Blog Post— 5 Min Runtime

The Machines Are Good at This

We spend a lot of time in education talking about the promise of technology in the classroom. Terms like “personalized learning” and “AI-driven instruction” and “adaptive learning” become buzzy. We view algorithms in awe and wistfully predict their power in shaping human cognition.

I think it’s prudent to ask what is edtech really good at doing? Consistently, and over a long time?

Take a step back to consider how we learn. A great model for this is Bloom’s Taxonomy. At the base, the foundation, of this taxonomy are two core objectives: Knowledge and Comprehension. It is very challenging, if not impossible, to demonstrate higher levels of cognitive processes without mastering a degree of knowledge and comprehension.


Bloom’s Taxonomy Poster by Dustin Johnson is license under CC By- NC 4.0
Changes made: Cropped Poster

In recent years cognitive scientists have found several effective ways to master remembering and understanding. For example, students will likely retain more knowledge when using the retrieval practice. This is when students tap into their memory after having learned a concept and bringing forth the knowledge on a repeat basis. Transferring memories from long term memory, to working memory, and back to long term memory strengthens knowledge retention.

Machines and algorithms are really great as tools to support the lower levels of Bloom’s taxonomy.

They struggle – for a host of different reasons – with the higher levels. In other words, the more abstract thought required by the students, the more the machine runs into trouble.

But that’s okay!

Because teachers and students both wrestle with the finite constraint of time – especially constructive use of time. When teachers and students offload their lower-level learning to software programs, the more time can be spent tackling higher-level learning that requires humans.

For example, when I taught social studies my students simply had to memorize many core American history facts. Sure, I could ask for essays or projects, but unless they had a basic understanding of the US Constitution, the outcome of such learning would be poor.

Flashcards work well with memorization. Flashcards with a little bit of spice (e.g. gamified) were the cherry on top for engagement. I used a (still) popular program called “Quizlet” with my students. With constant use, the software program helps students with rote memorization so they can move up the taxonomy ladder and build towards a different level of mastery.

Edtech freed me up to spend more time working with my students on higher level thinking. Technology worked as intended when applied to an appropriate level of learning.

Like many industries, the hype of AI is well ahead of the help it can provide. In most cases AI is really good at replacing the lower level function so we can spend more time working on higher order problems.