Dr. Hari Balakrishnan (BT/CS/1993) is a Professor at the EE and CS Department at Massachusetts Institute of Technology, and leads the Networks and Mobile Systems Group at MIT Computer Science and Artificial Intelligence Laboratory (CSAIL). He has made many contributions to the field of computer networks and networked computer systems.
Dr. Anand Rajaraman (BT/CS/1993) is a successful entrepreneur and computer scientist. He is the founder of several successful startups (Junglee and Kosmix), has headed @WalmartLabs and has worked as Director of Technology for Amazon. He is also an early-stage venture capitalist.
Hari and Anand both received the Distinguished Alumnus awards this year. They were also hostel-mates and project-mates during their stay at IITM, and proceeded to do their MS and PhD degrees in Computer Science in the US. In part 1 of this interview, Hari and Anand chat with Tanmai and Sarathram from T5E about their thoughts on academia and research in India and the US.
Tanmai: Are there any stark differences that you have seen between student groups in IIT, and in MIT, Stanford and Berkeley?
Hari: I’ve noticed that at MIT or Stanford, students are more independent and willing to try out a lot of things on their own. Undergraduates do a lot more in terms of research and side-projects than in IIT. I think in IIT there were too many courses. Reflecting on it now, the right model is to trust undergraduates once they are in their 3rd and 4th years to do things to develop themselves. They must be encouraged not to take more than 4 courses in a semester, but do something more independent.
Anand: That’s true. Also, at Stanford, there is a greater variety of courses available to students. Undergraduates don’t declare their majors until midway through their programme. At IIT, there was a bit of a herd mentality. Most of us in Computer Science used to take almost the same set of courses over the four years. In the US, you see undergraduates taking a more diverse set of courses. I think this leads to some interesting outcomes.
Tanmai: Do you think the fact that students in the US have to pay a lot for their education makes a difference to their attitudes?
Anand: It’s not clear to me. At Stanford and MIT, the top scores are often of students with scholarships, who are not paying huge amounts. I don’t think that how much a student pays affects behaviour, since most of the good students have scholarships.
Hari: I agree. The single biggest determinant of what a student does in college is the level of motivation, which depends on two factors. The first one is having inspirational teachers. It doesn’t take every teacher to be inspirational, you just need one or two to inspire you- I think that’s enough to kick a 20 year old into action. The second is the quality of the company you keep, the friends you have in college. I’ve seen it over many years, as a professor and student.
Anand: This is not to sell IIT short- I think IIT had excellent classes as well and we had interesting and inspirational faculty. It was enough to inspire me and Hari to go on and pursue interesting careers in computer science. We developed a lifelong love of the stuff that we started doing in IIT.
Sarathram: Could you contrast the industry and university scenarios in India and the US?
Anand: My experience of US industry comes primarily from Silicon Valley, which is a unique phenomenon. It’s an ecosystem where good ideas come out of many places (including places like Stanford and Berkeley) and are converted to real products that make a big impact. Lots of innovations have come out of this place. I was fortunate to end up at Stanford and move into this unique ecosystem after IIT.
In India, it is more difficult to take ideas at a nascent stage and make them really big. To do that you need a propensity to take risk, a tolerance of failure, the existence of VCs in industry to invest, and mentors. But Bangalore is coming up as a similar ecosystem in India.
Hari: I agree that Silicon Valley is unique. I also think there are other parts of the world which have perhaps less impact, but certainly enough commercialization going on. For instance, Boston has a number of start-ups and a similar ecosystem. I don’t know much about Indian industry at all, though.
With reference to universities, there are two differences that immediately come to mind. I think the tenure system in the US, where people have to produce fairly significant results in order to get promoted to tenure position is a good thing- because ultimately it requires you to build a reputation amongst well-known people in the field, and be considered an expert in some topic. The second is that research and teaching are a lot more integrated. As I teach my courses I end up finding a lot of useful things to say both from research that I’m doing as well as companies that are involved in one capacity or another. I can’t say much about Indian academia as I have no first-hand experience.
Anand: I’d like to add to Hari’s point about research and teaching. There are a lot of people, including me, who are active in the industry and actually come and teach classes. This brings a lot of industrial problems and perspectives into the classroom, which I think is very valuable.
Tanmai: What are your views on Computer Science and Engineering and its applications in the Indian context? Do you think we are in tandem with the requirements of Indian society? Is the reason that universities in the US have so much collaboration with industry because the market in the US is ready for the products that Silicon Valley makes?
Anand: If you discount cutting edge research on the frontiers of Computer Science, the market for the research is worldwide. For example, think of Google, or a browser. However, the US has a big enough domestic market that you can bootstrap using the US domestic market and then have a big worldwide impact. Traditionally, the Indian market is not big enough to do the same, except in a few specific cases. So there is an advantage to being in the US.
That said, I worked with the administration at IIT Madras to help start an entrepreneurship workshop at IIT M, and saw diverse ideas that had come up from students. Many of those ideas are applicable in the Indian context and they look very different from what we do at Silicon Valley. For example, there are were ideas for companies that make inexpensive medical devices, or build construction equipment suitable for Indian conditions. There are very interesting ideas for start-ups that I’ve seen coming out of IIT Madras, which are applicable to the Indian context.
Another important trend is that there are companies that are now starting in India that are capable of attacking the global market. A classic example is InMobi- they are based in Bangalore but their market is worldwide. This is more feasible now than before because of the cloud; it does not matter where you are as a service provider- you could be based in India and provide cutting edge research to anyone in the world.
Tanmai: Sometimes it is felt that there are gaps between academia and the short-term requirements that industry and society has. How does this gap get addressed?
Anand: I think that research universities should not look at only the needs of the industry right now. The short-term needs of industry are one thing, but you also need a blend of revolutionary new ideas and perspectives- and perhaps these will end up solving some of the short term problems as well.
Hari: Also, there’s a big difference between education and training. IITs should be in the business of education and not in the business of training people, say, just in a particular programming language because it happens to be what’s needed in the industry in the next quarter. I think the graduates of IIT should change how industry does things. Industry and customers don’t always know what they want, and if you show them a better way to do something, they will adopt it.
This is one thing that US universities have understood well. They have at least a 5-10 year horizon on the projects they’re doing. I only investigate research that has that time horizon. The advantage of that model is that you’re actually not competing against what industry might be doing in the next quarter or next year. You’re working at a longer horizon and when your project is at the end of its completion cycle, if it turns out to be a good idea, you have an edge in bringing it to the market. This is what research universities, and IITs, should focus on.
In part 2 of the interview, Hari and Anand talk about their views on careers and their personal journeys, and share some anecdotes from their student life at IIT.