What this suggests is that activities like the creation of mind maps (reference from 1990s research in this Google documentation) and dual coding theory (1991) can and should be outsourced to automation.

For a student, how do we define the value of actually doing this work (proven decades ago) vs having an AI translate it to them?

Is there anything in this “research” that suggests outcomes for students are better based on this insight?

We see in figure 9 in the study a response based on an immediate test that that is marginally statically significant from a population of 60 students where the latter population is only digital readers.

Why not actually compare to a textbook?