Rebecca Dorn
@rebeccadorn.bsky.social
PhD candidate at USC. Studying inequitable social systems in NLP, social media and content moderation. They/She
http://rebedorn.github.io
http://rebedorn.github.io
Reposted by Rebecca Dorn
I've spent the past year obsessed with how police are using facial recognition. What I found: Police in 12 states have used these tools to find and arrest suspects when they have no other evidence, leading to wrongful arrests
wapo.st/4fQYlXn
🧵 what I learned from 1,000 docs & dozens of sources ->
wapo.st/4fQYlXn
🧵 what I learned from 1,000 docs & dozens of sources ->
Arrested by AI: Police ignore standards after facial recognition matches
Confident in unproven facial recognition technology, sometimes investigators skip steps; at least eight Americans have been wrongfully arrested.
wapo.st
January 14, 2025 at 2:27 AM
I've spent the past year obsessed with how police are using facial recognition. What I found: Police in 12 states have used these tools to find and arrest suspects when they have no other evidence, leading to wrongful arrests
wapo.st/4fQYlXn
🧵 what I learned from 1,000 docs & dozens of sources ->
wapo.st/4fQYlXn
🧵 what I learned from 1,000 docs & dozens of sources ->
Data imitates life imitates data 🤯
December 25, 2024 at 4:35 PM
Data imitates life imitates data 🤯
My recent work on gender-queer dialect bias in content moderation is referenced in this CDT analysis!!!!!!!!!!! (I’m not crying, you’re crying)
New blog from @amariezaner.bsky.social, CDT’s Summer Research Extern, analyzes the intersectional disparities within automated hate-speech detection across US-centered social media content. cdt.org/insights/soc...
Intersectional Disparities within Automated Hate-speech Detection Across US Centered Social Media Content Social Media Content
This blog was authored by Amanda Zaner, Summer Extern for CDT Research. Navigating the socio-linguistic nuances of communication across social media platforms is no simple task. When it comes to online content moderation, platforms frequently rely on automated detection systems for the challenge of identifying and addressing abusive language posted in these digital environments. Recent […]
cdt.org
December 17, 2024 at 10:25 PM
My recent work on gender-queer dialect bias in content moderation is referenced in this CDT analysis!!!!!!!!!!! (I’m not crying, you’re crying)
Reposted by Rebecca Dorn
📢 NEW Paper!
@siree.sh, Lucy Suchman, and I examine a corpus of 7,000 US Military grant solicitations to ask what the world’s largest military wants with to do with AI, by looking at what it seeks to fund. #STS
📄: arxiv.org/pdf/2411.17840
We find…
@siree.sh, Lucy Suchman, and I examine a corpus of 7,000 US Military grant solicitations to ask what the world’s largest military wants with to do with AI, by looking at what it seeks to fund. #STS
📄: arxiv.org/pdf/2411.17840
We find…
December 9, 2024 at 2:18 PM
📢 NEW Paper!
@siree.sh, Lucy Suchman, and I examine a corpus of 7,000 US Military grant solicitations to ask what the world’s largest military wants with to do with AI, by looking at what it seeks to fund. #STS
📄: arxiv.org/pdf/2411.17840
We find…
@siree.sh, Lucy Suchman, and I examine a corpus of 7,000 US Military grant solicitations to ask what the world’s largest military wants with to do with AI, by looking at what it seeks to fund. #STS
📄: arxiv.org/pdf/2411.17840
We find…
Reposted by Rebecca Dorn
Our Open Source Developers Guide to the EU AI Act is now live! Check it out for an introduction to the AI Act and useful tools that may help prepare for compliance, with a focus on open source. Amazing to work with @frimelle.bsky.social and @yjernite.bsky.social on this!
Open Source Developers Guide to the EU AI Act
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
December 2, 2024 at 5:06 PM
Our Open Source Developers Guide to the EU AI Act is now live! Check it out for an introduction to the AI Act and useful tools that may help prepare for compliance, with a focus on open source. Amazing to work with @frimelle.bsky.social and @yjernite.bsky.social on this!
Reposted by Rebecca Dorn
Hi, so I've spent the past almost-decade studying research uses of public social media data, like e.g. ML researchers using content from Twitter, Reddit, and Mastodon.
Anyway, buckle up this is about to be a VERY long thread with lots of thoughts and links to papers. 🧵
Anyway, buckle up this is about to be a VERY long thread with lots of thoughts and links to papers. 🧵
First dataset for the new @huggingface.bsky.social @bsky.app community organisation: one-million-bluesky-posts 🦋
📊 1M public posts from Bluesky's firehose API
🔍 Includes text, metadata, and language predictions
🔬 Perfect to experiment with using ML for Bluesky 🤗
huggingface.co/datasets/blu...
📊 1M public posts from Bluesky's firehose API
🔍 Includes text, metadata, and language predictions
🔬 Perfect to experiment with using ML for Bluesky 🤗
huggingface.co/datasets/blu...
bluesky-community/one-million-bluesky-posts · Datasets at Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
November 27, 2024 at 3:31 PM
Hi, so I've spent the past almost-decade studying research uses of public social media data, like e.g. ML researchers using content from Twitter, Reddit, and Mastodon.
Anyway, buckle up this is about to be a VERY long thread with lots of thoughts and links to papers. 🧵
Anyway, buckle up this is about to be a VERY long thread with lots of thoughts and links to papers. 🧵
Reposted by Rebecca Dorn
Oh my god. I had no idea:
Medicaid: 73 million low-income people are exposed to AI-related decision-making
SNAP: 42 million low-income people are exposed to AI-related decision-making
Social Security disability benefits: About 13.8 million people are exposed to AI-related decision-making
Medicaid: 73 million low-income people are exposed to AI-related decision-making
SNAP: 42 million low-income people are exposed to AI-related decision-making
Social Security disability benefits: About 13.8 million people are exposed to AI-related decision-making
"92 million low-income people in the U.S. states—everyone whose income is less than 200 percent of the federal poverty line—have some basic aspect of their lives decided by AI"
www.techtonicjustice.org/reports/ines...
this is a damning report
www.techtonicjustice.org/reports/ines...
this is a damning report
November 23, 2024 at 6:42 PM
Oh my god. I had no idea:
Medicaid: 73 million low-income people are exposed to AI-related decision-making
SNAP: 42 million low-income people are exposed to AI-related decision-making
Social Security disability benefits: About 13.8 million people are exposed to AI-related decision-making
Medicaid: 73 million low-income people are exposed to AI-related decision-making
SNAP: 42 million low-income people are exposed to AI-related decision-making
Social Security disability benefits: About 13.8 million people are exposed to AI-related decision-making
We come to this place… for magic.
We come to SoCalNLP theaters to laugh, to cry, to care.
Because we need that, all of us,
that indescribable feeling we get when the lights begin to dim.
And we go somewhere we've never been before;
not just entertained, but somehow reborn.
Together.
We come to SoCalNLP theaters to laugh, to cry, to care.
Because we need that, all of us,
that indescribable feeling we get when the lights begin to dim.
And we go somewhere we've never been before;
not just entertained, but somehow reborn.
Together.
November 23, 2024 at 4:57 AM
We come to this place… for magic.
We come to SoCalNLP theaters to laugh, to cry, to care.
Because we need that, all of us,
that indescribable feeling we get when the lights begin to dim.
And we go somewhere we've never been before;
not just entertained, but somehow reborn.
Together.
We come to SoCalNLP theaters to laugh, to cry, to care.
Because we need that, all of us,
that indescribable feeling we get when the lights begin to dim.
And we go somewhere we've never been before;
not just entertained, but somehow reborn.
Together.
Reposted by Rebecca Dorn
Of the 95 people @propublica.org heard from 30 "said they tried to recover their belongings but hit obstacles, such as being unable to reach anyone at the facility or the site not having everything that was taken. Only one person got back all of his items."
www.propublica.org/article/home...
www.propublica.org/article/home...
Cities Say They Store Property Taken From Homeless Encampments. People Rarely Get Their Things Back.
Storage programs are meant to protect people’s property rights and allow them to reclaim their possessions. But they rarely accomplish either objective, according to a ProPublica investigation of citi...
www.propublica.org
November 21, 2024 at 6:31 PM
Of the 95 people @propublica.org heard from 30 "said they tried to recover their belongings but hit obstacles, such as being unable to reach anyone at the facility or the site not having everything that was taken. Only one person got back all of his items."
www.propublica.org/article/home...
www.propublica.org/article/home...
Reposted by Rebecca Dorn
it’s about to get very, very hard to be transgender in the united states. likely harder than it’s been in american history. do not let them scare you into backing down. do not give into transphobes framing. stand up to them however you can. it shouldn’t be this way but we have to fight
November 20, 2024 at 11:34 PM
it’s about to get very, very hard to be transgender in the united states. likely harder than it’s been in american history. do not let them scare you into backing down. do not give into transphobes framing. stand up to them however you can. it shouldn’t be this way but we have to fight
Reposted by Rebecca Dorn
Almost all 92 million low-income Americans now have some basic aspect of their lives determined by AI or automated decision-making software, a new report finds. Those systems are often flawed, and rarely in poor folks' favor.
Today's newsletter: www.washingtonpost.com/politics/202...
Today's newsletter: www.washingtonpost.com/politics/202...
Analysis | Flawed AI denies benefits to low-income Americans, report warns
As Trump prepares to slash spending, a new nonprofit emerges to fight automation in government benefit programs
www.washingtonpost.com
November 20, 2024 at 9:53 PM
Almost all 92 million low-income Americans now have some basic aspect of their lives determined by AI or automated decision-making software, a new report finds. Those systems are often flawed, and rarely in poor folks' favor.
Today's newsletter: www.washingtonpost.com/politics/202...
Today's newsletter: www.washingtonpost.com/politics/202...
Reposted by Rebecca Dorn
Excited to see them all at SoCalNLP this Friday!
Started a SoCal AI/ML/NLP researchers starter pack! It's a bit sparse right now, and perhaps more NLP heavy, but hey, nominate yourself and others! go.bsky.app/6QckPj9
November 19, 2024 at 7:04 PM
Excited to see them all at SoCalNLP this Friday!