We met with Dr. Zvi Galil to discuss the OMSCS program at Georgia Tech and the ways in which this program is influencing higher ed around the world. Enjoy!
Influential computer scientist Dr. Zvi Galil discusses the “earthquake” created by his development of the massive open online course (MOOC) master’s degree computer science program at Georgia Tech—and for a cost of just $6,600 total. He talks about how the OMSCS program is the largest MS in CS in the world, and it is influencing worldwide higher ed experimentation with similar programs. Former John P. Imlay, Jr. Dean of Computing, Storey Chair, and professor at the Georgia Institute of Technology, Dr. Galil talks with Dr. Jed Macosko, academic director of AcademicInfluence.com and professor of physics at Wake Forest University.
But it was against our principle of accessibility. Our motto is accessibility through affordability and technology. And I cannot one day say, "No, no, no, we will accept only…" [chuckle] Because everything in higher education is... You get your reputation by denying students.” – Dr. Zvi Galil
For more on Dr. Galil and the Georgia Tech OMSCS program, click here.
See Dr. Galil’s Academic Influence profile
See additional leaders in computer science in our article
Top Influential Computer Scientists Today
Considering a degree in computer science? Visit Our Computer Science Page, where you’ll find the best computer science colleges and universities, career information, interviews with top computer scientists, influential scholars in the field of computer science, great books, a history of the discipline, online computer science degrees, and more.
(Editor’s Note: The following transcript has been lightly edited to improve clarity.)
Jed Macosko: Hi, this is Dr. Jed Macosko at Wake Forest University and AcademicInfluence. And today, coming to us from Georgia Tech is Professor Zvi Galil, and he is a professor of computer science and has taught a huge class that has made computer science available at the master’s level to so many people. So I’m gonna find out a little bit about that today.
But first, tell us where you’re from originally. We can hear that you have an accent, and how did you end up at Georgia Tech? What was your trajectory?
Zvi Galil: It’s a very long story, but I’ll try to be succinct. I was born in Israel, actually in Mandatory Palestine, that preceded Israel, and I went to Tel Aviv University, BA and MA, and went to Cornell to do a PhD, and a postdoc at IBM, and then I went back to Tel Aviv University and was on the faculty. Then I was at Columbia University for 25 years as a professor, then as Chairman, and then Dean of Engineering. And then I went back to Tel Aviv University to be the President of Tel Aviv University, and I was President for a little over two years and stepped down. That’s a long story, not for today, and then Georgia Tech called, then I became... They asked me to be a candidate to the final list for the dean position at Georgia Tech, and I was Dean of Engineering for nine years. I stepped down in June of 2019, and went back to the faculty after two terms.
Jed: Wow. Well, that was very succinct, and we can kind of imagine you going between the two countries, United States and Israel, and building up your expertise in being an administrator and working with people.
But how did that lead to this giant master’s level computer science class that you’re now teaching?
Zvi: So I always was concerned about affordability and accessibility of higher education, and it all started in actually 2012. Sebastian Thrun , who was the founder of Udacity, visited Georgia Tech, and 2012 was called by The New York Times, “The Year of the MOOC.” That’s when the MOOC arrived, massive open online courses, and the first one was actually by Thrun and somebody else at Google, artificial intelligence, but he created a startup called Udacity, that provided a platform for MOOCs. Same time, also the Coursera, another one, started later, edX joined them. So these are the three first ones, and he suggested to consider starting a master’s degree that will be based on MOOCs. And that’s the idea immediately.
The MOOCs, the trouble with MOOCs was that they gave no credit. So it was great courses, a very large courses, very high attrition, only single-digit survival rate ’cause there were no credentials, very determined people that wanted to learn could, but 7% finished the MOOC; 100,000 started, 7000 finished.
So the idea is... But students, they want degrees. They want... And that also incentivized them to study, because if you take a MOOC and you have the smallest difficulty, the smallest challenge in life, in work, or in the family, or at home, you drop out. That’s what happened. So I liked the idea, and then there was a process. I was only the Dean. Dean, you’re in the university, you know, the dean cannot tell the faculty, “You do X.” They will laugh him or her out of the room.
And the most important part for everybody, for them, for the task force, for the faculty, and for me, was quality, to have a top quality program, actually the same one that is on campus, not at all different.” – Dr. Zvi Galil
I organized a task force to learn it, and actually I didn’t interfere with them. I told them, “You don’t want to do it, we won’t do it.” I wanted the faculty to want to do it, but I didn’t interfere. My Senior Associate Dean, who is now the Dean, Charles Isbell, we didn’t intervene. They led it for six months, they had town halls, Sebastian came, I came when they invited me, but we did not interfere, and after six months, they voted 75%, they decided to do it, plus they developed a plan. And the most important part for everybody, for them, for the task force, for the faculty, and for me, was quality, to have a top quality program, actually the same one that is on campus, not at all different. Same requirements, the same difficulty, no discounts. And they did it.
And then there were another two months where we needed approvals, Board of Regents of the University of Georgia. I know they’re bureaucratic, and then they approved... He visited in September 2012. The Board of Regents approved it in May of 2013. We embarked on creating the first five courses, and already announced that in January 2014, we would start. And then we started with five courses and 380 students. That was the start. Since then, it grew immensely. Actually, we don’t... The sky’s the limit. We don’t see it, how large it will become. I can tell you a little bit more if you want to know.
Jed: Oh, absolutely, I wanna know more. So first of all, clarification. You said no discount...
Zvi: No, no, no.
Jed: So does that mean...
Zvi: No discount in the academic requirements, or residents.
Zvi: Actually, there are four distinguishing features for our program. The first is that it’s MOOC-based. Initially, for the first program, it was based on MOOCs. Now they follow us, so it’s not unique anymore. The second that caused an earthquake was the price. The cost for the whole program was less than, actually, $6,600 the whole program, and you pay by course, 10 courses. So, you pay by course. So it was highly affordable, that caused an earthquake. That was the thing that attracted everybody’s attention, so these are the first two features that distinguish it. So it’s highly affordable. That’s why, that’s big part of its attractiveness, but the quality also.
Jed: So yeah, the quality is just the same. No discount on quality. So the affordability... Now, can you give us an idea of what it would cost a person before 2013, 2014, when they started to go to Georgia Tech to get a master’s in computer science?
Zvi: It depends if you are resident or non-resident of the state of Georgia.
Jed: If you’re a resident of Georgia...
Zvi: If you’re resident it costs like $20,000, if you’re non-resident costs $42,000.
Jed: Okay. So $20,000 to get the entire master’s degree. And now you only pay $6000 or $7000.
Zvi: Yeah, it’s now 7000 or 200...
Jed: Okay. So, it’s a third of the price...
Zvi: It’s... Of the 70,000.
Jed: Okay. Wow, that’s amazing. So you said it’s over 70,000 if...
Zvi: In a private university.
Jed: Oh, yeah. In a private university, it is a lot. Yeah, I did know that.
Zvi: They cost, out-of-state, something like $40,000, $42,000.
Jed: Yeah, but for an in-state person, you’re still saving 2/3 of the cost, and for an out-of-state person you’re saving 5/6 of the cost, roughly. So that is why you created an earthquake.
So what has happened to the normal program at Georgia Tech?
Zvi: Well, surprise surprise, it didn’t cannibalize the normal program.
Jed: That’s amazing.
Zvi: Actually in the first five years, I don’t have the up-to-date numbers, applications more than doubled.
Zvi: The explanation is that the online program is more than half residential within domestic students. The on-campus students, the face-to-face classes, that’s international, because they don’t want only to get the degree, they want to get into the States, which online program doesn’t enable that.
Jed: That is so unexpected. I would have predicted the exact opposite. So you’re saying that half of the people who take the online...
Zvi: No, not half, 64%.
Jed: 64% of the people who take the online class...
Zvi: Right now. When we started, it was 87%.
Jed: But right now, 64% of the people who take your online class to get their master’s in computer science are from the United States. And when people come to Georgia Tech to get a computer science master’s, they are more than half from overseas, is that what you’re saying?
Zvi: Yes, roughly...
Zvi: Roughly 55%. Right now, in our master’s degree, on-campus they are 55%...
Jed: That is so surprising. I would have guessed it, it would be the opposite way around. So...
Zvi: Also, it was not... A little bit unanticipated.
Jed: Yeah. So, because of that, you used that as an explanation for why your online program didn’t cannibalize your residential program. And your reason is that the people who are coming to the residential program, at least 55% of them, want that excuse to come to the United States on a student visa...
Zvi: It’s a foot in the door.
Jed: Foot in the door, okay. So that makes a lot of sense.
Zvi: It’s several reasons, but I believe this is the main one.
Jed: Interesting. Now, did you hear about other programs in the United States losing people because yours is so cheap and you get a degree that’s just as good and just as high quality?
Zvi: No, but we don’t know, but frankly, computer science is the hottest field. It’s the hottest field.
Jed: And you’re providing a way...
Zvi: That explains all this increase, because the increase of the size of our... Our program is called OMSCS, Online MS degree in Computer Science. OMSCS, that’s the name of the program. OMSCS has grown every semester. This month, it’s seven-year old. Every year it grows, every semester it grows. And actually, in the fall, we passed 10,000 students. In the spring, we passed 11,000. And we now have the applications, and also number of applications is growing up every semester, except fall and spring is different. The application for the fall is usually higher, because that’s the normal starting point. So if you compare fall to fall, and spring to spring, every fall it went up, and every applications for the spring it went up. Right now, applications for next fall, not finished, but at the same level of last year which was the record, we are 10% higher. So we may reach 15,000, maybe even higher, I cannot predict, but it’s still growing, it has grown every semester.
Jed: Wow, that is so incredible. So just to put it in perspective, what is the largest computer science master’s program in the world that’s residential?
Zvi: We don’t know, we don’t know. Our program...
Jed: What was your program?
Zvi: Somebody say to me... But I think it’s only a conjecture, I don’t think anybody went and checked it that our program is the largest master program by a top-ranked university in the world in any area, online or not online.
Zvi: But it’s one of the largest. I don’t know. I’m not making any claim, because I didn’t put the time to verify that claim, and I’m not sure he knows what he’s talking about.
Jed: Okay, well, either way, 11,000 students that are currently enrolled. That’s... Is it a one-year or two-year program?
Zvi: No, no, no, that’s... Everybody that is now enrolled, in different stages, and they finish in different periods of time.
Jed: Right. But how long is it set up for if you wanted to do it kinda fast? What would be the general...
Zvi: On campus, if you do it really, really fast, it’s one year. I think in OMSCS, there were few, very few that finish in two. Usually, they finish in three, because they take one course usually. Georgia Tech, one of our brands is difficult, really difficult, and our students usually have families, they have jobs, so taking two courses is quite something. So actually, the average is 1.3 courses, so a big majority take only one course.
Jed: And they have to take 10?
Zvi: A not insignificant minority takes two, and very, very few takes three.
Jed: Okay, so you’re saying that they have to take 10 courses to graduate, and you can take courses in the fall or spring.
Zvi: Also summer.
Jed: Also summer, okay.
Zvi: During the summer, we don’t have all the options, not all the courses are offered. Neither summer, which includes many of the courses. Which, by the way, we started with five courses. We now have over 50.
Jed: Wow! But you just have to take 10 to graduate?
Zvi: According to some specifications. So we have specializations.
Jed: Yes, of course, of course.
Zvi: It’s like [inaudible].
Jed: Yes, absolutely. I can imagine. So basically, what you’re saying is that to get done in a year, you’d have to take three...
Zvi: No, no...
Jed: Every single period, three classes...
Zvi: Most students, in a year, they usually take four, four, and two in the summer, or five and five. But online, nobody can do it. Nobody does it.
Jed: Nobody can do it, nobody does it. Okay. Very interesting though. So when you say that there’s 11,000 people currently enrolled, you can kind of think about it as having about 3000 or 4000 people per year. Is that how many people come in per year?
Zvi: It’s the number of students that started in the fall term. I bumped into it yesterday so I can tell you. It’s a little bit of over 2900, so it’s almost 3000.
Jed: Almost 3000, and then of course you get some in the spring time. You get some in the spring time too, right?
Zvi: Yes, yes, but it’s a smaller... Usually, the spring, it’s not the natural start.
Jed: Right. It’s a little bit smaller.
Zvi: Compared to the fall, it’s always smaller.
Jed: Okay, but you get 3000 in the fall and some more in the spring, and that’s what builds up the 11,000 that you have now.
Zvi: Yeah, but you have also people graduating. So far, we graduated almost 5000 students. And by the way, the 11,000 are more than 25% of the students of Georgia Tech.
Jed: Of all the students at Georgia Tech? Wow! That’s a pretty impressive number. So, now looking back on all the things that you just said, you’ve had a huge career. You’ve been at Columbia, you’ve been president at a university.
Is creating this program one of the things that you’re most proud of, and the thing that really makes your career feel like it was really a great career?
Not one of the things. It was the thing. This is the biggest thing I've ever done.” – Dr. Zvi Galil
Zvi: Not one of the things. It was the thing. This is the biggest thing I’ve ever done.
Jed: Wow! That is so cool.
Zvi: You can look me up. I had a pretty good career as a scientist, but this impact is the biggest thing that I never dreamed of.
Jed: And it could change the way everything is done, but obviously, computer science is probably the easiest subject to teach online.
Never say never, never say, "This area we cannot do it." Try, and try to solve it, and technology improves all the time, so maybe even you cannot do it now, next year you may be able to do it with some additional technology.” – Dr. Zvi Galil
Zvi: Yeah, it’s the most natural, partly because the technology comes from there, partly because in some sense, immediate interaction is somewhat less important. Though many areas, and I recommend places to try all areas. Never say never, never say, "This area we cannot do it." Try, and try to solve it, and technology improves all the time, so maybe even you cannot do it now, next year you may be able to do it with some additional technology.
Jed: That’s exactly what I think. So I teach physics, and of course, during the pandemic, we couldn’t have laboratory classes, so we found a little special device that we could mail to each student, and it had all the ways to test the laws of gravity...
Zvi: And when I give lecture, I am a big supporter of virtual labs, because you cannot be poisoned, and nothing can explode.
Jed: That’s a good point, that’s a good point. Well, is there anything else that you want to talk about that might help, let’s say a younger person follow in your footsteps to do something that is as impactful as creating this program at Georgia Tech? Are there any things that you think helped you be successful?
To my mind, it was like you gave the professors at Georgia Tech autonomy. You didn’t interfere. You said you guys had a task force.
Zvi: It’s their decision and you cannot tell them what to do, your faculty.
Jed: Yeah, I know, but that was probably one of the key things that you could pass on to somebody else, say, “Hey, if you’re gonna try to start something big, when you make a task force, let them make the decisions.”
You know II Rabi, Nobel laureate in physics, said, "Mr. President, I apologize, but I have to correct you, we the faculty are not employees of the university, we are the university." And I believe strongly in it. Your faculty need to want to do it, of course…” – Dr. Zvi Galil
Zvi: It’s kind of bottom up, so even if it’s your idea, it has to be their idea. You have to be, to a little bit backtrack, you have to go back and let them. And they wanted to do it. 75% voted to do it, because they were excited about the innovation, about being the first. And I cannot... I’m totally indebted to them. They always joke about Eisenhower coming to Columbia and telling the faculty, when he was president “You, the faculty, are the employees of the university.” So II Rabi... You know II Rabi, Nobel laureate in physics, said, “Mr. President, I apologize, but I have to correct you, we the faculty are not employees of the university, we are the university.” [chuckle] And I believe strongly in it. Your faculty need to want to do it, of course, if...
Zvi: Academic leader sometimes know how to direct, to suggest, then sometimes backtrack. I told them, “You don’t want to do it, you won’t do it. I will not force you. I will not corner you. But it will be your decision.”
Jed: And these people on the task force, were they all from computer science?
Jed: Were they all faculty in computer science? And did they stop teaching the regular classes and...
Zvi: No, no, no. Everybody that teaches in OMSCS, it’s extra work, and extra pay, not big, but not insignificant.
Jed: Very cool. Interesting.
Zvi: They have some additional income, actually, which has three pieces. But I won’t get... If you want, I could give you it.
Jed: Yeah, I think we’ll leave that off in case people feel a little weird, you talking about their salary. But I feel like that’s an important piece of why it worked. Because people who are university professors are often, in my opinion, as a university professor, a little bit underpaid, compared to our counterparts in the industry.
Zvi: Yeah, they won’t double their salary, but they get... In business schools, usually when they teach executive MBAs, they’re very richly paid. But it’s a decent addition. It was the way, because I also am a strong believer in incentives. People will only do voluntarily once or twice, or maybe even, if they are non-tenured, three times. But you cannot count, and you cannot sustain, a big effort without compensation.
Jed: And with 11,000 students paying on average, $600 per course. Right? Taking 1.3 courses per semester, that’s enough money to give the incentives.
Zvi: No, no we now do the... People now get... There is some sharing between the various units that are involved. I don’t know the exact numbers, like $6 million every year. So it brings... That’s after the net. So, after all the expenses [chuckle] including [inaudible]. So with this large number, even though the tuition is small, you multiply it by a large number, it’s still pretty good.
Jed: Yeah, that’s great. And it didn’t happen overnight. Georgia Tech had to invest in it to be able...
Zvi: We were lucky. It doesn’t hurt to have luck. So the initial cooperation, collaboration was with Udacity, with Sebastian Thrun, that The New York Times says a great physicist, called Jim Gates. I’m not sure you know him. He’s a presidential medalist. He was quoted saying that we are the Wright brothers, and I said, “No, no, no, we are the wrong brothers,” not because we invented the airplane, but because we will actually show whether the MOOCs can fly, and we did. Sebastian Thrun collaborate a brother. We call one another brothers, but then we also went and got AT&T into the picture. And we went to AT&T, actually directly to the CEO of AT&T at the time, Randall Stephenson, together with number 2, Bill Blase. And we went to him on a Friday, January 11th, before even the faculty voted. On Monday, he gave us $2 million. He liked the idea. And he later gave us another $2 million. So they were crucial.
Zvi: They were crucial. Georgia Tech is a public university. Its endowment is about 5% the endowment of Harvard. If they had to invest $2 million, I’m not sure they could do it, we could do it. I didn’t realize it at the time, but that was crucial. To create a course, cost us $300,000, each course. Now we are better, it’s about $120,000. But, it’s cost. We needed this $2 million.
In the first two years or so, or three, because of this twice $2 million, we were in the black and we didn’t have to invest. We were always in the black because this wonderful support by AT&T, and people ask me, “What did they get for it?” They didn’t get anything, but they sent many of their employees, and they didn’t get any special treatment. We dealt with them as any candidate applying. They had to have the requirements. The fact that they were from AT&T didn’t help them.
Jed: Interesting, so wonderful and I was gonna say, the fact that Georgia Tech became the first to offer this, and not Harvard with its huge endowment, is really because of one thing, Zvi Galil, right? Galil, you are the man. You’re the guy who had the vision. You said...
Zvi: I have a big part in it, but it’s the faculty.
Jed: It’s the faculty, but you helped them.
Zvi: And my Senior Associate Dean, Charles Isbell, who is now the Dean. There were two some other important people, so I won’t mention all of them. But Charles also... And Charles is amazing. He has two courses in the program, and still as Dean teaches them. I as Dean didn’t teach any course. He teaches every semester, I’m not sure summer, but for the spring he teaches two courses. Of course, it’s easier because the classes are recorded. But he’s still instructor off record. And he did build the courses together with somebody from Brown. That what you can do, you can have faculty from all sorts of universities. The guy from Brown, a great guy, Michael Lichtman, long gone. Charles is teaching every semester two courses, and the students love him.
Jed: So how does Brown feel about taking away one of their...
Zvi: No, no, that was an extra... Faculty can consult. A university does not own the free time. So I don’t think... I’m not sure what we did with Brown, maybe asked their permission. I don’t know. But that was way back when they... Like five, seven years ago, when they developed the course.
Jed: Interesting. That is so fantastic.
Looking at the pandemic, do you think that creating these MOOCs in 2014 prepared Georgia Tech to be better than other universities in teaching computer science, when the pandemic hit?
Zvi: Of course.
Jed: Okay, tell us about that.
Zvi: It was kind of prescient. The pandemic is of course horrible, nobody wanted it and it’s awful results, but there are some silver lining. So, when the... Now everybody moved to online, so did Georgia Tech. But we had 50-something courses ready, ready to go. We use them also... By the way, the tapes are used also by their own campus. They have access. So people will take it on-campus and go to classes, or don’t go to classes, they can use the tapes also. Usually, the professor use them. It’s another resource for them. But we were ready to go. [chuckle]
And many universities were forced to move to online with no experience whatsoever. And it was quite rocky. People suffered, but that was it. We had this 50-something courses in excellent shape. And now the result of the pandemic, some, not all, some that moved to online will use much more online, not necessarily in full, but maybe in part because you can combine. You can have a hybrid and change the mode of your teaching, and probably that’s a very likely case. Because to have a talking head, talk to them and everybody writing. If you have it on tape, it’s much more efficient.
Jed: So what you’re saying is that the people who normally taught in-person, who had to teach online because of the pandemic, didn’t just do a talking head, they let the students watch the taped version?
Zvi: No, it’s complicated, I don’t get into all the details, but they... We have some modes of interaction. It’s MOOC-based, but it’s much more advanced than the old online. Old online, where somebody’s just recorded in class, they’re still online and some of them charges a full price. Though there are 40 programs that followed us in 30 universities. The first one was Illinois, who calls us the big brother. So they have now three different programs, and we have three. We have two others in Georgia Tech, and they have three. They were the first one, but they came and learned everything from us, and copied essentially everything but admitted. The others even don’t know that they copied from us. But there are now 30 universities that have a much more reduced price MOOC-based master degree.
Jed: And it’s interesting that the first two were both public universities. Again, I’m coming back to this question about why not Harvard? Why Georgia Tech? Now you’re telling me the second one was University of Illinois, they didn’t really have you to get them started with the idea. They came and watched you, but they came and watched you because they must have already had the idea. So what was it that gave them the idea? Did they have somebody like you that was interested or...
Zvi: So they already saw us, and they liked the idea. The guy there, Rajagopal Echambadi , who is now Dean of Business in Northeastern, but he came here and he called us big brother and we called him little brother. Could be sisters or no gender, please.
Jed: But when he saw your courses, and he said, “Ah, I want to go down and meet this guy and see how they do this,” you indirectly convinced him by showing him courses, but he had a fertile mind that was prepared, and what was...
Zvi: Of course he did... Like iPhone, lowercase I, MBA capital. Like iPhone, they have MBA program.
Jed: So they were the first MBA program? They copied your computer science program to do an MBA program. Is that what you’re saying?
Zvi: Indeed, indeed, and they are quite successful. They have a small number... Thousands of number of students, nothing compared to 11,000. Maybe 3000, I don’t know exactly. And maybe they also... We accept everybody that we’d never still capped it. I had some pressures, I won’t name names, to cap the site, to stop at 4000. But it was against our principle of accessibility. Our motto is accessibility through affordability and technology. And I cannot, one day, say, “No, no, no. We will accept only... ” Because everything in higher education is... You get your reputation by denying students, by... You know everybody... Stanford accepts 5.5% and they...
Jed: Oh, it’s below 5% now.
Zvi: Different number weekly, I didn’t follow. It may be five, maybe two years ago it was 5.5. And they put it on everything, so as they denied people, they feel better, which is kind of upside down world. My goal was accessibility, and the funny thing that happens is, our online program... That’s the second. The first one was MOOC-based, the second one, the price, which caused an earthquake, the third one is the admission. So we accept everybody till we believe we can do it, and it’s about 70%. The on-campus is 10%. And guess what? The people in the online program sometimes preform better. So they are not inferior. [chuckle]
So the selectivity is, of course, on-campus, you have to be selective because you don’t have the facilities, you don’t have the classrooms, and you don’t have the other goodies that you give all the students, gym and everything else, and eating places and dorms. So I’m not criticizing, but in terms of selectivity, there are much, much more talented students that are rejected just because you don’t have room. So the 70% comparably perform sometimes a little bit better, sometimes a little bit worse, because we didn’t lower the admission requirements. The role of the admission requirement is to accept everybody that can do it.
Jed: Wait, wait, wait. Let’s repeat these numbers: 70% of the people who apply to your online program get in, okay? And the percentage of people who apply and get into the residential, is that a...
Zvi: 10%... 10.5.
Jed: 10%, 10.5. Okay, so you’re saying that the bar for admission is equally high in both cases? So...
Zvi: No, no, no. In the on-campus, you are stricter.
Jed: That’s what I thought. Okay, what were you just saying then about, “We didn’t lower the standards for... ”
Zvi: The number dictates how many you can allow in. So we probably could have accepted 50% for the on-campus, but we can not.
Jed: So you have to raise the bar, but it’s kind of, you’re saying it’s artificial. So if you took the bar at what it should be, it would be equal to the online, but because of space limitations, you have to artificially raise the bar for the residential program.
Zvi: And I’m sure the students that apply to Harvard... They accept in the single digit. 30%, 40% that are rejected are excellent students.
Jed: Absolutely, yeah. I know from personal experience.
Zvi: You know, don’t get down, don’t be down. And the process is quite random, you help some old lady cross the street, you get two points, and they admit you because of it. So...
Jed: We could go in that direction, but we did a whole interview on a guy who wrote the book Who Gets In And Why. And he spent three years embedded in three different schools, including Emory University, right next to you, to see how they do the admissions process, and he agrees with you that there’s a lot of randomness to it.
Zvi: Of course, we need to because of the space limitation. And we need the...
Zvi: But then there are all these things that are not exactly fair, like the legacy, and the donors, or the potential donors. But Georgia Tech is pretty good. I’m not talking about any other university, Georgia Tech is pretty good. They’re a public university. The process is fair.
Jed: Yes. It’s a weird process that we had a whole interview about it, so if you’re listening to this interview and you wanna know more about that, you can check out that interview. But I have just a few more thoughts for you I wanna discuss because you seem to have this great idea that you can get a Brown professor and a Georgia Tech professor to teach the perfect computer science class in the perfect way, and your model allows for that kind of flexibility.
What will education look like in the future?
Could we imagine a situation where a student maybe from international, wants the experience of coming to United States, wants to be surrounded by people their own age, on a nice campus or a nice place, so instead of doing this at the schools that already exist, Georgia Tech, Harvard, University of Illinois, instead, some entrepreneur comes and just rents out a resort and says, “Let’s all go here, and the professors are gonna be great professors from Brown University, Georgia Tech, Harvard, anywhere else, and we’ll just have it right here.” Do you think that that might be a future of education?
Zvi: It’s a little bit too optimistic, but it can happen. Also in our case, this is essentially the only case where we have two... Actually, the two professors co-teach this class, I told you, the one who is now Dean, and there’s a guy from Brown, and they kind of play good cop and bad cop. He asks a question. He answers the question. Sometimes they sing, it’s a beautiful... If you watch my talks, I have a clip, a video clip from the class, because they are good teachers and also quite entertaining.
Jed: That sounds fantastic, but why couldn’t that happen? Why couldn’t...
Zvi: Potentially, I have a friend. He taught online, and also online at Columbia, 40 years after he got his PhD at Columbia. And he basically warned, already 30 years ago, that maybe one day when we did online, and we did online at Columbia when I was Dean. We did it in the normal way of, that I told you, and not in sort of way and charge the full tuition. And he said that one day, actually, we will need much fewer teachers because we can get the best teacher in the world for every course.
Jed: Yes, but you also need that personal attention. So maybe you can have the best teacher, but then also, you get somebody there to help you work through it.
Zvi: These are the problems we are trying all the time to enhance, the engagement, because online, so... But...
Jed: Well, in the last few minutes that we have, can you just tell us, since none of the people watching this show have probably been through your course…
…what makes it better than the old-fashioned taped courses that’s just me talking? What makes it more interactive? What have you done? It cost you $300,000 to make each course. What did you do to make it so interactive?
Zvi: We have Piazza, Piazza is a system where students can interact. Now, in our case, and that’s the third... I told you the there are three distinguishing: MOOC-based, price, admission. And there are some other details in the admission that are somewhat different, I won’t get into them, but then the fourth thing that’s unique to our program: The use of social media. So there are 70, maybe now more, of student groups. And they interact on social media, give advice, say... Some of them by course, some of them by gender, some of them by place. The one from Shanghai, they have... Some of them by gender, the women group. I’m not sure we have minority groups, there might be. But there are 70-plus, and I think all of them... Most of them are still active.
And they interact, they help have one another, they even give advice to potential students. And we developed kind of the community, which is part of the amazing part. And many of our OMSCS students TA, and they get peanuts. They TA, and some of them TA after they graduate, some dozens of them. We need a huge number of TAs. We have a TA for every 50 students, on-campus for every 30 or 25. But for every 50 students, you need to grade, and this, we cannot also make the grading, and we don’t give only four choices questions in exams or homeworks. Sometimes projects, you need to grade things. They volunteer for little pay because somehow they love the program and give back. And actually without them, we wouldn’t be able to do it.
Jed: That’s fabulous.
Zvi: So almost all our on-campus students are TAing here, the master’s students. But we need... If you do the computation, 11,000 divided by 50 is... I don’t know. It’s more than that, because everybody takes 1.3 courses. So it’s 14,000, or 15,000 divided by 50 is 280, so you need 300 TAs.
Jed: That is amazing. Well, it is incredible. I guess we’ll just have to go and watch the taped videos. Are they shorter than the old-fashioned ones? Are they done in little bits with tasks in it?
Zvi: No, no, no... Yes and no. The MOOCs... One of the properties of the MOOCs which distinguishes them from the old online, you asked about the differences, is that the class is the same timeframe, maybe 75 minutes, but it’s broken to five, six pieces. And every piece has a quiz, so that will make sure that the student understood the concepts. So it has breaks and quizzes, but it’s usually... We put it... The class is in the same semester. So initially, if Sebastian always has had the wild ideas like the self-driving cars, I think he’s the first one to sell the idea, said, “Oh, they can have any time. We don’t have... Leave by semester,” but the people in our administration almost got a cardiac arrest.
You need to handle them, different semester, different times, somebody comes a few years later, we’re not gonna say we cannot do it. So we did it by the semester. So the things inside is flexible, but also not entirely flexible, because there are deadlines for homeworks, deadlines for exams, but in between it’s totally flexible. And plus, they can watch their class from the beginning to the end if they want because they get all the tapes right away.
Jed: Yeah. That’s great. So these 75-minute lectures are broken into five or six pieces, 10 minutes each, and there’s a quiz after it, and that’s what makes it a little bit more interactive than the old-fashioned taped lectures?
Zvi: That’s in the class itself. But then they have office hours. They can meet the TAs, they can meet the professor, and they interact between themselves. And actually, we’re also fortunate right now, there was a superb Executive Director that is not now anymore because he did other things, now he does only the top things, his name is David White, but David Joyner there is now Executive Director and his research is in computational technology. So he’s experimenting in all sorts of things. So first of all, he introduced a lab so people can do research. Because usually, our master’s students have one option to do research, initially we didn’t do it, very hard to scale, but now we do some, we got much more. Before research or thesis or a big project, it was only available to on-campus students because very they’re hard to scale, but now the research is... So we allowed in a few cases, people, not... Unofficially, but now there are much more possibility. There I’d say one...
Jed: That’s great.
Zvi: Of the courses now experimenting with synchronous, so they can have... But it’s hard with a thousand students in the class. So we’re trying with 200 in the class. So one of the meetings every week for this one course is synchronous, so they can interact like we’re interacting now.
Jed: And is there one professor of that course?
Zvi: Yes, but there’s a bunch of TAs.
Jed: And what... And are the TAs there during that synchronous time to meet with people in smaller groups and then they all come back together?
Zvi: Some of the TAs, they make it to small groups. Yeah, yeah.
Jed: Okay, very interesting.
Zvi: It’s now... It’s not part and parcel yet. It’s an experiment.
Jed: It’s an experiment with 200 students or so.
Zvi: You do something new, you experiment, and learn what you can or cannot do, and which you cannot, never say never.
Jed: So this has just been so much fun for you to do and fun for me to hear how it all went. I can only imagine that you’re just gonna keep working on this until you can’t work on it anymore. This is it, right? You have no plans except for... To keep this going, is that true?
Zvi: I was actually involved in the... My first three years as a Dean, I’m ready to serve on the faculty and I teach. But I’m also an Executive Advisor to online learning, so... But I’m still a kind of global ambassador. I gave 76, you’re number 76 or 7, talks about it in 16 countries. In March, I will give a webinar and I will give a keynote in ICISS, Information School Conference about it. So, still there is big interest seven years later, and I’m an ambassador and I’m very involved and I’ll do everything, but the leadership in the college have to want me to do it. But they do. [chuckle]
Jed: Well, at Georgia Tech. At Georgia Tech they do, but not everywhere else. So hopefully, you can get the message out to other universities so that they can become the little brothers.
Zvi: It hasn’t happened yet, I won’t mention the name of one university, almost happened. Because there are universities that cannot offer a master’s degree in computer science. They’re kind of limited, but they could use others, ours. Plus, maybe somebody there to manage it, so people will be on campus but they will use our tapes, our problems. As we will do with the students, we will not charge them a high price. So, we want it to be yours.
Jed: That’s great, wow. That would really get the vision you had to have other schools using your tapes, your program, and spread it to more than just the 15,000 students that you’re projecting will be in your program, get it out to the other programs around the world, and it could be in the hundreds of thousands.
Zvi: And it could, I don’t know, because faculty are funny people. And NIMBY, and not in my backyard, or... And then with that university, actually, the Computer Science Department wanted it. It’s a very, very big... Maybe the biggest university in the US, I won’t mention names. They were very interested, but then went to the Senate and they said, “No, no, no. It doesn’t reflect well on us.”
Jed: Yeah, okay. Well, I wish you the best in that endeavor, because the more universities that can use your stuff, the better, or if they can develop their own, that’s great too. But why reinvent the wheel if you guys have already done it for computer science. So anyway, I wish you all the best. I hope this interview continues to spread your message about this amazing project, and thank you for spending your time with us today.
Zvi: It was pleasure. As you see, I love to talk about it, I love to spread the word, I love people to know about it. And whoever invites me to talk about it, it doesn’t have... He or she do not have to ask me twice. So you saw that I talked about it with pleasure, and I appreciate your interest.
Jed: Well, thank you, Professor Galil, we really appreciate it.
Zvi: Thank you.