Comic Steve Martin as soon as stated that educating is like present enterprise. I attempt to preserve this in thoughts after I’m giving faculty lectures.
However what occurs when the entertaining professor will get upstaged by a chatbot that may produce the lecture in addition to write pupil papers and take the ultimate examination? Does the faculty class change into a meaningless joke?
There are individuals who concern that ChatGPT, Bard and different generative AI bots will let college students outsource their very own studying. However I educate media historical past. I do know that new media applied sciences don’t make individuals out of date. Video didn’t kill the radio star.
So slightly than slip some language about ChatGPT within the coverage part of my syllabus about plagiarism (which received’t cease college students who know concerning the apps that may rewrite papers to evade detection), my plan this fall is to concentrate on creating interactive classes that incorporate chatbots straight into my educating.
As an alternative of letting chatbots change the educational course of, I’ll present my college students that something that chatbots can do, they will do higher.
Lots of my college students had been already attempting out ChatGPT final yr. As a result of chatbots might be particularly helpful for performing routine duties, one pupil defined that she had began to make use of ChatGPT at her job in customer support to generate fast responses to complaints, which she would then rewrite to enhance.
Whereas chatbots are in a position to do this form of job nicely, extra difficult duties, comparable to historic essays, is usually a catastrophe. However these limitations additionally open the door to educating workouts that present college students tips on how to use this know-how of their work.
Professors educating writing abilities can have chatbots generate outlines, drafts and different lists of concepts. Then, the professor can direct college students to work in small teams to rewrite the textual content for larger originality.
Chatbots additionally supply a chance to show crucial pondering and media literacy abilities. ChatGPT is susceptible to creating up false data out of the data-driven cloud — a phenomenon its handlers euphemistically name “hallucinations.” Which means that college students should learn to examine details and confirm data, utilizing citable sources and databases.
Professors may also educate college students to be alert to the systemic racism and sexism that AI bots can perpetuate and amplify due to the supply texts they’re drawing from. I as soon as requested ChatGPT to write down an inventory of a few of the main students of the U.S. Structure and the first Modification. Its response included solely white males — as if no individual from one other background, ethnicity or gender ever studied the U.S. Structure.
An answer to this drawback? Present college students how they can provide the chatbot follow-up prompts that generate extra full solutions — say, particularly to incorporate individuals of coloration, totally different genders and numerous backgrounds. After I did this, ChatGPT readily listed Kimberlé Crenshaw, Ange-Marie Hancock and different distinguished constitutional students.
For my courses this fall, I’m additionally creating “AI Moments,” the place my college students will get an opportunity to see who does it higher: the robotic or the professor.
After I current a brand new lesson and discuss it with my college students, I’ll immediate ChatGPT to offer a lecture on the exact same topic.
To check out this concept over the summer time, I requested ChatGPT to rewrite my quick lecture on the historical past of broadcast media. Unsurprisingly, the textual content it generated was horrible. Only one cliché after one other. The AI-generated draft additionally made unhealthy phrase decisions — changing the phrase “media” with “platform” (not all media are platforms). It modified my query, “Did the emergence of broadcast TV imply the top of going to the flicks?” and as an alternative requested, “whether or not the emergence of broadcast TV resembled the demise of cinema attendance attributable to the rise of radio.” This phrase selection altered the which means of the purpose, which is that new media don’t substitute the outdated.
After I re-create this train in my classroom, I plan to have my college students search ChatGPT’s lecture for unhealthy writing that they may rewrite, turning every cliché into authentic imagery and poor phrase decisions into one thing extra exact. I’ll additionally ask them to search out and eradicate bias and fact-check for inaccuracies.
What I realized from my follow matches with ChatGPT is that I do know extra about educating journalism, writing and media historical past — despite the fact that the chatbot can draw from huge quantities of knowledge on the web. And extra importantly, it can’t share concepts precisely or in a inventive and interesting method.
I need my college students to study to make use of AI successfully, since these instruments will change into ever extra widespread and perhaps even indispensable in workplaces and in training. I believe the easiest way to do this is to make this know-how a part of their classroom expertise.
Elizabeth Blakey teaches media historical past at Cal State Northridge. This text was produced in partnership with Zócalo Public Sq..