitated too quickly by things that are not working, or they push against things that are too difficult to change. You must first look at what’s happening around you. If you can’t step back and observe the big picture, there’s not much else you can do.
Lesson #6: Think of your career as a series of projects.
I met Jim Gray, a Turing Award winner and great technical leader, at MSR.
I once asked Jim, “You worked in MSR, and in SQL. It seemed like you never worried about whether you were on a product team or research team.”
Jim’s response was that you shouldn’t define your career by your title or discipline—he said that “I follow projects where I can make an impact.” He wasn’t worried about whether it was product or research.
Instead, he wanted to think about the kinds of interesting projects you can work on, the hard problems where a team can come together to solve something big. Don’t get caught up in categories. Instead, dive in.
Lesson #7: Always walk in the middle of the road.
Wherever you are on your career path, you’ll do a lot—you’ll make decisions, you’ll code, you’ll create, you’ll achieve. But more than doing, be. Who will you be? What will you be known for because of who you are?
Long ago, as a young Chinese student, I was introduced to Confucius’ teaching: the doctrine of the golden mean. The Chinese is actually written as 中庸之道, which literally is about walking in the middle of the road and keeping your direction.
To me, the essence of what Confucius taught was about listening, with balance, to both sides, being thoughtful and respectful. People can be extreme without knowing if their extreme position is right.
As you walk the middle, don’t burn bridges. You never know which one of your peers will become your next boss, which of your interns will go on to create the next unicorn.
Be generous, be open, be kind.
You never know what’s on the horizon. And maybe one day down the road, I’ll be an even better professor.