Her final essay was on Claudia Rankine's Citizen & the white gaze. She turned in two versions. Her original essay cited Fanon's Black Skin White Masks to great effect. The AI version cleaned up some of her grammatical errors, but also deleted every mention of Fanon from her work!
Comments
Wow, wow, wow.
I mean, I come to the AI conversation as an anti-AI curmudgeon, but this.....
Wow.
- Aime Cesaire (or similarly cited Black male anticolonial thinker)
- James Baldwin (or other commonly cited Black male writer)
- Black women writers
Paging @ruha9.bsky.social & @alondra.bsky.social!!
In my experience the only way to prove this kind of thing is to meet with my students 1-on-1 and ask about it. Sometimes they choose not to sustain a complete lie for very long.
I’ve seen a few different people test this out by asking if Claude knows who Fanon is (it does, and it can recite his bio)
<= you can ask GPT-4 about the relationship between Fanon and Rankine
DH scholars: is there a way to go about this?
They don't always agree in the reports and are still meeting to talk about ideas. Leo Flores & Liz Losh are on this task force, which bodes well
1/
:)
Also, Reddit is pretty mainstream these days.
I want to know if it is smart enough to adapt and not be as bias as the algorithm or does it reset?
But that’s where we are and where we will continue to be
Better ask how we can create “quiet”, adaptive, monitoring in perpetuity
I can 100% guarantee that Claude will happily discuss Fanon at length: https://pastebin.com/sg06KJNP
sometimes it absolutely is.
but often times, that's just the beginning that kicks it off… it just becomes "how things work" after
We are all blessed with it to a degree.
I challenge my beliefs and worldview to be more accurate.
I’ve thought about how to dispel collective illusions recently.