A Double-Edged Sword: LLMs, Writing, and You

Speaker: Graham Oliver, Adjunct Instructor, AWEC
Time and Date: 5/8 (Thursday) 13:40-15:10
Venue: Room 501, Zonghe Lecture Building
Sign Up: 活動報名系統

Speaker’s Introdction:
Graham Oliver is a teacher, writer, and editor living in Taipei. Currently, he serves as an Adjunct Instructor at the Academic Writing and Education Center of National Taiwan University, where he’s taught multiple LLM-related courses. His writing has appeared in Harvard Educational Review, Commonwealth Magazine/天下雜誌, Guernica, and elsewhere. His areas of interest include the teaching of writing, literature in translation, and video game narratives. He holds an MA in Rhetoric and Composition as well as an MFA in Creative Writing from Texas State University.

Synopsis (conducted in English) :
Come learn how to stand out and avoid the LLMs’ many pitfalls. LLMs are a writing tool like any other: you need to know how, and when, to use them. Join us to be a step closer to that goal.

Outline – LLMs are everywhere. Almost every person reading this has used ChatGPT, Gemini, Claude, Perplexity, or DeepSeek to edit their work, generate code, act as a therapist, give feedback on an idea, or create entire homework assignments. LLMs have a lot of benefits when it comes to speed, efficiency, and following general writing conventions. But as you also know, LLMs come with huge drawbacks. There are ethical considerations, information hallucinations, poor writing quality, and more. On top of that, studies have recently come out showing longterm effects in the minds of people who use them. There’s a lot to worry about!

In this lecture, we’ll discuss the pitfalls of using LLMs, and how you can adapt to minimize their damage. This includes thinking about which situations LLMs work best and which they don’t, how to let LLMs be a good editor for you but also how to be a good editor for them, and what to watch out for each time you type in a prompt. In the end, LLMs are one of many tools you can use to help your writing. They’re especially great for working in your non-native language. But like any other tool, it doesn’t fit every situation, and you have to learn to use it before you can use it well. After this lecture, you’ll be a big step closer to that goal.