You don't need to have a computer science degree from Stanford to be working on one of Google's hottest teams
While the terms have kind of become buzzwords in startup land, Google has had teams doing research and building AI-driven applications for years.
For example, the company established its "Brain" group five years ago and the team has since penned dozens of papers, built an open-source AI system called TensorFlow, and influenced a bunch of Google products and services like Photos, SmartReply, and speech recognition.
The team held a question and answer session yesterday on Reddit, and one of the most striking parts (to someone not entrenched in that world, at least) was reading about the crazy-diverse backgrounds that Google Brain team members have.
You might think that to be working at one of the preeminent machine learning groups, you would have to have a degree in computer science from Stanford. But you'd be wrong.
Here are some of surprising paths that Google Brain employees have taken:
- Chris Olah: Did one year of pre math at University of Texas, dropped out to help his friend dispute an arrest, got a Thiel Fellowship to do 3D printer research, started being interested in machine learning, joined Google Brain as an intern.
- Martin Wattenberg: Had a background in math, but worked in journalism for his first six years out of school.
- Doug Eck: Undergrad in English literature and creative writing, while self-training as a programmer. Eventually did a PhD in computer science, focused on music and AI.
- Dan Mané: Majored in philosophy.
- Geoffrey Hinton: A degree in experimental psychology, followed by a year as a carpenter. Before getting to Google, he did get a PhD in AI.
"Machine Learning is such a new field though that degrees matter less than you might think," senior research scientist Greg Corrado writes. "I think all you really need to get started is a college level foundation in Vector Calculus & Linear Algebra, plus proficiency in Python, C++, or similar."
Brain's newly launched Residency Program specifically looks for people with different specialties to go through a 12-month internship-like role that dives into deep learning techniques.
"We actively encourage people who have non-traditional backgrounds to apply," TensorFlow product manager Zak Stone writes. "We believe that mixing different perspectives and types of expertise can spark creative new ideas and facilitate closer collaborations with other fields."
Read the rest of the fascinating AMA here.