Wednesday, October 19, 2016
GHC16 Plenary Session: Astro Teller, Captain of Moonshots at X
Dr. Astro Teller, Captain of Moonshots (CEO) at X
Dr. Teller's talk focused mostly on the mantras followed at X to foster innovation. He began with an anecdote illustrating how his management style has developed over the years to what it is today, and how that impacts X's culture. It started with a photograph of Dr Teller with his two famous grandfathers, one a Nobel-Prize-winning economist (Gerard Debreu) and the other the famous theoretical physicist and founder of the Lawrence Livermore Labs, Edward Teller. He spoke about how, given his illustrious family background, he always had the feeling that he was the "dumb one" in the family, and how that spurred him on to work harder and study harder than anyone else. This was what led him to eventually graduate with a PhD in Artificial Intelligence. Soon after graduation, Dr. Teller started a business, and as CEO, felt that he always had to play speed chess with his employees to be an effective manager. Six months down the line, he realized that that wasn't actually the case, and the best way to lead was to amplify his employees' skills. This is what he has dedicated the past 20 years to, and at X, although his title is "Captain of Moonshots", he spends all his time on people and culture, rather than hands-on involvement in tech work.
Dr. Teller then went on to define what a moonshot really is. At X, anything is considered to be a moonshot if its a huge problem that needs solving, if there's a radical solution that's being proposed to fix that problem, and finally, if that solution involves breakthrough technology. Some of X's recent "moonshot" projects are Google Brain, self-driving cars, and Google Loon.
The first of X's mantras to more effective innovation is to fall in love with the problem being solved, rather than the technology. For instance, saying that one wants to work on machine learning is akin to saying that one wants to build something that has transistors in it. Dr. Teller gave the example of a recent X project, headed by Kathy Cooper, which tried converting sea-water to methanol in a carbon-neutral way, as an alternative to gasoline for the approximately 4 billion internal combustion engines in use today. While the team was trying to figure out ways to effectively make the conversion, they were also thinking about alternative solutions if their original plan didn't work. Although they were eventually successful in making methanol from sea-water, the associated cost was too high, about $15 a gallon. The project therefore had to be killed, but still garnered Kathy and her team a lot of recognition at X and beyond. The main take-away from that story is that its OK to fail - if people aren't given credit for all the hard work they've put in on failed projects, then no one would ever have the courage to walk away from ideas that didn't work out. At the same time, sharing Kathy's work with the world means that other folks who want to solve the same problem have a body of work to look back at. X even makes it a point to celebrate all projects that have failed on the Day of the Dead.
Another mantra followed at X is "inspiring innovation". When people are inspired by the problems that they are trying to solve, they bring more of themselves to work, and are also encouraged to think differently. Astro Teller mentioned the famous "mutilated checkboard" problem as an illustration - simply putting the same problem in different ways causes people to shift their perspective on how to solve it. This is known as perspective shifting, and is deeply ingrained in X's culture.
A third mantra at X is "T-shaped people". X doesn't necessarily believe in hiring persons only with a suitable background for a given position. On the contrary, at X, people from different fields are welcomed in the belief that they bring fresh perspectives to the table. As an example, Astro Teller spoke of a fashion designer who was brought on board the Loon team to design and perform quality assurance tests on the balloons, ensuring that they survive in any kind of environmental conditions.
The fourth mantra - get in touch with the physical world as quickly as possible. This puts projects out into the real world early on, increasing the chances of discovering bugs and problems. For instance, recently, a self-driving car was out on a test drive in Mountain View, when it came across an old lady in a wheelchair, wielding a broom and trying to get a duck off the street. This simply wasn't a situation that the self-driving car project team had thought up unit tests for :-). The car passed that little test successfully, waiting until the old lady had crossed the street. Similarly, there are countless other similar situations that would be difficult to account for until the project is out in the real world. Sometimes, this strategy could backfire, causing the project to be killed. But the thing to note is that if one couches failure as failure, then people are going to be wary of taking new projects on. However, if things are turned around and failures are, instead, viewed as learning opportunities, then people become much more willing to experiment, even if they fail. This is a culture that managers need to foster.
The fifth mantra is that we need to balance both audacity and vulnerability. A mix of both is needed to make sure that we innovate effectively. While audacity is what drives us to try new things, vulnerability and humility are required to admit when we go wrong, and to learn new things. X's culture is to help people to achieve this balance.
(This post was syndicated from http://kodbalekaapi.blogspot.com/)