A safe and supportive forum for philosophers.
Owned & Moderated by Marcus Arvan (University of Tampa)
Contact: marvan@ut.edu
Blog mission & moderation policy
In our new "how can we help you?" thread, a reader asks:
What AI tools have you found most useful for doing philosophy? What philosophy-related tasks have you found them most useful for? (I am asking because I use very few AI tools but want to start using more.)
Do any readers have any helpful tips?
I’ve been reluctant to use AI for fears of “authenticity” and whatnot, but I have to admit I’ve found ChatGPT helpful for tedious work that doesn’t require intellectual creativity. For example, formatting citations. Putting the source info into ChatGPT and having it match properly to the Chicago Manual of Style has saved lots of time.
For teaching, I’ll also ask ChatGPT questions about my quiz questions e.g. “do you think students will have enough time in a 75-minute period to read and answer these?” or “is this question worded ambiguously? Suggest improvements.”
I have found https://elicit.com/ to be a useful AI resource for locating articles. Good for some generic lit review kinds of things too.
I had a workshop on the book manuscript that I’m finishing up. I recorded each session, and then I used an AI app (I used Aiko, but there are several) to turn the recordings into PDF transcripts. It did a great job; the only thing it really messed up was proper names, and it was usually able to tell when different people were speaking and put line breaks between each. It could work for conference q&as too (depending on the kind of conference it is).
A relevant, recent discussion from Daily Nous:
https://dailynous.com/2025/02/05/professors-how-are-you-using-ai-in-your-day-to-day-work/
I use Claude Opus to help revise sentences and paragraphs
I find any leading LLM helpful for brainstorming.
If I’m reading difficult math, I find they’re decent at helping me to break into the text — they make some mistakes, but so do I, and it’s nice to have someone to talk through examples or confusions with.
They’re immensely helpful with tikz diagrams and other fussy bits of LaTeX coding, and I’ve used them to do some coding in other languages (e.g. netlogo for agent based modeling) as well. Good results for requests like `how do I say ___’ but not great results for requests to write long complicated programs.
I have a fairly strict rule on ChatGPT. I will use it for robustness checks, but not to initially generate any substantive argument.
So, in other words, I use it to run an audit on mathematical or formal arguments. I feel fairly comfortable with this as, right now, ChatGPT seems like a much better editor than writer.
Doing this doesn’t guarantee that my math or formal arguments are correct, but it’s a good way to make sure I didn’t miss something basic that reviewers will spot.
I have utilized perplexity becuase whatever information it gives you, it provides direct citation for its sources so that you can create your own paper trail. Very useful.
Leave a Reply