The impact of over-reliance on AI 

By Phuong Thao Tran (S00389283)

 Throughout this week, I’ve been putting my critical digital literacy skills to the test by searching for information about an internet safety issue. I’ve been looking at human overreliance on AI, basically this points out how AI in the future will assist and benefit people or will they detrimentally impact if humans over rely on AI. I will apply the TRAAP method, which stands for Timeliness, Relevance, Authority, Accuracy and Purpose to assist my writing to be more credential.

    AI, which stands for Artificial Intelligent, refers to the ability of computer systems to perform tasks which normally require human intelligence. It is a very broad field that encompasses theories and developments in computer science, cognitive science, mathematics, philosophy, neuroscience, linguistics and more. (Lumsden, M. (2024, March 20). How do I guides: Artificial Intelligence: Introduction - What is AI?)

    However, many people, mostly students in school, tend to use AI to mischievously finish their work off without using their abilities to proceed on the work on their own. At a certain point, this is considered as over abusing on Artificial Intelligence. Over reliance on AI refers to an excessive dependence on artificial intelligence systems to make decisions, solve problems, or manage tasks without appropriate oversight or human judgement. (Zhai, C., Wibowo, S., & Li, L. D. (2024). The effects of over-reliance on AI dialogue systems on students’ cognitive abilities: a systematic review. Smart Learning Environments)

    Over-relying on AI is a massive problem. I have found a source from the University of Stanford that will give us a clearer view for how AI is deviating people from their capabilities of assessing and addressing problems. According to Helena Vasconcelos (2023), an Undergraduate from University of Stanford majoring in Symbolic Systems, a handful of people tend to trust AI's recommendations without going through it logically, even when they give out false information. Vasconcelos also conducted several experiments to test out her thesis, one of which has caught my attention and made me think deeply about how people are over-relying on artificial intelligence. Here, she set up an experiment where she asked people work together with an AI on a reading comprehension task. The AI first summarised a passage and then give out explanatory for that summary. The human collaborators had two options for evaluating the AI's performance: they could either complete the reading comprehension task on their own or review the AI's summary and explanation. Even though both options were quite relatively as challenging, most of the collaborators chose to choose AI's work instead of going through it on their own.

    This is quite surprising thus a sad reality because people are slowly depending on AI without ethically looking out for legitimacy, and this discovery by Helena Vasconcelos has only been around for a couple of years, and with this pace of modern technology era, people may have difficulties in analysing and evaluating on their own in the near future.

    This problem is emerging, and it is heading towards the younger generations. Without the right guidance, people will have a misconception over AI's value towards humanity. AI is a tool that assist people throughout tasks under certain conditions, not something that can single-handedly replace human's work to succeed on its own.

    Of course, there are counter tactics this issues. Personally, I think people should try to do researches and take notes and make out key points of the topics on their own when they are facing the issues. Other than that, we should all raise our awareness on self-dependence when working without the help of AI tools.

    Here's the link to the website from University of Stanford that I used to support my work, if you readers want to check it out https://hai.stanford.edu/news/ai-overreliance-problem-are-explanations-solution (Katherine, M. (March 2023). AI Overreliance Is a Problem. Are Explanations a Solution?)

How did I use TRAAP for my blog?

T- Timeliness - When was this source published?

R- Relevance - Does this source meet my information need?

A- Authority - is the author of this source an expert? how do you know?

A- Accuracy - is this information accurate? does it match what I found in other sources

P- Purpose- Why was this written? 

    The article that I specifically picked is recent- March 2023 and it is relevant to my topic of AI over-reliance. This text was given out from the University of Stanford, hence it meets the expectations of authority. Since the released date and the author is from a big source and is newly updated, it is accurate for the current time. Other relevant information from other websites are also recent. I chose the website from Stanford University specifically because it gives out clear reasons that will support my blog, also they have their own experiments from the University's researches, so they give out an iconic example to my topic.

Thank you for digging in my blog. Bye for now.