You can convince me that AI brainstorming is no worse than handing a student a list of possible topics for an assignment, though not as good as a suggestion or two from a teacher who is familiar with the student's interests and strengths.
But outlining the work? No, no, a thousand times, no.
Part of demonstrating understanding of complex ideas is showing that you have a grasp of how they fit together, how one connects to another. That structure and connection is what drives the organization of a piece of work. The structure and organization also reflect the process of deciding what to include and what to leave out. Without selection and structure of ideas, you end up with a pile of unvariegated details in a paper best entitled (as I often told me students) "A Bunch of Stuff About This Topic."
This has been my eternal beef with the traditional shake and bake "research" assignment in schools. You know the one-- go find some sources about your topic, then write a paper in which you re-state what they say, but in such a way that you aren't technically plagiarizing.
What is always taken from sources (usually just one) is not simply facts and data, but organization and structure. When an author goes to write, say, the sixty gazillionth biography of Abraham Lincoln, the author's most important work is to first decide what the point, the thesis, of their book will be, then to use that filter to select which details and source materials from Lincoln's life to include (a process that is often looplike-- search through materials, develop a thesis, look at more materials, modify the thesis, and on and on) and then figure out how to best arrange the details to support that thesis. There may be more looping back; in the writing, the author may decide that Source Material X doesn't really fit, so it's rejected. The author may also decide that to build a bridge between Point A and Point C they need to do additional research to find material out of which that bridge can be built.
When some high school student grabs the resulting biography for their own paper about Lincoln, they are taking not just facts from the book, but the thesis, the organization, all the decisions about what to leave in and what to take out. And that student is unaware of it all, because if the author did the job well, the book will seem like it just had to be the way it is, that there could be no other way to write about Lincoln.
Except that, of course, it is the result of deliberate choices made by the author, including uncountable choices that all other Lincoln biographers chose to make differently. It's not just the bricks you collect, but how you choose to put them together.
The foundation one builds decides much of what house can be built atop it. To imagine that AI can build the foundation and that leaves the student free to make any sorts of choices about the structure built atop it is just silly. The notion that structuring the product is a minor part of the job, and the actual marking of words on a page is the major portion is just wrong. If the writer has been thorough with the selection and structure of the work, the actual writing portion is a smaller part of the labor or creation.
Most writing problems are thinking problems, and a major portion of the thinking takes place before the actual placing of words on the page begins. To outsource that to a machine that doesn't even think is a recipe for bad writing, and worse, for a product that cannot be reasonably used as an assessment of the student. Which takes us back to the post I was writing when I started writing this one.
No comments:
Post a Comment