Udacity aims to teach 160,000+ students statistics

Author: Stephanie Kovalchik

With a research lab at Stanford, a VP position at Google, spots on Popular Science's list of the brilliant 10 and Fortune's 2010 list of the Smartest People in Tech, you might think that Sebastian Thrun would be ready to rest on his laurels. But the innovator whose drive launched the driverless car isn't about to slow down. So what, you may ask, does the visionary behind Google's Street View and Project Glass plan to do next? To teach the world statistics...in 7 weeks.

Sebastian Thrun (center) listens intently to a

discussion about SpaceX (SpaceX founder

Elon Musk is left of the picture).

Image by Erik Charlton/Wikimedia.

The story behind this ambitious mission began in 2011 (you might have guessed by now that Thrun functions on an accelerated time scale). In the summer of that year, Thrun and colleague Peter Norvig, Google's Director of Research, decided to embark on a not-so-little experiment. Inspired by Salman Kahn's presentation at TED 2011, they were curious to test whether the Kahn Academy model of online instruction could work for college-level courses in the computational sciences.

So, Thrun asked engineers David Stavens and Mike Sokolsky to join his team and help him build the online infrastructure that would put his and Norvig's Introduction to Artificial Intelligence on the Web, where anyone anywhere would be able to take it for free. What happened eight weeks later, no one expected nor is likely to forget. By the time the first Intro to AI lecture went live, 160,000 students from 190 different countries had enrolled.

The success of the Stanford experiment fundamentally changed the direction of Thrun's career. When the last final of Intro to AI was graded and 23,000 certificates of completion sent out, Thrun calculated that, by opening up one course to the Internet, he had been able to reach more students than most university professors would in an entire career. With the potential impact he could make through online education, Thrun knew he couldn't return to the traditional classroom. He had, in his own words, "taken the red pill" and "seen Wonderland."

Today, that Wonderland is Udacity, a private educational organization launched by Thrun and his team, whose mission could be the most radical and most challenging its architect has yet undertaken - the democratization of higher education. Since February of this year, with a team of eight engineers and the financial backing of Charles River Ventures, Udacity has been building on the success of the Stanford experiment to bring hands-on computer science training to everyone. Even in that small period of time, the university has attracted more than 50,000 participants to study topics ranging from building a search engine to programming a robotic car.

There are a number of reasons why some commentators believe Udacity is leading a 21st-century revolution in education. Udacity courses are delivered through an easy-to-use interactive online environment, which can be accessed at any time and on any device with an Internet browser. Udacity has no tuition fees and an open enrollment policy. The only requirements for admission are a valid email address and an enthusiasm to learn. And, unlike free courses on iTunesU, Udacity's classes are designed specifically for the Web and are like no other college course out there.

At Udacity, courses are organized into units covering roughly one week of material. Each unit is divided into a collection of micro-lectures of 2- to 3-minute duration, just long enough to cover a single concept. The YouTube-length clips are ideal for reviewing challenging topics, as students can repeat them as often as they want like favorite songs on their iPod. Believers in "learning by doing", the Udacity team punctuates the micro-lectures with hands-on tasks to reinforce the concepts introduced. These mini-challenges come with automated feedback and instructor-guided solutions. As impressive as these interactive, self-directed features are themselves, it is only when they are combined with high-quality instruction that they become powerful tools for education. Like the Kahn Academy, Udacity's greatest achievement has been its ability to connect students to awesome teachers. Udacity professors, like David Evans and Westley Weimer, have an enthusiasm for their fields of expertise and an infectious, engagingly geeky love for problem-solving that make being one of their students a reward and delight.

Next week, when Udacity begins its third hexamester (1/6th of the calendar year), it will face a milestone. The curriculum is going to broaden to include several courses in mathematics and science. One of the new offerings is an Introduction to Statistics (ST101) to be taught by Thrun himself. This will be an extraordinary opportunity for students on the Web to learn how the leading mind in probabilistic robotics thinks about data. Deciding to enroll could also help Thrun make history, since he is determined that ST101 will be so embarassingly u-dacious that it will break the enrollment record set by Intro to AI.

Does this mean that Udacity's ST101 is the future of statistics education? Similar efforts of massively open online education like Stanford Professors Andrew Ng and Daphne Koller's Coursera and Harvard and MIT's joint venture edX are underway, which, at a combined investment of 76 million dollars (59.8 million €), suggest that the movement is more than hype. Still, despite an exciting momentum for this new paradigm of higher education, its quality and sustainability are uncertain. Economies of scale would suggest that online universities could be more economical than traditional ones, but if education wants to be free it could get costly.

In this respect, Udacity has an edge over its competitors. A recent partnership with Pearson VUE to offer voluntary official credentialing for a fee and a resume-delivery system for a select list of employers could each help the university remain viable and tuition-free even as it continues to grow. But what really makes Udacity special is the sheer zeal of its founder, whose dedication to the project's success has already cost him $200K (157K €) and a tenured position at Stanford. There are few people who think it would be the coolest thing to teach 160,000+ students statistics - and probably just one who has the skills to pull it off. Whatever the outcome, Thrun should be applauded for his effort.

Just consider what the implications would be if he succeeds. It would mean that over the 7-week course, for every practicing statistician in the US there would be more than 8 Udacity students, from every corner of the world, learning statistics. It would mean that the students of ST101 would out-number the populations of Greenland, the Cayman Islands, and Liechtenstein combined. It would mean that, if Thrun read at a pace of one name per second, he would need 44 hours to read through his entire class roster! By any measure, ST101's debut on June 25th is an event every statistician ought to follow.