What WALL-E Teaches Us About Adaptive and Personalized Learning (David Wiley)

David Wiley (@opencontent) is co-founder and chief academic officer of Lumen Learning. This post appeared in EdSurge February 24, 2016.

 

One hundred years from now, Earth is an epic wasteland of French fry containers, water bottles and other remnants of mankind’s consumerism. Humans have evacuated on a starship, leaving behind an army of robots to clean up the planet.

So goes the plot of WALL-E, a lighthearted, animated film set in a post-apocalyptic future. While much of the movie’s screen time is dedicated to the exploits of our rusty bot hero, the film also makes some gentle commentary on the impact that living with advanced technology for centuries might have on people.

The starship that humans call home has fully automated systems designed to provide for all their needs—food, drinks, entertainment and even transportation around the intergalactic spacecraft. Although exercise is critically important for people in space to prevent bone loss and other deleterious effects of weightlessness, the humans are moved about by intelligent hovering chairs. They get even less exercise than their Earth-bound ancestors and the effect is, in a word, flab-bergasting.

Just at the moment when exercise is more important than ever, AI has stepped in to prevent the human characters from engaging in it. This sentence, perhaps more than any other, sums up my concerns about adaptive and personalized learning as currently imagined in today’s mainstream edtech marketplace.

Learning on Autopilot

Our world is awash in information. There is some disagreement over the exact amount, but reasonable estimates state that humanity records and transmits a little over three exabytes of information every day. Now, more than ever before, people are desperately in need of skills that will help them determine what is worthy of their attention, and how to effectively study and learn over their lifetime in this increasingly ill-structured and information-rich environment.

Yet what is the primary purpose of most adaptive or personalized learning systems? To eliminate the complexity of deciding what to study, how to study or how long to study. The highest aspiration of these systems seems to be automatically selecting, sequencing and presenting just the right information for the learner at just the right time. All the learner needs to do is sit back and click “Next”—no judgment or thinking required.

Some will no doubt argue that the brainpower students spend trying to decide what to study or how to study represents “extraneous cognitive load” that prevents students from giving their full attention to biology or economics, and that this distraction will manifest itself in assessments as “construct irrelevant variance”—irrelevant variables that affect results. And from the narrow perspective of the discipline, that may be true.

However, now more than ever students need explicit support developing the skills that will allow them to successfully navigate—rather than drown in—the ocean of information that awaits them post-graduation. When faculty choose to let adaptive or personalized systems make these choices on behalf of students, we are complicit in the atrophy of these critically important skills in our students. No wonder so many adaptive platform websites feature student quotes saying, “I wish all my teachers used [insert product name here]!” Prolonged exposure to these systems makes traditional studying increasingly difficult for students as their study skills slowly fade away.

From Automation to Empowerment

This is not a Luddite, anti-edtech rant. I firmly believe there is an important role for technology to play in enabling more adaptive and personalized experiences for students. But rather than designing learning technologies that actively deskill students, adaptive and personalized providers should look for ways to use technology to support students in the development of their agency, metacognition and learning-to-learn skills. Rather than making complicated decisions on behalf of students in a black box, these systems should surface their data and support students in evaluating them and making their own decisions about what and how to study.

These criticisms generally apply to the faculty context as well. Take the above paragraphs and replace “student” and “learning” with “teacher” and “teaching,” and the logic largely holds. Adaptive and personalized systems appear to be actively deskilling faculty as well, attempting to replace them rather than augment them. There are tremendous opportunities for these technologies to empower teachers, broaden their agency and improve their teaching skills. Oddly enough, at the same time they deskill faculty, these systems also make unrealistic skills demands on them by presenting them with data dashboards that are only valuable for faculty who—in addition to their disciplinary knowledge—also have graduate degrees in data science and instructional design.

In the long run, the true power of adaptive and personalized systems will only be realized when they are designed to simultaneously support student learning in the discipline and increase human agency, giving students and faculty the chance to develop their metacognitive and pedagogical skills rather than contributing to their slow demise. I’m looking forward to that future.