Individualizing Education with AI

As I surf the web I find an increasing number of posts about “deep learning”… I’m always disappointed to discover the learner is a neural network computing machine, not a human – not even an animal. See, for example

Deep Learning in a Nutshell: Reinforcement Learning

I’d like to apply this kind of deep machine learning to a problem in human education: the matching of learners’ characteristics (background knowledge, learning styles and goals) with learning objects (more properly labeled “teaching objects” including Open Educational Resources, free-lance teachers and communities of practice). The machine (AI) would scan the profile of the learner and search online for the most appropriate collection of teaching objects to meet the learner’s goals given his/her abilities and background knowledge.

This is actually not a very challenging logical problem for a computer but it does involve gathering a lot of data and learning (by the machine) about which matches are useful to the learner. Two metadata considerations make this difficult to implement:

First: We don’t have very elaborated ways of describing learner characteristics. At the moment we usually note academic subject, language (natural, not machine) and educational level of the learner. There’s a lot more to know about learning styles, for example, whether the student is

  •  self-directed or other-directed
  • predominantly visual, auditory or kinesthetic
  • intellectual strengths and weaknesses (see SOI)
  • physical abilities/handicaps (accessibility)
  • solo or social (independent or classroom)

to name just a few.

Second: When we catalog teaching objects (Open Educational Resources (OER) in, say, OpenStax CNS, OERCommons, MIT’s Open CourseWare, we don’t provide the learner enough information about the resource to figure out whether s/he can benefit from the material offered.

By combining crowdsourcing of feedback on materials as learners try to use them and developing better descriptions of learners we would have the prerequisites to use deep learning AI’s to teach many more students much more effectively.

In future blogs I’ll explore learner characteristics, OER and other ideas we might use to implement a new “open educative system” that could support learning in this century better than our current classroom-teacher-school-based “educational systems” do today.

If this topic interests you, I’d love to hear from you. Please leave a comment here or email 

oercommons-fishing