Elon Musk, founder of SpaceX, Tesla Motors and Hyperloop, has some serious reservations about artificial intelligence. And just as his inventions sound like something out of a sci-fi novel, so do his fears about AI: He worries that we risk unintentionally creating a super-smart machine that could obliterate humanity. One reason Musk is so eager to colonize other planets, in fact, is that he fears AI will wipe out the human race. We need a place to which to flee, he argues, even if it’s Mars.
Some dismiss these doomsday predictions as farfetched, pointing out that there’s too much uncertainty to extrapolate where Siri, Watson and Alexa will lead us. But it seems clear that AI will have a deep impact on our work and our lives.
In 2015, NPR came out with its tool “Will Your Job Be Done by A Machine?” which forecasted that the jobs of telemarketers, umpires and even fashion models would become computerized. At the most recent ASU GSV Summit, machine-learning pioneer Andrew Ng (and co-founder of Coursera) warned audience members to steer their children away from becoming radiologists. Machine learning would most definitely render the highly specialized field obsolete.
So, what is higher ed supposed to do? How can we help students chart a course for themselves and prepare them for the careers and gigs of an AI world? If we wish to meet the rapidly changing demands of the workforce of the future, we first have to acknowledge that the educational pathways of the future might very well involve skills-based learning that does not necessarily add up to four years, or even a degree. We will have to embrace new and alternative postsecondary educational models, even if they make us feel uncomfortable at first.
A college president once explained to me that there is a reason that we distinguish college from postsecondary education. She argued that the scaffolding of learning that occurs in college cannot be replicated in other postsecondary programs, particularly those that aren’t liberal arts-oriented. But these sorts of arguments rarely benefit the “new traditional” students — working adults — who need us to uplift the notion of vocational education to help them get where they want to go. If we cling to the idea that there is something sacred and untouchable about four years of college, we will shut out myriad possibilities for our students.
At the very least, we have to examine our own assumptions about the scaffolding of degree programs, especially when an emphasis on research has been built into the professoriate. I only know because I went down that route, and it turns out that most graduate students intent on teaching at the college level actually learn very little about teaching. It can take anywhere from six years to a decade to complete your doctoral studies, and the bulk of that time is dedicated to the rite of passage known as the dissertation.
Academics can speak broadly about the concept of scaffolding in terms of prerequisites, core curricula, majors and minors, but we should ask: What does scaffolding really entail? We may understand how to guide a student from a subject like introductory biology to more advanced levels of biology, but how good are we at connecting across disciplines? Yes, most universities engage in cross-listing courses, but how do we enable students to transfer and apply knowledge from one domain to another? Institutions of higher education know how to sequence learning, but they aren’t as good at integrating and creating coherence across a vast array of courses. In fact, faculty most often need to secure special dispensation to co-teach a course across two disciplines.
In “The One World Schoolhouse,” Salman Khan of Khan Academy blames the “balkanizing habits of our current system” for denying students “the benefit — the physiological benefit — of recognizing connections”:
Khan describes the detrimental effect of artificially separating narrow specializations and “ghettoizing” learning.
As if in an effort to challenge such silos of learning, the National Academy of Engineering has created a Grand Challenges Scholars Program to give students broad, interdisciplinary problems to solve. Students in the program work on real-world problems, such as making solar energy economical, providing access to clean water, securing cyberspace, restoring and improving urban infrastructure, developing carbon sequestration methods, and many others. This same model can be expanded beyond engineering to what Stanford’s d.school executive director Sarah Stein Greenberg calls purpose learning:
By working on real-world problems, students are empowered to connect ideas without getting bogged down in identifying which disciplines they come from. Identifying how a body of understanding fits together can be more useful than understanding the boundaries between disciplines.
But this means letting go of the idea that the academy knows best how to sequence learning through departmental and disciplinary structures, or even at the course level. We may believe that we are the experts in organizing content and learning experiences for the best student outcomes, but until we figure out how to break out of educational silos, we are a long way from teaching our students the skills they need to adapt to and thrive in situations of ambiguity and uncertainty.
The future of teaching and learning should prioritize purpose learning. There’s a reason why NPR’s top 10 jobs least at risk of computerization for the next 20 years were roles like social workers, choreographers and elementary school teachers. These kinds of jobs synthesize technical skills with the uniquely human ability to judge, empathize and nurture. We must enable students to best marshal all of their skill sets along with the resources, content and knowledge out there today to engage with the problems they wish to solve in the future.
Purpose learning provides students the foundation to learn how to learn for a lifetime. Scaffolding sounds nice, but synthesis is key.
This column was first published on EdSurge.