In the late summer, professors leading three traditional Stanford classes—an introduction to AI, to databases and to machine learning—decided to offer their classes online for free to anyone in the world. Those who successfully pass the class will earn a "statement of accomplishment," autographed by the profs. (Stanford students taking the classes in situ will also rack up credit hours, too.)
Professor Andrew Ng is teaching machine learning; professor Jennifer Widom is serving up an introduction to database management. But the class that has proven wildly popular is "An Introduction to AI." The duo that jointly teach the AI class are hardly your rumpled-shirt, pocket-protector kinds of profs. Both have day jobs helping run research at the grand Googleplex, not all that far from Stanford. They’re also heavyweights in the AI world, particularly Sebastian Thrun who several years ago built an autonomous car that snagged top prize in a DARPA "Grand Challenge" contest.
By the time the class started on October 10, 160,000 people had signed up online. A huge number were just doing brain-buzzing window shopping: Professor Peter Norvig says he wondered whether some really had the math chops to take the course or just watched Terminator and though AI was cool. Thrun warned many off by cheerfully promising they’d have to clock the same amount of time on homework as a "good" Stanford student—up to 12 hours per week.
Three weeks into the class, EdSurge was privy to some stunning statistics. Yes, there was a huge attrition. Even so, a full 35,000 people handed in the first three weeks’ of homework, making this likely the largest single class on the planet. (A distant second: a multi-site dance class for 26,797 rollicking primary school students in Liverpool on a day back in July 2008.)
The students are lurking all over the world; most of them have full-time jobs, leaving only the wee hours of the morning for reviewing lectures and tackling homework sets—a habit shared by many of Stanford’s engineering students but for different reasons. Some watch the videos as many as 20 times apiece. One geographic disparity: Chinese students account for less than 1% of the online cohort, largely because the videos are delivered via YouTube and China is (for the most part) blocking YouTube.
Norvig and Thrun’s videos are spare and oddly low tech: most are only two or three minutes long and feature a pen scratching out equations on a piece of paper and a voice over talking through the math. That simplicity promises the fewest technical hiccups for students—and the easiest way to record the lectures, Norvig says. Thrun concedes that he may spend seven hours making an hour-long video, just to get every detail right. He had better:
"If I make a mistake, I’ll get about a thousand emails about it."
Even to the professors, lecture by video can feel surprisingly intimate. Thrun concedes.
"I can get away with things when I teach a lecture class that I can’t online. Now I’m thinking intensely about the student experience."
Some pedagogical approaches just don’t work: Thrun likes to pepper his lectures with questions that he may not be ready to answer. That doesn’t fly online. Norvig notes.
"The students aren’t too happy because they expect the instructor to know everything in the virtual environment."
Discussion groups seem to have sprung up on platforms as diverse as Reddit, Facebook, and Stack Overflow. And cheating? It will probably happen, acknowledge the professors. Stanford will award those who pass the final exam a "certificate of accomplishment," signed by Thrun and Norvig—incentive enough to take the class but hopefully not enough to go to the trouble of cheating, say the profs.
The biggest emotional reward for many may be proving—at least to themselves—that they can keep pace with a Stanford student, Norvig says. There are 175 Stanford students taking the class live and in situ, for course credit. The final grades of both live and online students will include a class ranking, Thrun promises. So far, he’s been impressed that the percentage of top performers in the online contingent is on par with the percentage of top performers at Stanford.
Just how disruptive such online classes will be to traditional bricks-and-ivy schools is unclear. Norvig feels it isn’t a stretch to think that online classes have some of the same needs as physical ones. For example, homework is due from everyone every Monday, regardless of geographic location, because in physical environments, "homework is done because it’s due," he contends. Norvig may have a point. While online class technologies can spur unprecedented access and collaboration, instructors may need to find ways to recreate the social norms and motivators that foster student engagement in the physical world to create real and lasting success in online education.
COMMENTARY: A bold experiment in distributed education, "Introduction to Artificial Intelligence" is being offered free and online to students worldwide from October 10th to December 18th 2011. Enrollment for this class is now CLOSED, but you may follow the coursework on YouTube. Check the Introduction To Artificial Ingelligence website for future classes.
Artificial Intelligence is the science of making computer software that reasons about the world around it. Humanoid robots, Google Goggles, self-driving cars, even software that suggests music you might like to hear are all examples of AI. In this class, you will learn how to create this software from two of the leaders in the field. Class begins October 10. The course will include feedback on progress and a statement of accomplishment.
Taught by Sebastian Thrun and Peter Norvig, the curriculum draws from that used in Stanford's introductory Artificial Intelligence course. The instructors will offer similar materials, assignments, and exams.
Sebastian Thrun is a Research Professor of Computer Science at Stanford University, a Google Fellow, a member of the National Academy of Engineering and the German Academy of Sciences. Thrun is best known for his research in robotics and machine learning.
Peter Norvig is Director of Research at Google Inc. He is also a Fellow of the American Association for Artificial Intelligence and the Association for Computing Machinery. Norvig is co-author of the popular textbook Artificial Intelligence: A Modern Approach. Prior to joining Google he was the head of the Computation Sciences Division at NASA Ames Research Center.
I am happy to report that I finished Unit 1 - Introduction to AI. I found the introduction fairly straightforward, and uncomplicated. The Intro to AI just covered the basic AI concepts. There was a one question quiz at the end of each lesson, which was answered in a separate video. I got through all 15 videos in about 45 minutes. Now its lunch so I am posting this sucker.
For students, both in high school and college, there are a number of online courses. MIT Open Courseware provides several in the fields of architecture and planning, engineering, humanities, arts and social sciences, management, science, and others. I have taken a few of their marketing courses, finance courses, statistical analysis, to name a few. If you surf through the website of many universities you will find online courses, some are open only to students, but many are open to the general public. It's a good time to challenge yourself, and learn something new.
Stanford Engineering Everywhere offers free online engineering courses, including computer sciences, artificial intelligence, linear systems and optimization and other engineering-related courses.
Open Culture offers 400 free online courses in biology, computer science, economics, engineering, history, literature, math, philosophy, physics, political science, psychology and others. You will also find games of all types and a variety of videos on a broad range of subjects from music to the war in Afghanistan. Well worth checking out. And, all free.
If you are too lazy or don't have the time to find it out of a textbook, you may wish to consider online services like: Quora.com, Yahoo Answers, Student of Fortune and Khan Academy.
Courtesy of an article dated November 14, 2011 appearing in Fast Company Co-Exist
Comments