Blog: Discussion on AI Assistance in Academic Writing
During this week's group meeting, I2SC group members shared specific ways they have used LLMs to improve their writing and discussed best practices for incorporating LLMs into a writing workflow.
As a teaser to the discussion, Jianlong prepared a "Journalist or AI" game in which group members were invited to attribute authorship to pairs of news-like paragraphs. We noticed that:
- LLM-generated text tends to have a perfect sentence structure but is often verbose.
- Human-written text may break secondary-school conventions but is often effective in succinctly communicating ideas and data.
- Ingmar is very good at this game.
Consensus reached by participants included:
- Never using LLMs to create an initial draft of a paper, because *you* should be the one to decide the overall structure and direction of the writing.
- LLM-generated feedback for your writing can be helpful, but it tends to be overly positive--even if you specify that you are aiming for a competitive venue where most papers are rejected. Be sure to ask for critical feedback.
- LLMs can also be a great tool for: brainstorming, refining word choices, and grammar checks.
Other interesting ideas that emerged:
- Academic prompt repository: group members were encouraged to share effective prompts and models for generating critical and constructive paper feedback.
- The DeepSeek shock: the reasoning steps from more recent "thinking" models, such as DeepSeek-R1, can expose us to new ways of approaching a research question and identify relevant literature
- Switching back to Grammarly: even though LLMs can potentially add flavor and create a more "professional" appearance, some actively reject stylistic suggestions from these models. The pre-LLM era grammar checker was seen as a preferable tool to ensure correctness while maintaining an original voice.
(This post was written by Jianlong without AI assistance.)
