cfiesler.bsky.social
information science professor (tech ethics + internet stuff)
kind of a content creator (elsewhere also @professorcasey)
though not influencing anyone to do anything except maybe learn things
she/her
more: casey.prof
767 posts
14,477 followers
325 following
Regular Contributor
Active Commenter
comment in response to
post
hahaha exactly
comment in response to
post
haha I didn’t even notice that! At least it doesn’t show up in slide show view. :)
comment in response to
post
Thank you so much!!
I don’t know if you were there for the Q&A but at some point I mentioned small online communities and was thinking in part of that paper of yours. :)
comment in response to
post
This was so thoughtful!! I’ve also always wanted to do an internet nostalgia kind of study. I’ve gotten some of that from the fandom migration work but so much has changed since 2018…
comment in response to
post
Ah I always appreciate a good “live tweet” of pieces of a talk! (No worries about the typo. 💙 ) bsky.app/profile/emma...
comment in response to
post
hahaha I was just going to ignore the comment but this might be the correct response
comment in response to
post
And I do think I’ve accomplished my goal of my YouTube videos making their way to a new audience.
comment in response to
post
And I don’t just mean people unreasonably hyping up AI (though lots of folks decrying this research and talking about super intelligence) but also discounting the tech in ways that suggest being misinformed about how it works (e.g. “can’t answer a question no one has asked before”).
comment in response to
post
I think there will inherently be a difference between an obviously temporary technical glitch and thinking something has been shut down for good :)
comment in response to
post
Sure thing, adapt however you like!
comment in response to
post
Anyway, I really hope this will be helpful for folks, especially teachers! I believe a lot in the "accidental learning" aspect of my unexpected appearance on someone's social media feed (and I think I've been successful at that!) but of course lots of folks seek out learning too. :)
comment in response to
post
And of course most of this content originated on TikTok, which isn't exactly a good archive for lots of reasons. So I'm also in the process (slowly) of uploading old videos to my YouTube channel. So as I continue to populate this syllabus you'll see that some but not all videos have YT links!
comment in response to
post
I've been working on this in the wake of my NSF grant termination; losing that funding meant that I likely can't continue the project where teams of college students would have created educational content about AI for social media. So I thought... well at least I can make MY content more accessible.
comment in response to
post
And felt obligated to share!
comment in response to
post
It would have made the article at least 50% more millennial cringe haha!
comment in response to
post
guess I’ll make a TikTok 😂
comment in response to
post
Nope! Though if there aren’t errors that represent carelessness then I don’t see why it would matter.
comment in response to
post
I’m about to read that for a book club, suspect that will give me pause too…
comment in response to
post
Sorry I just keep thinking of more to say on this topic! Is generative AI going to provide some version of the autonomy paradox like blackberries and zoom? “Oh it’s so great that AI can do some work for me! … but also I have to do twice as much work.”
comment in response to
post
It's actually even worse than that though - technically they were only allowed to cite papers we read in class or from a list I gave them and there were STILL fabricated citations. 🤦
comment in response to
post
This is an undergraduate class and a certain number of scholarly citations were required.
comment in response to
post
My department has. :) (I have very strong feelings about the GRE heh.)
comment in response to
post
🥰
comment in response to
post
Thanks again for this reminder! Someone else asked me about this exact thing so I made a follow up. :)
comment in response to
post
Also was reminded of the parallels between not only generative AI and mechanical looms, but also generative AI and washing machines. When does “labor saving” technology actually reduce work as opposed to just redistributing it or increasing expectations?
comment in response to
post
Well there's "ok" in that I probably wouldn't know they did that.
However, if they say "I really enjoyed your paper X" but they never actually read that paper, then I consider that not OK ethically. But of course, I would not know.
comment in response to
post
Very high.
comment in response to
post
I suspect that even faculty who don't mind students using LLMs would be horrified about them not checking the output though.
comment in response to
post
nah it was supposedly published last year
comment in response to
post
The same thing happened in my class this semester. Not only did I get assignments turned in with fabricated citations, but with fabricated citations with me as the author...
comment in response to
post
I wonder if I need to do an updated version of this video literally just to talk about this issue. youtu.be/XPM22bWTjRk
comment in response to
post
Definitely worse. At least getting a real paper title involves a google scholar search or something!
comment in response to
post
In response ("if I may also offer a piece of advice") I gently pointed out that if you're interested in computing fields you should probably be aware of the limitations of LLMs, and also that if you're not familiar with a professor's work they are likely not a good fit for your research interests.
comment in response to
post
Yep.
comment in response to
post
Nah apparently the fictional paper was published at CHI 2024. :)
comment in response to
post
It was not particularly creative, honestly. I've seen LLMs produce far better hallucinations for me. :)