This week on our blog we have been asked to reflect on AI, it’s pros and cons, and uses in the classroom. Generative AI (GenAI) is mainly comprised of Large Language Models, which take mass amounts of text as it’s training data and produces what is effectively extremely advanced “auto complete.” Some GenAI tools have gone further and are able to produce image or video, however in this post I will be focusing on Today I will be addressing the prompts: What are some of the major limitations of GenAI? and Talk about how you’ve found GenAI useful or not for educational purposes.
What are some of the major limitations of GenAI?
GenAI is one of the major revolutions of our time, along with the communications revolution of the 2010s, it may be the most disruptive outcome of technology to everyday life. Since the beginning of its widespread use in 2021, the use of AI in everything from education to art has been utterly transformative for society writ large, specifically widely publicly available models such as ChatGPT or Gemini. The limitations of GenAI are clear, tools that seem intelligent offer students another way to find information, but the information that they receive can sometimes be wrong. Hallucinations, oversimplification, and questions of accuracy are all huge limitations to GenAI. There is major risk for students of becoming overly reliant on AI and then having it give you information that is simply false. Additionally, students may outsource much more of their thinking to GenAI. With something so easy and quick, it is more of a draw than ever to simply get AI to do your assignment for you, or tell you how to do it instead of doing it yourself. As a result, student’s abilities to evaluate information, construct effective arguments, or even fundamentally reason can be shaken or destroyed before they get a chance to grow. Lastly, the ethical implications are a huge limitation on GenAI. Environmental costs are real, and the ownership of ideas legally can become a huge problem for people who are using AI creatively. Additionally, this does not address where the training data for the AI came from, and whether or not it was available to the public.
The following are videos that I believe are relevant to this topic but are quite long.
My Experience Using GenAI for Educational Purposes
Here, I have to flip the script. As much as I harp on GenAI, and have legitimate grievances with it, I have also used it in my education so far, and plan to use it again while teaching. I have used GenAI to help me correct grammar, spelling, and punctuation, as I am often not the best when it comes to this kind of type editing. GenAI in my education has been much like a fancy word processor, while I use it to quickly help me format or edit my work, I refuse to turn to GenAI instead of thinking of my own ideas. For this assignment even, I have used it to help me format, giving it my ideas and getting it to suggest where in the blog post I should put things. But, I also only use its suggestions a fraction of the time as having it around helps me to be more confident in my own work, in comparison.
In the classroom I believe that I can take this a step further, AI is a fantastic tool and I believe that using AI to make some of the menial tasks of teaching is not just a benefit, but necessary to keep up with increased workloads. Teaching has long since been a job where you are expected to work hours well above the 40 per week that is allotted to you for pay. I am firmly against AI teaching students, but if there is any moral use for it, I think that using it to enhance student learning by allowing teachers to create quick, individualized, and effective assessments is as close as we can get.
