Working in EdTech, I can’t open a single professional-focused media/social-media platform without seeing articles about AI in schools, our workplaces, and our personal lives. Authors are concerned parents, optimistic investors, cautious engineers, curious hobbyists, and heady neuroscientists.
The articles span a huge range of opinions, from apocalyptic warnings for our health, environment, and future ability to think to visions of a future where physical work and engineering is optimized to offset any environmental damages, our relationships have built-in mediators, and learners have highly personalized curriculum that keeps them in their ideal Zone of Proximal Development (ZPD).
That last area is my field of work and former field of academic study, and I can tell you that the future is not here, yet. I can also say that I think this particular future–with personalized and adaptive–is possible, and that I think we will get there. I am also convinced that this future does not need to require students on a screen for 6 or more hours a day, devoid of peer and teacher interactions.
Why we “need” devices right now
When a student sits down at a desk to work on a paper assignment, we have limited signals about that student’s learning experience. Are they stuck? Are they digesting information? Are they distracted? Are they crushing it?
A really great teacher is going to use a host of hard-won skills to find this out: walk-arounds, visual inspection, different ways of checking in. In a well-managed class, they’ll use strategies like peer-discussions, share-alouds, and more to promote student-to-student teaching and learning and to surface misconceptions that need addressed in small groups or as a whole. While every student isn’t getting immediate feedback from a teacher, most students are getting some feedback, and from the research shared around work, doing these things is more effective than not doing them.
In our current evolution of educational tech, we see so many ways that companies are trying to increase the depth and accuracy of those signals. The idea is that if we can shorten the time to give a student quality feedback and that if we can identify students who are struggling and even find common struggles between students, we can augment a teacher’s ability to address students’ needs.
More than that, we see a clear potential to adjust the content itself, and bring the “next step” into the “challenging, but achievable” range for a student for whom the content is otherwise inaccessible due to gaps in knowledge or understanding. And given the large gap we are seeing between where a student is compared to what is being taught… this feels like an increasingly important potential to achieve.
And the default approach to achieve all of this is to put every student on a screen and to have the applications track the students’ work and use their algorithms to adjust content and roll up reports to teachers. And more radical approaches start to look like “Big Brother,” with microphones and cameras throughout a classroom.
Tensions
I’d be lying if I said that I want students on screens for hours per day in school. I don’t want that.
I’d also be lying if I said that I think that we can tackle these challenges at the scale we need to address them without applications, devices, and even (gasp) AI in the classrooms.
We’ve been thinking about this tension a lot, both in my current company and also in the EdTech community as a whole. There are real concerns over privacy, social disconnection, and the teacher’s role in a classroom. There are nuanced discussions to be had over who is ultimately responsible for the content that is presented to a student when the decision-making process is algorithmic or AI-guided.
There are also very real differences in capability between what can reasonably be done on paper vs on an application.
Not long ago, we entered the era of near-instant micro-applications–the ability to create an on-the-fly visual and even interactive explainer for almost any kind of topic. Struggling with ratios? Maybe a quick recipe scaling demonstration. Need a visual to show the “area method” of multiplication? Not a problem. Need a quick understanding check for Lewis (dot) Notation in chemistry? Here’s a quick question that uses the common misunderstandings as distractors.
Higher quality translations and feedback have never been more possible than now, and for students who are learning in their second language, this can be critical to their understanding what a question is asking.
I won’t even get into the gamification (and literal games) of education apps or the way that some of these applications are built to work together across subjects.
The Future we Design
When I am honest with myself, I don’t see these tensions as wholly solvable, and I don’t think it’s a simple matter of a “balance” to be found. I am more convinced that this is a matter of decisions to be made and directions to be prioritized. And with every decision we make, we influence the future engineering of the digital side of every compliant tool and system.
For instance, I would push to decide that a teacher must ultimately be responsible for the broad content that is shown to students, and in order to have dynamic content, all solutions must enable a teacher to review the content a student might encounter or at least the different adaptive strategies an “intelligent” curriculum might take.
Without this kind of decision, there is too little clarity and a lack of capability to have a responsible party. We’re left in a limbo where the teacher will point to a black box and say, “I don’t know where that came from,” and an application company will say, “The teacher provided the base set of content, so it can’t be our fault.”
I would also push for a decision at school levels to prioritize systems that offer analogue versions of their content without losing the adaptability or that incentivize moderated human interaction. This gives teachers the ability to mix modes without giving up too many benefits.
And, of course, I would push that schools have a system of measuring both implementation and impact of any new system they bring in as more than just a one-off. And if this triggers a little alarm in your brain around your student being a “test subject” for new technology, I think you are justified. But I also think that without this kind of measurement, we risk introducing technologies and systems into classrooms that are no more effective than standard classroom practices or worse–tech that is loved by students and teachers but that is entirely ineffective.
Leave a Reply