|Posted On February 7, 2014|
MIT Electrical Engineering and Computer Science Professor Anant Agarwal has personified the educator-entrepreneur, as he has had a foot in academe and a foot in new ventures for more than a decade. He has led CSAIL, MIT’s Computer Science and Artificial Intelligence Laboratory, just as he was a founder of Tilera Corporation, which created the Tile multicore processor. He led the development of Raw, an early tiled multicore processor, Sparcle, an early multi-threaded microprocessor, and Alewife, a scalable multiprocessor. He also led the VirtualWires project at MIT and was the founder of Virtual Machine Works. His start-ups have largely been focused on his areas of research and areas of interest, but he had not focused on the education space itself until late 2011.
It was at that point that Agarwal taught what would become MITX’s first massive open online course (MOOC) on circuits and electronics, which drew 155,000 students from 162 countries. This overwhelming response showed the promise of having his academic and his entrepreneurial pursuits coincide. Agarwal developed a partnership between MIT and nearby Harvard to establish edX. Unlike rivals Coursera and Udacity, edX is a not-for-profit. Therefore, when Agarwal thinks about the competitive landscape among the MOOCs, his perspective is “the more the merrier.” In fact, in June of last year, edX became open sourced, and the source code, OpenedX, has led to interesting collaborations with Google, Stanford University, and even with countries such as France and China.
I spoke with Agarwal multiple times in recent months to ask him how edX is evolving, and what he foresees for the future of edX and for the academic institutions that they draw from.
(To hear an extended audio interview with Anant Agarwal, please visit this link. This is the seventh article in the Education Technology Innovation series. To read past interviews including interviews with the CEOs of Udacity, Coursera, and Khan Academy, please visit this link. To read future articles in the series, click the “Follow” link above.)
Peter High: As edX enters its third year in existence, what key lessons have you drawn thus far?
Anant Agarwal: The power of edX and of MOOCs more generally is to democratize education. People want to learn no matter their circumstance or their age, and the experience of our students shows definitively that this is the case. We have many people who are in the workforce who use edX to develop new skills to employ in their jobs. Therefore, we are thinking more broadly.
A related example is our partnership with global steel manufacturer Tenaris. Through their adoption of the Open edX platform, Tenaris will enhance their existing training programs delivered through Tenaris University to nearly 27,000 employees worldwide. We have established a comparable relationship with the IMF.
We also have announced a partnership with Davidson College and the College Board to host Advanced Placement (AP) course modules for high school students, as well. So what began as university-centric idea is migrating to the pre and post university settings.
PH: Anant, please talk a bit about the genesis of edX.
AA: At MIT, we’ve been very heavily involved in online learning as a resource for 15 years starting with OpenCourseWare. In the early part of this decade, I played around with an online laboratory called WebSim. building an online platform where several hundred students from around the world would come and conduct experiments. I had always felt that I would be able to do online lab experiments, and then in late 2011, we decided to build a platform and put MIT courses online. We formed MITX and announced it in December of 2011.
The first course we offered was on ctircuits and electronics, and the response was staggeringly good as we had 155,000 students from all parts of the world take that course. This was a difficult course with differential equations as a pre-requisite, so the enrollment really surprised us. I spoke with others at MIT and at Harvard, and we decided to partner up launching edX as a collaboration between MIT and Harvard. We decided to expand the partnership, and then we worked with other universities offering courses on our platform.
PH: Were there any misgivings, especially in the early stage, about the extent to which this might cannibalize the existing offerings? Naturally these courses do not replicate the experience of being on MIT’s, Harvard’s, Berkeley’s, McGill’s, or Georgetown’s campuses, but at the same time, it is access to some of the professors that make those universities famous. Was the possibility considered that this was a something that was cannibalizing the value of these universities’ educations?
AA: The course material from MIT courses was already available to the public with thousands of courses on OpenCourseWare. OpenCourseWare was definitely very good for MIT, and good for the world. Therefore, there has been no real strong concern as to whether this would damage MIT’s or Harvard’s reputations. Rather we felt that increasing access to learning for students around the world was a very good thing to do. A major motivation was to re-invent how education happened on our own campuses, so that we could apply new technologies to learning and really reinvent how education happened on the campuses. Online learning is a huge disruptive force, and we certainly felt it was better for us to be part of it and to fuel it for both improving on campus education as well as increasing access beyond it. That is why we launched edX.
PH: How does edX change the educational experience for students who were on those campuses? For instance, for your MIT students’ part of this is access to materials and to university professors who are not at MIT but are part of edX. I assume also that part of the value is a foundation that is laid through the use of edX’s material that then can be built upon through the classroom setting. Is that a fair synopsis or at least a portion of the value intended?
AA: Absolutely. A big motivation for starting edX was to improve the campus experience. For example, lectures with online learning sequences are at extra places on campus with interactive exercises and virtual laboratories. Now the students can watch these on their own campuses in their own dorm rooms at their own time and pace. Just imagine, you can pause and rewind a lecture, which you can’t do in a classroom. You can watch this at midnight, and we find that the largest amount of access happens between midnight and 2:00 am.
The students like it a lot, and it improves the quality of learning because the students can pace themselves. This is a form of learning called active learning where rather than sitting in a lecture and passively listening for an hour, after five or ten minutes students interact with the content answering questions or conducting virtual lab experiments. This active learning has been known to improve learning outcomes, and it is a good thing for our campuses. I like to say that online learning is like a rising tide, it lifts all of boats, and it will improve campus learning as well as anchor that sort of education everywhere.
PH: How do you foresee this changing the experience for professors themselves? As a professor, how has it changed the way in which you prepare for classes?
AA: About 1,500 students at MIT are accessing the edX platform on the campus in some way, which is about 25% of the students at MIT. This is seen as ample use on campus. The way it impacts professors like me is that in the past, I would have to go into class and use the same lectures each semester. Of course, the lectures change a little bit where I would have two or three demonstrations and I would cycle between the demonstrations. But still you are cracking the same old jokes, and the students and some professors such as myself, don’t like early morning lecture. This is good for all of us.
Also grading is the real chore that professors have to do that often times would result in arguing between themselves and the teaching assistants. Instructors would spend all day grading mid-term exams, and now we have computing technologies that do online grading and provide instant feedback to the students, which they love. In fact, students get a little green check mark when they get something right and they get multiple tries. So they don’t make silly mistakes anymore because they can check their answer and they can get multiple tries. Students are telling us that they go to bed at night dreaming of the green checkmark. There are many studies that have shown that instant feedback also improves learning outcomes.
Automatic grading helps professors and online videos and lectures help professors leverage their time for higher value added duties like interacting one-on-one with students including doing projects and performing research with students. Professors can establish a variety of apprenticeship models leveraging the technology with students resulting in higher levels of interaction with students and helping students learn how to learn.
PH: Are there certain kinds of courses, or are there certain kinds of professors for that matter that seem to be particularly well received on edX? Are there topics that suffer when transformed onto that platform?
AA: We have about 30-35 courses in a platform that are in disciplines ranging from the science of cooking, Greek history, the humanities, engineering, public health, and law. We provide courses in a number of areas, and we are constantly surprising ourselves.
Our platform technology enables us to grade free form essays and free form answers. We have three technologies to do that including the artificial intelligence (AI)-assessment, peer assessment, and self- assessment. With the AI assessment, we have innovated heavily and use machine grading to grade essays. We also have peer grading where the students grade each other’s work. And self-assessment where students grade themselves. We are grading free form essays routinely using AI assessment that I would have thought would have taken a few years to develop, but we are there. It is a really huge innovation.
The edX team has also come up with a way to replicate small group discussions. They have invented a course with a discussion forum on our platforms where students can ask questions and get answers on the discussion forum. Cohorts are introduced and small discussion groups are created where the professor can divide up a large class into smaller groups, and students can have a much more intimate discussion in a smaller group setting. In fact, the professor can also create discussions groups based on students’ answers to questions. So this cohort concept is a very handy concept and something surprising.
PH: Your company is very data driven, as you have noted. What metrics do you monitor to gauge the success of edX?
AA: We keep a lot data on the success rates of our students in passing courses, and the how long they stick with the courses. It is important to emphasize, though, that we are two years in, and therefore we are young company. The general evolution of the MOOCs runs from Sal Khan’s establishment of Khan Academy to Sebastian Thrun’s founding Udacity, to Stanford’s broader work on online courses and us. All of that has happened in a two to three years’ time. If you compare that to search, the evolution from Alta Vista to Google as we know it now has been 20 years. Who knows where we will be in 20 years.
Education is far too important to make drastic changes at the first sign of difficulty. We’ve faced a lot of challenges and have overcome them in the past and we’ll continue to strive to do better, but we are in this for the long-run. For instance, we tried offering job placements and that didn’t work, as we quickly realized we couldn’t be an education platform and a job placement platform, so we unbundled it, but we need to keep experimenting.
PH: What experiment are you currently working on?
AA: We are quite excited about open source. We elected to make our platform open source a few months ago, and that has hastened the pace of establishing partnerships as others leverage our technology. It led to collaboration as different as ones we’ve established with Stanford University to ones we now have with countries like France and China.
PH: How do you pull together the kind of team at edX that can both operationalize the ideas as well as be the content experts?
AA: We focus on non-profit principles, but with respect to delivery, it is run like a startup. We have pulled our team together both operationally and strategically with startup experience. For example, Rob Rubin is the Head of Engineering and he came from Carbonite and he has a sequence of startups under his belt. Matt Malloy, who is our VP of Marketing, used to be Head of Marketing at Zipcar and he has done a number of startups himself. We also have academics so our effort combines both academics and content. We also have to build a cutting edge platform, so we have some great technologists. Therefore, it is a great combination and mix of university leaders, startup experience entrepreneurs, and operational leaders. I think a team like that is really what you need to pull off an academic and a social entrepreneurship startup like edX.
Peter High is President of Metis Strategy, a business and IT advisory firm. He is also the author of World Class IT: Why Businesses Succeed When IT Triumphs, and the moderator of the Forum on World Class IT podcast series. Follow him on Twitter @WorldClassIT.