Just like most other students I use Generative AI tools. Most of my experience lies in using ChatGPT, but I have certain rules I live by when using AI tools for university related activities. First of all, I fit the Dutch stereotype of being stingy, or at least for monthly subscriptions. I do not want to be bound by monthly payments, neither for AI tools nor for services like Spotify. So, my options are limited to the free versions. Secondly, I want to be able to do everything myself, so I only use it as support. Finally, I have some trust issues with AIs, particularly ChatGPT.
During my bachelor I did not use AI much, even though I come from an ICT & Business background, were students and teachers picked it up pretty quickly. So, most of my experimental phase took place last year. However, my experiments made me more reluctant to use AI, specifically ChatGPT. After experimenting I found the following issues:
- ChatGPT often does not catch the real meaning of your input, while other humans would be able to do it.
- ChatGPT makes stuff up. I remember thinking that I could maybe use ChatGPT to find something in a document I uploaded, but the results it gave were not mentioned in the document. Especially quotes, it made them up left and right. It answered what its thought I was looking for even if that was not the truth.
- During my interactions with experts and companies I remember often being told to be very careful with the information you put in ChatGPT as it can share that information. For example, if you share your ground-breaking idea with ChatGPT, it can give your idea to someone else that is asking for inspiration etc. Similarly, you have to be careful with personal and sensitive information.
However, in the current study environment, I often feel at a disadvantage for not using generative AI tools often. Even though I do not know how many people actually use ChatGPT or similar tools, I feel like I spend unnecessary time on doing all the preparation by myself. I am afraid that the workload is going to grow increase as more and more students are going to use AI tools to save time. If that happens, it is going to come down to who wants to spend most money on AI tools, which I think is not fair.
On the positive note, I do see great opportunities for AI for students with learning disabilities, like dyslexia and AD(H)D. They often struggle with things like taking notes during lectures and writing. And while I am still a bit conflicted on the writing part, since I am of the opinion (as someone with a learning ability) that we still have to be able to do it ourselves, AIs for note taking seem very useful to me. Sadly, the tools I found are too limited with their free versions to be useful or do not have a free version at all.
I really liked your blog post! I can relate to what you’re saying about feeling like you might fall behind if you don’t use AI, because more and more people are relying on it. Sometimes it even feels like there’s this unspoken expectation that we should be using it.
Your point made me think back to our discussions about platforms and freemium models in class: If you pay for ChatGPT, you get all the premium features, while the free version is more limited. That creates a kind of inequality between students.
I also thought of the “black box” concept we talked about. You mentioned not being able to trust what ChatGPT gives you because it sometimes makes things up or misinterprets inputs. The algorithms are so complex that it’s basically impossible to trace how your input leads to a certain output. And why Chatgpt gave this output.