[00:00:00] Hello, this is Mark. And welcome back to the Learning Machine podcast. Today, the topic is MOOCs, or massively online open courses. They're like courses you would take in a school, except that since they're on the Internet, there's generally no barrier. Anyone can participate in them and there's no limit to the number of students that can be in one. And since they're online, there are a few other differences, such as obviously, professors don't have time to give as much individual feedback, if any. They tend to rely on forums where students can help each other and they tend to be focused on videos. So most MOOCs I've seen have videos front and center. They might be full lecture length or they might just be short,10 minute modules. They'll often have some readings that accompany the videos and then some MOOCs have quiz questions as well. And in the case of programming focused courses, there might be an automated grader that will check your program and see how it does against a variety of criteria. In some cases, you can just keep submitting it and improving it and submitting it again.
[00:01:10] I would say the two defining MOOC platforms that are what most people think of, if you say the word MOOC would be Coursera and edX. Coursera was started by a couple of Stanford profs. It's modeled pretty closely after actual university courses, or at least it was in the beginning.
[00:01:27] Classes had start dates, homework with due dates, tests, occasionally peer-graded homework assignments, and students who achieved a high enough grade would receive a statement of accomplishment—a PDF, generally signed by an instructor. Those certificates don't actually hold any value in terms of academic credit, but they do still serve as a motivator and, more recently, as a cash cow.
[00:01:52] Coursera very quickly grew in popularity and signed up instructors from universities all over the world. And nearly all of these instructors were world class, at least in terms of their own careers. When I took the course on the Scala, a programming language, it was taught by none other than Martin Odersky, the inventor of the Scala programming language. And, their machine learning courses were taught by the co-founder, Andrew Ng, who was one of the leading researchers in the world on the topic and so on, with many, many different Coursera classes. The quality of the instructors was and still is absolutely top notch.
[00:02:27] Not long after Coursera got traction, Harvard and M.I.T. opened a competing platform called edX. And all of the things that I've said so far about Coursera are also true of edX. Although the two platforms are a bit different. Coursera was launched as a for-profit venture, even though its courses were free to take without certificates, but Ed X is entirely a non-profit that survives off of donations and those same course certificates.
[00:02:55] From a user perspective, I haven't seen too many ways in which that makes a difference so far, but Coursera has been a bit more aggressive about its monetization recently, so that may be due to the difference in their models.
[00:03:08] On the whole, I find the emergence of these two platforms is extremely exciting. There are hundreds, if not thousands of free courses online taught by world class instructors. Sometimes the material in them is the same as what's in the university class. For example, I took a linear and integer programming course from the University of Colorado Boulder. One of my friends was taking it in person at the University of Colorado at Boulder, and the material was identical. And that's incredible when you consider how much people have to pay, especially in the US, to attend a university course in person. But there are some drawbacks. If you're a very motivated student and your goal is to replace a four year degree through Coursera and edX, it's extremely difficult to make a coherent curriculum to do that. That's for a couple reasons. There's a flood of lower level courses, and very few high level courses. There are some and there are some graduate level courses, but the vast majority of courses are aimed at first or second year students or remedial students.
[00:04:11] And to give you an idea, I would say the two subjects which have the best coverage would be computer science and mathematics. And yet, over 90 percent of the courses don't have any hard prerequisites. So, when you take a probability and statistics course, it generally won't depend on calculus. Same thing for a linear algebra course, which is not how it works at the vast majority of universities. In the lower level courses, including calculus, you can generally get a great replacement for a university course that will go through all the same things.
[00:04:47] But, the higher level courses are generally a watered down version. Or, what also happens fairly often is, there will be a full higher level course, but there won't be a suitable chain of prerequisites to get there. Part of this is due to courses coming from different schools all over the world and obviously they don't all teach the exact same things in the exact same courses. So, you kind of have to look for one institution that's put out a lot of courses and follow through the whole track. And as I've said previously, those don't tend to exist past the first and secondary level.
[00:05:27] And another drawback is that you're still generally tied to course schedules, for most courses. By that I mean, when you sign up, you won't have access to the whole course. You'll just have access the first week or two. And then every week, there will be required quizzes or assignments that you have to do. And if you don't get them done in time, then you'll be penalized points on your your score for the course... even though you're not actually in a course. You're doing it over the internet and the content was available the entire time. So as a result, a lot of courses I sign up for, I started off very impatient to get started and I usually tended to start during vacation or a time when I didn't have too much work and I would get as much done as I could. Then, there wasn't any more course content available, I would have to wait, and then later, after my vacation was over, I would be too busy to finish everything during a given week and I would miss assignments or quizzes and then eventually just get demotivated and quit the course, because it became impossible for me to pass it. And that's ridiculous! I know they want to have a lot of students going through the same content at the same time on the forums, but one of the biggest advantages to having content online is that it's always available. Imagine if YouTube said you can only watch this video at five thirty p.m. on Thursdays. It would be really annoying and probably a lot more people would watch TV instead of YouTube if they did that. Same thing for Netflix.
[00:06:58] The final drawback, which shouldn't be any surprise at all, is the certificates. They don't grant academic credit, employers don't care about them, but you do have to pay for them, and Coursera is increasingly locking down the platform for people don't pay. I tried a few classes last year and found that I couldn't participate in the automated quizzes, so I couldn't find out if I had actually learned the material or not unless I paid to get a certificate. Otherwise, I was auditing and auditors no longer get quizzes. So, that kind of sucks. edX doesn't have that problem as far as I know, though. So that's Coursera and edX.
[00:07:36] Another platform that could be considered a MOOC is the Khan Academy. That was started by single person, Salman Khan, who had a background in mathematics and finance, and he started just by helping his cousin over Skype. But he did a great job of it, and over time his videos went viral and then it became a big platform. And now there are a lot of courses on it. He taught a fair amount of higher level material, but at this point, the site is aimed primarily at K through 12 and it's got a lot more than just videos. In fact, it's got quizzes, but they're different from the quizzes at Coursera or edX. They use spaced repetition, which was a topic of a previous learning machine podcast, and all in all, I think it's incredibly efficient way for a kid to learn math or even an adult if you need to brush up. I think it's definitely worth checking out if that's your goal.
[00:08:32] And another platform that's gained a lot of steam in recent years is Udemy. It's not really like the other MOOCs since it's not based on traditional academic courses and it's much more open. There isn't a high bar, if you want to create a course and put it out there. So as a result, there's a lot of garbage. In fact, I would say that good content is rare. I have bought dozens of courses on Udemy and there wasn't a single one that I would call high quality, although there were a few that kind of got the job done. They tend to be aimed at the lowest common denominator, and that's because it is for-profit. It's dominated by Internet marketers and people who were basically just trying to make an income, which is... there's nothing wrong with that at all. But, the result is, the courses are made based on how much money can be made from making them, which means they're almost all aimed at beginners. And there are a lot of people that take shortcuts and there are a lot of aggressive marketers that promise a lot and make very lackluster courses.
[00:09:40] My advice on Udemy is just stay away. It's not the place that I find to be productive use of time even. Even if the courses were free, I think I would still just go to YouTube courses over Udemy courses.
[00:09:55] Then finally, there's MIT's Open Courseware, also abbreviated as OCW. It's the oldest of them all goes back at least the early 2000s, and it stands out from the others in a number of ways. First of all, they just recorded the whole lectures from every lecture for an entire semester. So if you take an MIT Open Courseware course, you'll see the exact same lecture that students say in 2011 or 2005, or whichever year's course you're looking at, saw. And then you'll do the same homeworks that they did and you'll take the same test that they did. There are no automated graders and there's no course schedule. It's just totally free. Anyone can drop in, watch whatever they want. The videos are. I think they're all on YouTube. The homework assignments, all of that is just there for the taking, whatever you want. So, if you have a hard time motivating yourself, you won't do anything with OCW. But if you're really motivated, you could put yourself through the exact same trials that actual students did at one of the best universities in the world. And it's not limited to math either. They have many courses from many different departments and in general, the coverage is great. Like you can go through an entire bachelors degree or in some cases an entire master's degree's worth of material. So it's definitely worth checking out. May or may not, well, it's definitely not the easiest to use, which is why it never became as popular as the others.
[00:11:25] So looking at all of the MOOC options, here's the bottom line:
If you need to review your basic math up to a college level, then use the Khan Academy. It's what they do best, they have a fantastic system and they are investing into it all the time and it's totally free and there are no BS certificate barriers or things like that. If you want the most rigorous education and you're self-motivated and can do it, use MIT's Open Courseware. If you want a structured course with support, then use edX or maybe Coursera. And I would also say if you want to do something that's just a one-off course, like a specific history course or learn a specific programming language or learn about baseball and sabermetrics or the history of comic book heroes or something like that, then actually edX and Coursera are a pretty good option for that as well. Thanks for listening to the Learning Machine.