It is said that information wants to be free. But what about college courses from elite universities?
A new Mountain View, Calif., start-up founded by two Stanford computer science professors believes that technology has the ability to democratize education. To prove it, Coursera has put online 40 free courses taught via recorded video by professors from Stanford, Princeton, the University of Michigan and the University of Pennsylvania. The classes range from the highly technical (Machine Learning) to the ethereal (Modern Contemporary American Poetry).
"Top universities offer an amazing education to a fraction of the population," said co-Founder and co-Chief Executive Andrew Ng, who is on leave from Stanford to run the company. "With online technology, I'm hoping we'll be able to educate not just thousands of students, but millions."
The idea has attracted $16 million in venture capital funding from Kleiner Perkins Caufield Byers and New Enterprise Associates. Coursera will use the funding, in part, to expand its 12-person staff by as many as 17 this year. New hires will be split between engineering and course operations. The company is looking to hire front-end and back-end software engineers with experience in Web development and user interface. Course operations jobs entail working with professors to develop the specifications of how online courses will work, such as how long video lessons should be and how grading is done, Ng said.
The company hasn't yet developed a revenue model and Ng said that it is focusing on its products and attracting a large number of users. "If we can change millions of students lives, there will plenty of ways to keep the effort sustainable," he said.
Coursera is tackling challenging computer science problems in its mission of educating large swaths of the world, Ng said. For courses in which testing can't be done in a multiple-choice format, poetry for instance, the team has developed a peer-grading method and has had to develop guidelines for how students assess each other's work. For other courses, crowdsourcing will be used so that individuals will each do a small piece of grading and then have the results aggregated.
"We believe we can use peer grading and potentially attain grading accuracy that is as accurate or more accurate to that of a teaching assistant," Ng said.
Coursera's class platforms can handle as many as 100,000 users, which gives the company's engineers a lot of data about the way students learn and the ability to adapt teaching methods based on those habits.
For instance, when Ng taught a machine learning course online last year, he had 2,000 students make the same mistake on an exam. Students reversed two steps in the algorithm they were asked to perform, so in future exams Ng created a custom message that encouraged test takers to reconsider the order of steps.
Ng believes even greater innovations will come out of the melding of Web technologies and education.
"Online education is in a very immature phase," he said. "Even though our websites work, at some level we are barely scratching the surface."
Write to Joseph Walker at Joseph.Walker@dowjones.com