Is Artificial Intelligence (AI) Ruining Education?

The impact of AI on education and student work

When OpenAI released ChatGPT in 2022, I remember feeling a sense of dread. I remember feeding it my essay assignments (I’m a high school English teacher) and watching the blinking cursor spit out sentence after sentence. Paragraphs. Transitions. Conclusions. I was rather impressed.

Until I read the essay.

It was unspecific. Generic. Broad. Vague. It was glaringly a fake essay.

I’m not impressed by AI.

And I’m here to tell you, the large language model (LLM) systems that people refer to as Artificial Intelligence (AI) are not that impressive.

ChatGPT reminds me of SmarterChild, an AOL Instant Messenger chatbot that would talk to you and answer questions. The way that I view ChatGPT is a narrative version of Google search results. The essays read like they were directly lifted from Sparknotes.

Sure, AI creates poetry, fiction, lesson plans, essays, and more. It can also write your cover letter, resume, or business plan, but the product will always be an aggregate of information that is already out there. It’s like conducting research. I will admit that I don’t use AI that often as a teacher, but I can tell you that I use the internet for almost everything. ChatGPT scrubs the internet and presents the information to you as a chat.

This is why AI is not a problem in education. It’s a new research tool.

What limits do teachers set on the use of AI in their classroom?

Obviously, teachers can’t accept an essay or assignment completely written by AI. But what if a student used AI to write an essay, then re-work each sentence and paragraph into their own language. Is this allowable? Is the previous situation any different than a student telling another student what to write down?

The teacher has to establish a clear and fair boundary on the acceptable use of AI. My advice is to take the issue head on. Show students what is acceptable and what is not acceptable.

AI also has a credibility problem. Everyone should be skeptical about the results that AI produces--it’s not always correct. But this has been a problem since the internet. Teachers still have to teach the research process and how to evaluate credible sources. Nothing new.

In a study that surveyed students about cheating before and after generative AI, researchers found that there was no discernible difference between the rate of cheating before and after the introduction of AI. One surprising note from the survey: they found that 59-69% of students admitted to engaging in at least one academically dishonest action in the last month. That number is a problem, but that number was a problem before the introduction of LLMs. AI doesn’t even seem to increase cheating.

So what are teachers worried about?

Key Takeaways

• The fears about AI ruining education by creating more cheaters has largely been exaggerated.
• Teach students how you want them to use AI. Be clear and firm in what is not acceptable.
• Think creatively about assessments as a workaround to prevent cheating before it starts. You can have students write assignments by hand, produce a product, or participate in a lab or workshop.