Building Knowledge
with AI
Key Ideas
Writing with Generative AI is an Iterative process of determining where/how your ideas can develop and fit into the larger picture.​
Iterative - A process in which the same task or tasks are done several times en route to the final creation.
Triangulation - A process of connecting multiple points and sources of information. These connections help knowledge makers gain a better grasp not just of major ideas but also of their nuances.
There are some potential uses of AI that can support or complement students as they generate and organize ideas. There is also a challenge, however, in students not recognizing or possessing the subject knowledge they require to use AI effectively to organize their thinking (Eaton, forthcoming, a).
Effectively integrating generative AI into knowledge making processes depends on the user engaging on many intellectual levels: ​
1
Recognizing what subject area knowledge would be useful to stimulate a deeper conversation with the chatbot;​
2
Understanding where AI’s suggestions veer away from intellectual best practices in a subject area;
3
Knowing where their thinking can benefit from triangulation with AI tools and where it may be negatively impacted.
Without connecting these levels, there is a risk that students will end up dancing along the surface of the knowledge they're engaging.
Getting beneath this surface layer is challenging. People do not always have the disciplinary grounding to identify where their knowledge falls short, and this lack of grounding means they cannot always tell when the material they receive from AI tools exceeds what they understand.
It is important to consider that most knowledge making processes are hybrid (Pigg, 2024), requiring that the prominence of human intervention and AI output will vary. This opens opportunities for people to scrutinize outputs and move beyond this surface.
.png)
This model shows is that part of building knowledge is about determining what a person does and does not know.
Hypothetically, generative AI can help fill that knowledge gap. It can act as a point of triangulation that knowledge makers can prompt to make connections with other sources of information they have available. It can also serve as a tool that can lead people to other sources of information which can facilitate their triangulation (Eaton, forthcoming, b).
Note. It is important to see this AI's role in writing as an iterative process and not three static points.
Generative AI has made it so that written outputs can be established by non-human entities. It has also made it that outputs set by a non-human entity can build on other outputs by that entity or by other non-human entities. There is space and, indeed, a need for various degrees of human mediation in these outputs (see Graham, 2023; Knowles, 2024).
It is important to recognize that generative AI users can build from pre-established meaning, not just towards new meaning, which adds a layer of complexity to understanding how humans and chatbots may collaborate to create new knowledge (Eaton, forthcoming, b).
.png)
Remember
​Humans must graft their own knowledge into generative AI outputs to reshape, reset, and/or expand the knowledge that is created.
-
Writers must actively test AI outputs against disciplinary standards, contextual expectations, and their own intellectual commitments to ensure the accuracy of the ideas they create.
-
Seeing ideas represented by generative AI can set up metacognitive opportunities that can help people choose what fits the context, eliminate what doesn't, and expanding upon promising ideas.