Gen AI in the Classroom: The Good, the Bad and the Confusing

By: Sarah Roberts
students with computers in a lecture hall
Filters

When the buzz surrounding generative AI tools like ChatGPT started dominating public discourse in late 2022, it hit higher education institutions like a ton of bricks. Seemingly out of nowhere, students had a resource that could do much of their coursework without professors even being able to differentiate between authentic and machine-generated assignments. One year later, many educators are still struggling to understand how monumental gen AI will be for their courses and students.  

However, one clear truth is that this technology is here, and the only reasonable approach is to adapt to it. While a few universities are, recent studies found that just 3% have policies surrounding students’ use of this tech. For anyone instructing students, understanding the benefits, the risks and the unknowns of gen AI is becoming a necessary part of the job. To help them get started, here are the good, the bad and the confusing implications of gen AI in today’s university classrooms. 

The Good: A tutor that can level the playing field and improve accessibility 

According to research, 80% of students who receive one-on-one tutoring outperform their peers who don’t have this form of support. Additionally, tutoring improves graduation rates for university students. However, the cost associated with private tutoring in higher ed averages at around $75-$100 an hour. Those costs are prohibitive for many students, making the benefits of this form of support limited to only wealthy undergrads.  

Given the high costs, it should come as no surprise that many educators are excited about the possibility of affordable or even free gen AI options for providing individualized instruction that can mimic that of a more costly private tutor. Such tools can help students learn more and could be particularly useful for those who speak English as a second language or have learning disabilities.  

Verbit recently released its own gen AI offering, Gen.V, for students to summarize transcripts and pull important information that streamlines studying and learning. These applications of the technology showcase some of the ways that gen AI can benefit students and help them learn more effectively. Still, even those who are optimistic about the potential to improve learning with tech realize that there are plenty of drawbacks as well. 

a student working in a library

The Bad: A tool that hinders growth and learning 

It’s not an overstatement to say that gen AI sent many educators into a panic. The temptation to use ChatGPT to write an essay or answer questions will lead many students to take the easy way out. Furthermore, over half of students report using these tools compared to fewer than 25% of instructors. Students are, therefore, becoming far more familiar with the technology, its usefulness and its capabilities than their professors. Instructors are left wondering if a student completed an assignment and learned the course material or simply punched a few prompts into ChatGPT. 

One possible way to combat this is to use another AI-powered tool to check whether an essay was written by a human or a bot. While recent tools show better results in identifying AI-generated content, trying to fully prohibit the use of such readily available technology will inevitably create a bit of an arms race – with students finding more subtle ways to use gen AI shortcuts and professors searching for better AI-detecting tools.  

The problem won’t just impact courses that involve writing essays. Gen AI can also write code, which means computer science instructors also need to face the possibility of their students using bots to do all the heavy lifting on their assignments. In these cases, AI detectors are even less reliable. As a result, instructors concerned about the use of the tools may need to determine whether the code includes techniques the students haven’t learned yet – a strategy that’s potentially even more problematic. After all, it could penalize a student who learned more advanced coding in an earlier course. 

Another issue is that AI can sound convincing while turning out completely fabricated information. So-called “hallucinations” can lead students astray. If a student is using the tool to study, they might learn incorrect information that later leads to poor test scores.  

While some instructors are looking for ways to prevent the use of gen AI in their courses, another more complicated question arises. Now that the tools exist, will prohibiting their use ultimately help or harm students?  

two students sitting back to back working on the floor

The Confusing: A disruptor that could render some disciplines obsolete 

The goal of universities is to teach students. If instructors are successful, students will leave with skills and knowledge to help them navigate their future – especially their careers. When it comes to preparing students for the workplaces of tomorrow, understanding how to use gen AI will be imperative. Whether instructors like it or not, some of the skills they’re learning now might not be anymore.  

However, it’s not entirely clear that learning to turn out an essay or code by prompting ChatGPT will make students effective professionals. Can a person who doesn’t know how to write a decent essay truly review the output from a chatbot and say whether it’s good enough? If a person relies solely on one of these tools to write code, will they ever be able to troubleshoot if something goes wrong?  

As universities work to adapt their policies to address the use of gen AI for completing work in their existing courses, they also must determine which courses might need to be retired or altered and what new courses they should add to their curriculum. Any institution that plans to help its students enter the workforce must find ways to teach them how to use this new technology effectively to be more productive, not just to do the existing tasks in the quickest and easiest way. To achieve this, universities must be creative, adaptable and strategic. Luckily, they should have plenty of brainpower on hand to help them navigate the “fourth industrial revolution.” They just need to make use of it.  

Connecting with forward-thinking partners 

The evolving technological landscape is exciting, but it will also be a challenge to navigate. Fortunately, universities and other higher education institutions don’t need to travel through this uncharted territory alone. Leaders in education can find guidance and inspiration by forging partnerships with organizations working to make the most out of emerging technology and harness its power to improve educational opportunities for all. Reach out to Verbit to learn more about our gen AI offerings and other AI-powered accessibility solutions, including captions, transcripts and more.