When people ask me if they should work in industry or in academia, I usually advise them to first figure out what they want to do. I.e., what is the mission that you’re on? What is the change you want to make happen in the world? When you have an answer to that, you can then figure out if a company or a university is a better place for you to execute your mission.
Here’re some things I like about working in the business world:
- Access to significant resources to do big projects. AI research is increasingly capital intensive, requiring huge data and computational resources. These are easier to get in a company.
- Strong sense of teamwork. When you don’t need to worry as much about authorship order or making sure this work counts toward your PhD thesis, you can better execute with a strong sense of teamwork and go after team goals and celebrate team successes.
- Rapid decision making (depending on the company). I love working in a nimble environment where we can rapidly direct resources to where there’re needed, ranging from quickly building a new compute cluster, to buying a large amount of data, etc.
- The ability to directly help huge numbers of people, through launching novel products and services.
Here’re some things I like about the academic environment:
- Ability to explore almost any topic under the sun. For example, at Stanford I started recording educational videos. Initially no one considered this “real” Stanford work; but this turned into Coursera (and benefitted Stanford too). At Stanford, when my students and I like felt like building an open-source robotics platform, we could also just do it without justifying it to anyone. This led to the creation of ROS, a very successful open-source platform.
- The freedom to spend 100% of your time learning and not have any direct output even for years. Companies like Baidu are very supportive of employee growth and often have people spend months just learning/studying; but it would be harder to have employees do this for years.
- The ability to earn a degree. Even today, having an advanced degree is helpful. Universities and companies can both be very good at developing talent, and society is getting really good at recognizing ability, regardless of where you gained that ability.
For myself, one of the missions I’ve been excited about is creating universal access to the world’s best education, and I thought a company (Coursera) would execute best on that mission. More recently, I wanted to develop AI technologies that let us help hundreds of millions of people, and I thought a company (Baidu) would be best for that mission. But there’re plenty of other worthy missions, such as teaching students, and certain areas of investigation, that would be better to execute in a university.