AI and the Dumbing Down of Education

Education Policy Brief #88 | By: Rudolph Lurz | January 4, 2023

Photo taken from: www.linkedin.com

__________________________________

In many ways, 2023 was the year of AI. ChatGPT is ubiquitous in boardrooms and classrooms alike. Its usage is prevalent across grade levels and industries. AI bots have passed law school and MBA examinations, along with the United States Medical Licensing Exam (USMLE).

Chatbots are replacing customer service phone operators. They’re also available, and customizable, on apps such as Snapchat. In short, AI bots can provide full essays for students along with restaurant recommendations for celebrating that one-click B+ afterwards. The technology is progressing faster than school districts’ capacity to keep up with it. Many districts are placing bans on ChatGPT, although proving AI usage remains a challenge. In a recommendation to the G20 titled, “Reconsidering Education Policy in the Era of Generative AI”, the authors note, “While tools to detect whether content is created by AI are being developed and deployed, their effectiveness varies and the sophistication of means and mechanisms to get around them will continue to evolve in parallel.”

AI usage represents one of the rare areas where political party affiliation does not necessarily reflect where policy actors stand on the issue. Some Democrats and Republicans are alarmed by it. Others believe it represents a chance for economic growth and job creation.

Democratic U.S. Senator Richard Blumenthal noted that deep-fake, AI-generated voice recordings could have harmful effects on society and geopolitics alike. Many GOP policy makers are worried that AI is apparently too woke and that ChatGPT itself has anti-conservative bias. Entrepreneur Elon Musk noted that AI represents a civilizational risk. Former New Jersey Governor Chris Christie accused primary candidate Vivek Ramaswamy of amateurism, noting at a recent debate, “I’ve had enough already tonight of a guy who sounds like ChatGPT.”

Despite the fact that AI usage is considered plagiarism, AI sites like Magic School also exist for educators.

Therefore, within one year of AI’s widespread usage, teachers can design quizzes, assignments, and even full curriculum maps using AI assistance, and students can then complete them using ChatGPT.

In all of this heavily formulaic and repetitive AI-generated drivel, is learning even taking place? Is teaching?

One of the most frequent refrains in the GOP’s culture war is that teachers should “stick to teaching” and avoid “indoctrination”.

In that regard, AI provides the perfect answer for the GOP’s goals. GOP policy actors have frequently advocated for less rigid standards for education licensure, noting that veterans and business representatives should be allowed to teach in the classroom. There’s a severe teacher shortage nationwide, and positions are difficult to fill.

AI provides potential standardization of the curriculum, shovel-ready lessons and assessments for novice educators to implement, and potentially an end to critical thinking and all of the sensitive topics associated with it. Some GOP policy makers might claim to dislike AI, but it provides the inevitable conclusion of their culture wars on education: a vanilla, just-the-facts curriculum that avoids any contentious topics that make students uncomfortable.

If you’re a conservative GOP policy actor, what’s not to like?

ANALYSIS

I am an educator at a good private school. I teach English and a DE Communication course, which allows students to earn college credit at a nearby university. I frequently catch students using ChatGPT and other AI bots to complete assignments. AI detectors like GPT Zero are only a first line of defense. They will analyze a document and provide a likelihood that the document was generated by AI.

I tell my students that AI has a distinct, pedantic “accent” that is repetitive and needlessly verbose, but also uses perfect punctuation and grammar. If I suspect AI usage, I run it through an AI detector. I also have ChatGPT myself, and run my assignment prompts through ChatGPT to see what comes out. If a student submits work that is identical to what ChatGPT has produced for me, and the document shows up as 90%+ AI generated using an AI scanner, I have enough evidence to go to administration to refer the student for an academic honesty conference.

Perhaps the simplest tool I use to prove AI usage is the humble document history button on Google Docs. It shows each change the student made to the document, right down to individual backspace keystrokes. If there is only one edit made and 1000 words appear at once, it is likely that the document was copied and pasted. Students can claim that they did not use ChatGPT when they are sitting across from the Academic Dean. It is harder to claim that they wrote 1000 words with perfect grammar and punctuation in one single edit on their assignment.

With each of these academic honesty discussions with my supportive administrative team, we approach it as a teachable moment for first-time plagiarism incidents. 20% of college students do not consider AI usage as cheating. It is important to show clearly the line between using ChatGPT as a tool and abusing it as a 24/7 version of George McFly from Back to the Future to complete assignments using digital “handwriting” that looks like their own.

As an educator, I see AI’s potential and wish I had it during my doctoral program. Not to write anything for me, naturally. On my own writing website, I am highly critical of ChatGPT’s style and ability level as a writer. ChatGPT will also “hallucinate” when it does not know the answer to something, since it is programmed to speak confidently about any topic. If it does not know the details of an answer, it will make them up. Maybe that is why it is so bad at basic mathematics.

That said, I think ChatGPT is really useful as a source mining tool.

My wife is a surgical oncologist. I sat across the table from her when she was working on a pancreatic cancer presentation. I asked ChatGPT to provide the best treatment options for female pancreatic cancer patients under the age of 50. It provided a mix of pedantic, basic, and just plain useless information, along with some recommendations that stretched the truth, all without documentation or sources. My wife got a good laugh from Dr. ChatGPT. Then I asked it to provide a list of academic journal articles from reputable sources that discussed pancreatic cancer incidence in female patients under the age of 50.

My wife stopped laughing. In seconds, ChatGPT completes one of the most annoying parts of the academic research process-paring good information from garbage on the Internet. I show my students how to use ChatGPT to find good sources, but then stress that they must read the sources themselves.

In short, effective usage of AI in the classroom is more about coding than copy/paste. As the common adage goes, garbage in, garbage out.

The problem that I see coming is that education is going to start looking like the garbage out portion of that phrase. GOP policy actors who are so concerned about so-called “woke” indoctrination will use AI to create the most vanilla and non-controversial curriculum possible. Donald Trump Jr. and other GOP critics frequently use the term “Orwellian” in their criticism of “woke leftists”. I fear that these same GOP policy actors who incorrectly invoke fears of 1984 will create an education system that looks a lot like the world of Fahrenheit-451, and that AI will help build it.

Captain Beatty is Guy Montag’s boss in Fahrenheit-451. Early in the novel, he defends the education system in this dystopian society, in which critical thought is banned and trivial facts are memorized and regurgitated by students instead.

Beatty states,

“You must understand that our civilization is so vast that we can’t have our minorities upset and stirred. Ask yourself, What do we want in this country, above all? People want to be happy, isn’t that right? That’s all we live for, isn’t it? For pleasure, for titillation? Colored people don’t like Little Black Sambo. Burn it. White people don’t feel good about Uncle Tom’s Cabin. Burn it. Someone’s written a book on tobacco and cancer of the lungs? The cigarette people are weeping? Burn the book. Serenity, Montag. Peace, Montag. Take your fight outside. Better yet, into the incinerator. Funerals are unhappy and pagan? Eliminate them, too…Let’s not quibble over individuals with memoriams. Forget them. Burn all, burn everything. Fire is bright and fire is clean” (pp. 59-60).

Fahrenheit-451 uses firemen to burn books in the homes of dissidents, but in most cases, the firemen are not necessary. Adult citizens have their reality TV programs that provide mindless entertainment, after learning and reciting mindless information during their schooling. There’s no need for critical thinking. Just turn on the TV.

I am honestly more concerned about AI sites for educators than ChatGPT for students.

If AI writes the curriculum and students use AI to find the answers, we have no need to fear 1984. We will enter the world of Fahrenheit-451.

No so-called “woke” material will be discussed, and no firemen will be necessary. Students will not have to experience the pain of thinking about things which might trouble them. Donald Trump Jr., Ron DeSantis, and other Republican culture warriors will be happy.

I do not see the “civilizational threat” from AI that Elon Musk sees. I see us undoing the work of centuries of education and philosophy, all in the name of convenience.

AI is unlikely to destroy us. But if it is placed in the hands of conservative policy makers, it will make us all a little more stupid.

__________________________________

Further Reading

Check out usrenewnews.org/education for more coverage on this subject. Get the latest updates from our reporters by subscribing to the US Renew Democracy Weekly Newsletter, and please consider contributing to Keeping Democracy Alive by donating today! We depend on support from readers like you.

DONATE NOW
Subscribe Below to Our News Service

Pin It on Pinterest

Share This