AI and Case Writing

 I do seem to have AI on the brain lately. 

The more I think about it, if I were still coaching the less likely I would be to allow students to use AI for the writing of cases. It’s not that the AI of it is somehow ethically wrong; I mean, as far back as I can recall students have been buying pre-written cases from various purveyors. It’s the “not writing” of cases per se that bothers me, regardless of the source. Here’s why.

The act of writing something requires you to think about that something. If it’s a debate case, first you have to study the subject, then think about it, then come up with a main theme you wish to argue—that such-and-such is good/bad for this essential reason—and then you have to come up with contentions that expound on that main theme. This requires a lot of thinking on the part of the writer. Asking AI to write a case for you eliminates the need for most of that thinking. The problem here is that the lack of that thinking is the skipping of a lot of steps in your preparation for rounds. That original thinking on your part has massaged your brain with all sorts of information and thoughts and byways. When the round happens, that massaging makes you ready for just about anything, not just what you wrote but what you considered and what you didn’t write. You are as prepared as conceivably possible. Delivering a pre-written case eliminates that preparation. It also eliminates all the learning derived from that preparation. The educational value of case-writing no longer exists.

There is another advantage to writing your own case. One of the most important educational values of forensics is mastering public speaking. The speed of LD and CX undermines this to some extent but in PF, with its predominance of lay judges, effective speaking is very important. When you write something and then present it orally, it sounds like you because it is you. Your familiarity with the language used, the emphases, the general style, is implicit. Public speaking is a difficult and often frightening situation; comfort with the material being presented removes much of inherent difficulty. Being an effective speaker can definitely be a life skill learnt through our activity. Removing the prep from the speaking equation is another more indirect lessening of debate’s educational value.

Meanwhile, even if you do get an AI-written case, there is the question of its reliability. You got AI to write it for you presumably as a labor-saving action. Are you going to parse it all the way to hell in back—as you should—to ensure that its citations, if any, are accurate? Are you going to use it as an outline and write your own case based on it? Are you going to practice it to death until in fact it does sound like you wrote it? All of these would seem to eliminate the efficiencies you were seeking in the first place by not writing the case yourself, while at the same time still eliding the educational benefits that writing it yourself would have granted.

Nevertheless, the NSDA and NCFL are perfectly happy with you using AI to write your cases. It is, in a word, no skin off their noses if you skip the educational basics. After all, most of the activities they support are IEs and Congress, and the numbers behind those activities are far greater than the number of debaters. And AI is a messy subject at best, so why not avoid it if you can, despite the fact that the Sept-Oct PF topic goes at it at the roots? I wonder how many educators across the country are really well-informed on the subject. 

There is one legitimate albeit depressing argument in favor of using AI to write cases. The world students will be venturing into in the coming years will indeed be a world of AI. AI will do most of the writing of a non-creative (and, God forbid, even creative) nature. Reports, memos, brochures, proposals, etc.: most people are already not particularly good at writing but presumably good at their jobs, so when their jobs require writing, why wouldn’t they use a shortcut? It could be said that learning mastery of AI in secondary education will be of value in the long run. But this is not mastery of AI, it is simply learning one application of AI, and not a particularly desirable one at that, for all the reasons stated above. I would like to live in a world where everyone is capable of writing workable prose for memos and reports and the like. I would like to live in a world that educates everyone to do that writing. This is not me being a luddite, any more than if I wanted all art to be created by humans as expressions of humanity. Casual writing is, to some extent, an expression of personality. Do we want to lose that by making all casual writing machine-generated? 

I would hope that the NSDA and NCFL, and educators in general, are hotly debating the use of AI. Actually, I know that a lot of folks are. But if the forces against AI in education lose the battle, it is education’s loss, and ultimately humanity’s loss. There really isn’t much to debate about that.