Sebastian Thrun smiles a little awkwardly as he explains why he no longer believes in the educational revolution he sold to the world just a few years ago.
The lean, balding robotics pioneer has been instrumental in convincing investors, governments, and colleges to splurge millions on the online college education platforms dubbed MOOCs, or massive online open courses, billed as opening up quality education to anyone on Earth (see “The Crisis in Higher Education”). Thrun, a Stanford professor, helped birth the frenzy when he put his introductory artificial intelligence course online in 2011, accidentally attracting 160,000 students.
Amazed by the response, he took time out from Stanford and also from a side job working on autonomous cars and other research at Google to found Udacity, a company offering MOOCs in computing, math, and physics.
It attracted $160 million in venture capital investment and teamed up with San Jose State University to offer courses valid for college credit. But within two years of Udacity’s launch, Thrun began to question whether MOOCs could make much of a mark on the world in their current form.
Udacity’s completion rates were as low as 2 percent, and the people who made it through were mostly the kind of well-motivated students already served well by conventional institutions. Meanwhile, it was clear that many people sought out MOOCs to improve their employment prospects, but providers seemed more focused on emulating colleges than serving that need.
Thrun, now Udacity’s president, didn’t waste much time turning the company away from its founding ideas to market itself as a gateway to a new job in the tech industry. The company partners with employers such as Amazon and Facebook to offer “nanodegrees” that are tightly vocational. “We are happy we moved out of the MOOC space,” he says. “We are able to outpace universities with curricula that you wouldn't find in any university, that empower students to find jobs.”
Udacity works with its corporate partners to create courses intended to produce job candidates with skills those same companies find in short supply, such as machine learning and mobile app development. More than 30 companies, such as Intel and Samsung, have signed up with Udacity as “hiring partners” that gain access to graduating students before they hit the wider job market.
Thrun says this model allows Udacity to fill an important—and lucrative—education gap that college-style MOOCs don’t. He argues that technology has created new jobs and changed existing ones faster than colleges can keep up, and that many people can’t afford the time and expense of conventional education anyway.
“There's growing mismatch between people's educational needs and this idea of once-in-a-lifetime education at college,” he says. “That made sense when people had one job for life. Now technology moves fast and people are forced into new jobs quickly.”
Udacity currently offers 12 nanodegrees, ranging from front-end Web developer (created with help from companies including AT&T and Google) to self-driving car engineer (shaped by partners including Mercedes-Benz and Uber’s trucking division, Otto). Some 3,000 people have graduated from nanodegree programs in the past two years, and 13,000 more are currently enrolled. Around 900 people have gotten jobs related to the program they studied.
Students pay $199 a month for most nanodegrees, and work through courses at their own pace. Taking six months to finish is typical, and to give students additional motivation, for most courses the company refunds half your tuition if you finish within a year. For some courses, paying $299 a month gives you the right to a full refund if you don’t get a job within six months of graduation. (So far only one person has taken up that deal.) Student fees provide most of Udacity’s revenue, although some partner companies contribute financially as well. By contrast, many coding and other technology boot camps require students to enroll full-time, and fees can be much higher.
Dan Haddigan signed up for Udacity’s front-end Web developer nanodegree late in 2014 in the hope of finding better prospects than he had working as an art seller, a job he landed after graduating with a fine arts degree. He praises the personal feedback provided on coding exercises and project work, and career counseling offered once a course is done.
Haddigan completed his course in about five months, working on it before and after work most days. Worries that his atypical education might hold him back evaporated when he got the first position he applied for after completing the Udacity class. He is now a Web developer with IntuitSolutions, a Philadelphia company specializing in creating e-commerce sites.
Haddigan thinks that a growing need for technology skills has made companies more open-minded to alternative backgrounds like his own. “They are willing to overlook things like whether you have a formal degree as long as you have the know-how,” he says.
That kind of thinking led IBM’s Watson group to work with Udacity to co-create the Artificial Intelligence Engineer Nanodegree (Amazon and Chinese ride-sharing company DiDi also contributed). Rob High, chief technology officer for the Watson group, says researchers inventing new artificial intelligence techniques will continue to come from conventional, elite educational backgrounds. But large numbers of less elite programmers and managers also need to understand the technology if IBM and others are to deploy it broadly, he says.
The way Udacity’s new model neatly matches the motivations of students and tech companies suggests it could become well established even outside technology, says David Passmore, a professor of education at Penn State University.
The company’s reasonable tuition and sharp focus on jobs has value in an era when tuition is more expensive than ever, he says. The way Udacity works with companies to create courses provides them a relatively easy and direct way to shape the skills in the job market, says Passmore, who could see it easily adapted to industries like manufacturing.
Thrun doesn’t have plans to expand beyond technology, where Udacity’s brand and his network are strongest, but Erik Brynjolfsson, director of the MIT Initiative on the Digital Economy, says that still leaves a large market, since technology skills are now needed in every industry (see “How Technology Is Destroying Jobs”).
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
Data analytics reveal real business value
Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.
Driving companywide efficiencies with AI
Advanced AI and ML capabilities revolutionize how administrative and operations tasks are done.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.