AI-Augmented Human Thinking Pedagogy

I asked chatGPT to create a concept map of the pedagogy we developed, based on two literature reviews of research into cognitive impact of LLM use and what to do about it. The image above is a jpeg export of the cMap, which should be clickable in its native form. (I've previously posted in more detail about concept maps). This is a good first draft, but very much a draft. The advantages of a concept map are multifold:

• Like a mindmap, a concept map both visualizes the whole system and the web of relationships between nodes.

• More powerfully than a mindmap, concept maps force the mapper to explicitly label the relationships between nodes ("four thinking capacities includes," "independent thinking awakens").

• Documents and URLs can be linked to nodes and connections, providing a layer of justification/explanation of nodes and connections and an expansion of resources needed to fully understand the map. I added two links to this map and over time I may add more. I'll post an enhanced map here when I finish it.

The work that I need to put in is multifold:

• The nodes need to be rearranged so it is both readable and coherent -- related nodes ideally are on the same vertical or horizontal plane and related nodes can be color-coded.

• The small document icons indicate links to resources. I need to add them manually.

• I probably won't have the time to do it, but a cMap can be the locus of a community of interest by attaching comment threads to nodes and connections and inviting others.

In general, I think there is a worthwhile exploration to be done in regard to AI-generated/human edited concept maps, involving:

• A basis in literature reviews of relevant research (instead of relying on LLM knowledge claims.

• A human-LLM partnership is required to turn AI first-drafts into useful and reliable instruments.

• Inclusion of comment threads can make knowledge mapping and understanding into a community effort.