Haley L.
@haleyhaala.bsky.social
86 followers 290 following 26 posts
phd candidate at stanford
Posts Media Videos Starter Packs
pay no attention to the companies beyond the ai (who are quickly amassing non-democratic control of public school functions)
charleswlogan.bsky.social
Check out the latest Mystery AI Hype 3000 newsletter for a list of episodes focused on education.

You'll find episodes featuring yours truly, Adrienne Williams, @hypervisible.blacksky.app, and @haleyhaala.bsky.social.

🧵
emilymbender.bsky.social
The latest on the Mystery AI Hype 3000 newsletter: Back to School is Better Without AI

buttondown.com/maiht3k/arch...

w/ @alexhanna.bsky.social and @deccamuldowney.bsky.social
At #4S2025 questioning all things "modern" including the earl grey

ps- come to our sessions on Saturday about language technologies!
Academic books are notorious for boring endings, but @emilymbender.bsky.social and @alexhanna.bsky.social hit it out of the park! The last chapter of #TheAICon is full of actionable (pro-labor!) steps for standing up to Big Tech's hype.
Fight silliness with silliness!
ridicule as praxis circled in book
Moreover, in elite higher ed, student unions may be the first (and hopefully not only) experience the children of capitalists have to learn about labor rights. Student unions are important pedagogical praxis.
A reminder to @aaup.org faculty to be allies to your grad student worker unions! It has been devastating to see @stanford.edu "justice-oriented" aaup faculty and deans complaining about and belittling union movements on campus.
sbagen.bsky.social
I'm proud that @aaup.org has been on the front lines fighting this administration's abuses. Trump is, indeed, an aggressively anti-labor president, despite the "workers party now" bs.

www.nytimes.com/2025/09/01/o...
Opinion | Trump Is Wiping Out Unions. Why Are They So Quiet?
www.nytimes.com
I am so excited to host a panel with @ajalvero.bsky.social at #4S2025! If you're around next Saturday, join us for discipline-blending sessions and conversation under the theme of "Writing Things Down: Textual Technologies and their Consequences." @4sweb.bsky.social
But differential performance by skin color is not something @flysfo.bsky.social should tolerate, and is one more reason to keep automatic face recognition out of public services...
To be fair, I'm not sure if this is bad or good- not being recognized may be a protective factor, especially under our current federal administration. See great work from @stanfordnlp.bsky.social researchers about
#AI surveillance systems www.nature.com/articles/s41...
Computer-vision research powers surveillance technology - Nature
An analysis of research papers and citing patents indicates the extensive ties between computer-vision research and surveillance.
www.nature.com
So why is #TSA @flysfo.bsky.social using software that works differentially based on skin color? If the feds won't deal with it, can California's gov @gavinnewsom.bsky.social pursue this as a civil rights issue?
When it was my turn I asked the agents if they saw that software worked better on some people rather than others.
Two agents told me it tends not to work on dark skin.
Now, the failure of facial recognition on dark skin has been a known problem for YEARS as documented by @poetofcode.bsky.social.
An #AI story: While going through #TSA at @flysfo.bsky.social I noticed that the facial recognition was not working on the Black man in front of me. The agent asked him again and again to get closer.
We wrote about this at the grad/faculty level in scientific publishing. As a raciolinguistic perspective would suggest, academics find new ways to racialize language even after people use AI to "sound" like "native speakers" dl.acm.org/doi/10.1145/...
“You Cannot Sound Like GPT
dl.acm.org
Reposted by Haley L.
abeba.bsky.social
New paper hot off the press www.nature.com/articles/s41...

We analysed over 40,000 computer vision papers from CVPR (the longest standing CV conf) & associated patents tracing pathways from research to application. We found that 90% of papers & 86% of downstream patents power surveillance

1/
Computer-vision research powers surveillance technology - Nature
An analysis of research papers and citing patents indicates the extensive ties between computer-vision research and surveillance.
www.nature.com
Reposted by Haley L.
stanfordeducation.bsky.social
Professor Thomas Dee analyzed attendance data from five school districts in the southern Central Valley and found a spike in daily absences this past January and February. These districts, which serve over 100,000 students, saw considerable jumps in all age groups. stanford.io/3HVr6Hb

#EduSky
Reposted by Haley L.
thanks Shaily!!
shaily99.bsky.social
Go checkout Haley's paper that won best paper!!
On language ideologies in peer-review and use of LLMs for "polishing" writing. Important work, and such a good read!

"You Cannot Sound Like GPT": Signs of language discrimination and resistance in computer science publishing"
arxiv.org/abs/2505.08127
Flattered and shocked for our paper to receive the #facct2025 best paper award.
facct.bsky.social
🏆 Announcing the #FAccT2025 best paper awards! 🏆

Congratulations to all the authors of the three best papers and three honorable mention papers.

Be sure to check out their presentations at the conference next week!

facct-blog.github.io/2025-06-20/b...
Announcing Best Paper Awards
The Best Paper Award Committee was chaired this year by Alex Chouldechova and included six Area Chairs. The committee selected three papers for the Best Paper Award and recognized three additional pap...
facct-blog.github.io
Brilliant presentation by @sepehrvakil.bsky.social and Mahdi Ganjavi on their new @mitpress.bsky.social book about engineering activism during the Iranian revolution. They close with this picture and caption: "faculty resists against state control over the campus"
faculty resists against state control over the campus
Thanks to @stanfordhai.bsky.social for supporting this project!
Our findings offer insight into the role of #ChatGPT in the reproduction of language ideologies which conflate producers of “good English" with producers of “good research." Future work should look into this effect in other writing scenarios, such as #education and #work.
Authors, aware of this development, described removing ChatGPTisms and other clues which could expose their “non-native” status to reviewers.
Though ChatGPT may mask some signs of language background, reviewers explain that they now use ChatGPT “style” and non-linguistic features as indicators of author demographics.