Join us for our
#NoTechForApartheid
event in with
@haymarketbooks
to amplify the
#NoTechForApartheid
campaign & learn what we can do as tech workers. We don't have the registration page yet but you can add your email here to get info:
Details in thread ๐
We are thrilled to welcome Dr. Alex Hanna as our incoming director of research. A sociologist by training, her work centers on the data used in new computational technologies, and the ways in which these data exacerbate racial, gender, and class inequality. 1/4
Today's my last day at Google. Starting tomorrow I'm joining
@timnitGebru
at
@DAIRInstitute
as Director of Research.
On my way out, here's some thoughts on the tech company as a racialized organization and the power of complaint.
Twitter's machine learning ethics & transparency (META) team led by our friend
@ruchowdh
, has been laid off along with countless other workers. What's happening there should be a wakeup call on the consequences of leaving our technological future to the whims of billionaires.
In this Op-ed,
@timnitGebru
&
@mmitchell_ai
note that "Scientists & engineers should focus on building models that meet peopleโs needs for different tasks, & that can be evaluated on that basis, rather than claiming theyโre creating รผber consciousness."
As we celebrate our launch day, we want to extend deep gratitude to our Advisory Committee Members
@safiyanoble
and
@CwaMaina
who have helped us get here, and our first research fellow Raesetja Sefala
Congratulations to Rylan Schaeffer, Brando Miranda, Sanmi Koyejo for winning a best paper award at NeurIPS for this insightful paper.
Are Emergent Abilities of Large Language Models a Mirage?
"AI has launched a new era in unpaid labor, one in which everyone participates in building the foundations for new companies and technologies while remaining completely oblivious to their contributions."
@adrienneandgp
and
@MilagrosMiceli
For our 1 year anniversary virtual events, we'll have talks, panels and breakout discussions with attendees. Sign up here:
Why should we always clean up after harmful tech rather than implementing our imagination of a technological future that serves us?
Today we welcome
@MilagrosMiceli
as a research fellow.
Milagros's research focusses on labor conditions and power asymmetries in outsourced data work as well as their effects on machine learning datasets. She has conducted ethnographic fieldwork, interviews and workshops...
Join us on March 17th for Stochastic Parrots Day marking the 2nd anniversary of the paper. The authors will be joined by scientists, teachers, artists, legal scholars & data workers to discuss the large language model landscape.
Sign up here:
Join us virtually on December 2nd and 3rd as we celebrate our 1st anniversary. We'll have interactive talks and conversations with DAIR's team members located across North America, Europe & Africa. 1/4
That's a wrap. Thank you for celebrating with us.
Stay tuned for our research philosophy which will be released in English, Spanish, Arabic, Tigrinya and Amharic along with our revamped website.
"When the contract with OpenAI was terminated eight months early, โwe felt that we were left without an income, while dealing on the other hand with serious traumaโ, said petitioner Richard Mathenge..."
We are looking for a full time summer intern to contribute to our goal of communicating our research to the public through creative means. Take a look at our job position and apply here:
big breaking news: LAION just removed its datasets, following a study from Stanford that found thousands of instances of suspected child sexual abuse material
PRESS RELEASE ๐ข๐งต
@turkopticon
& DAIR applaud the Congressional announcement of a letter by Sen. Markey...& Rep. Jayapal...which demands that Amazon, Google, OpenAI, & other tech giants answer questions about the working conditions of data workers,
โData is the last frontier of colonization,โ Mahelona says.
"In turning to AI to help revive te reo, the Mฤori language, Mahelona and Jones, who is Mฤori, wanted to do things differently."
Please join us in welcoming Asmelash Teka as a part-time DAIR fellow.
Asmelash is the Co-founder and CTO of Lesan (
@lesan_ai
). At Lesan, he has built state-of-the-art machine translation systems to and from Amharic, Tigrinya and English. ๐๐๐
"...a number of AI experts, including
@alexhanna
, noted that despite the widespread proliferation of ethical frameworks & principles for AI, virtually none of them take into account the extensive human labour that underpins the technology."
Congratulations to the workers who successfully unionized ๐
""For too long we, the workers powering the AI revolution, were treated as different & less than...โ said Richard Mathenge"
If you missed Richard's panel on
#StochasticParrotsDay
, watch here:
Can't quite believe I'm writing this: Today, 150 African workers behind ChatGPT, TikTok and Facebook voted to unionize at a landmark meeting in Nairobi.
These AI workers are invisible, underpaid, and the backbone of the tech in all our pockets:
We are excited to welcome Meron Estefanos (
@meronina
,
) to DAIR as a research fellow. She is an activist and radio journalist who has been fighting for the rights of kidnapped refugees, and negotiating their release for 17 years.
Join us on Dec 2nd & 3rd for research highlights & panels at our 1 year anniversary event.
Day1 research highlights:
*Language tech for Ge'ez based languages |
@asmelashteka
*Refugee advocacy & Bitcoin |
@meronina
*Nobody owns our tech future |
@dylan
Please join us in welcoming our newest fellow, Adrienne Williams (
@adrienneandgp
)๐๐๐
Adrienne started organizing in 2018 while working as a junior high teacher for a tech owned and operated charter school. 1/4
Image descriptions of 4 screenshots in the quote tweet of
@alexhanna
's talk from
#WiDS2022
today.
1. โTo begin to see how biases are propagated and reinforced via machine learning system training data, we need histories of the datasets themselvesโฆโ
"The EU AI Act will only entrench police and migration authoritiesโ existing culture of impunity โ encouraging them to keep deploying harmful AI systems against the most marginalised communities."
#AIAct
Despite the big pronouncements about the EU overcoming lobbying efforts to pass 'the world's 1st comprehensive' rules for AI, the AI Act is a failure from a human rights perspective & a victory for industry & police
Read
@accessnow
's take here:
We have now setup the registration page for our event where you can directly register below. While the event is free, you have an option to donate with proceeds going to
@pal_legal
@MedicalAidPal
&
@RespondCrisis
Join us for our
#NoTechForApartheid
event in with
@haymarketbooks
to amplify the
#NoTechForApartheid
campaign & learn what we can do as tech workers. We don't have the registration page yet but you can add your email here to get info:
Details in thread ๐
"Coming from the community she studied, Sefala said, was crucial to the projectโs success."
Our fellow,
@bonjora
, emphasizes the importance of valuing people's lived experiences.
"We've shown that you can build useful models by using small, carefully curated data sets," said Asemalsh..."We understand its limitations & capabilities. Meanwhile, Microsoft or Google usually build a single, gigantic model for all languages, so it's almost impossible to audit."
"By simply putting the term โpro-choiceโ into Copilot Designer, with no other prompting, Jones found that the tool generated a slew of cartoon images depicting demons, monsters and violent scenes."
Our fellow, Raesetje Sefala (
@bonjora
) will talk about her work at
@Blackathonic
on February 5 at 1pm EST. You can get your tickets at โฆ to attend the event.
"In the end, the real problem is arguably not that the data labelers shared the images on social media. Rather, itโs that this type of AI training setโspecifically, one depicting facesโis far more common than most people understand, notes
@MilagrosMiceli
,
Congratulations to our director of research
@alexhanna
๐๐๐
"On Valentineโs Day 2011, graduate student Alex Hanna helped to lead a crowd into the office of Governor Scott Walker and dump 3,000 โWe โฅ UWโ valentines on his desk."
Join us for the 20th episode of Mystery AI Hype Theater Monday Nov. 6 at 10:30am Pacific where
@alexhanna
and
@emilymbender
travel back in time & examine the AI Hype in the grant proposal that funded the 1956 summer project at Dartmouth that started it all
โDonโt be fooled: itโs self-serving hype disguised as raising the alarm,โ
@dylnbkr
...research engineer
@DAIRInstitute
...current discussions regarding hypothetical existential risks distract the public and regulators from โthe concrete harms of AI today.โ
Join us on March 17th for Stochastic Parrots Day marking the 2nd anniversary of the paper. The authors will be joined by scientists, teachers, artists, legal scholars & data workers to discuss the large language model landscape.
Sign up here:
We stand in solidarity with our Palestinian siblings who have lived under brutal Israeli occupation & apartheid for 75 years, and now face a genocide with access to food, water, and fuel shutoff while being bombed and burned with white phosphorus. 1/17
Join us for another edition of Mystery AI Hype Theater 3000 on Monday April 1st at noon pacific time, where
@timnitGebru
will guest host with
@alexhanna
. In the mean time, you can catch up on old episodes
Are you catching up with the story of the lawyer who used ChatGPT? Watch this Mystery AI Hype 3000 Theater episode hosted by
@emilymbender
&
@alexhanna
where special guest
@KendraSerra
warned about this type of issue with generative AI Hype and law.
Congratulations to our very own
@adrienneandgp
and
for becoming Just Tech fellow.
"Adrienne Williams is a labor organizer & researcher with Distributed AI Research Institute."
"
@alexhanna
sees...financial incentives to rapidly commercialize AI outweighing concerns about safety or ethics. There isnโt much money in responsibility or safety, but thereโs plenty in overhyping the technology,"
by
@kharijohnson
She also works in the area of social movements, focusing on the dynamics of anti-racist campus protest in the US and Canada. Dr. Hanna has published widely in top-tier venues across the social sciences, including the journals Mobilization, American Behavioral Scientist, and 2/4
We will have conversations with an incredible group of scholars, journalists and activists.
1. Caroline Hunter who co-founded the Polaroid Revolutionary Workers Movement which forced the company to pull out of apartheid South Africa.
Join us for our
#NoTechForApartheid
event in with
@haymarketbooks
to amplify the
#NoTechForApartheid
campaign & learn what we can do as tech workers. We don't have the registration page yet but you can add your email here to get info:
Details in thread ๐
Statement from the listed authors of Stochastic Parrots on the โAI pauseโ letter
"Regulatory efforts should focus on transparency, accountability and preventing exploitative labor practices."
w/
@timnitGebru
@mmitchell_ai
and
@mcmillan_majora
Prof.
@emilymbender
tells us:
"if we mistake our ability to make sense of language and images generated by computers for the computers being โthinkingโ entities, we risk ceding power โ not to computers, but to those who would hide behind the curtain."
I refuse to be delegated to the "skeptics box" in someone else's framing of a debate. Here is my response to
@stevenbjohnson
's NYT Magazine article about LLMs and OpenAI.
On NYT Magazine on AI: Resist the Urge to be Impressed
Our second panel moderated by
@alexhanna
will start in 5 minutes. Join us here:
Panelists are: 1. Caroline Hunter who co-founded the Polaroid Revolutionary Workers Movement which forced the company to pull out of apartheid South Africa.
Join us for our
#NoTechForApartheid
event in with
@haymarketbooks
to amplify the
#NoTechForApartheid
campaign & learn what we can do as tech workers. We don't have the registration page yet but you can add your email here to get info:
Details in thread ๐
"Currently, no one can contest if they have...text thatโs been picked up in a training data set...used to train a model, people canโt say no to that. So figuring out how to allow contestation brings with it a...bunch of research questions"
@mmitchell_ai
"But while authorities generally pitch facial recognition as a tool to capture terrorists or wanted murderers, the technology has also emerged as a critical instrument in a very particular context: punishing protesters." by
@DarrenLoucaides
Big Data & Society, and top-tier computer science conferences such as CSCW, FAccT, and NeurIPS. Dr. Hanna serves a co-chair of Sociologists for Trans Justice and as a Senior Fellow at the Center for Applied Transgender Studies. 3/4
Our fellow,
@MilagrosMiceli
was interviewed for part 2 of
@techreview
's series on AI colonialism, by
@_KarenHao
.
"newer data-labeling companies have sought to establish a higher standard for working conditions...
Dylan Baker (
@dylnbkr
) and Dr. Alex Hanna (
@alexhanna
) write:
"Our intent, as the first full-time employees of the newly formed...(DAIR), is to find a new path for technological research, which centers communities impacted most by emerging technologies.
FastCompany included Dr. Hanna as part of their 2021 Queer 50, and she has been featured in the Cal Academy of Sciences New Science exhibit, which highlights queer and trans scientists of color. 4/4
5.
@antloewenstein
who wrote the must read book The Palestine Laboratory: How Israel Exports the Technology of Occupation Around the World, which Verso has made available FOR FREE as an e-book. Get yours here:
Researchers and practitioners in responsible AI demand an end to the "decades long, tech-enabled apartheid and occupation of the Palestinian people" which has culminated in "a genocide of the Palestinian people."
You can sign the letter here ๐
The first episode of Mystery AI Hype Theater 3000 for the year will be on Monday January 8 where Dr.s
@alexhanna
&
@emilymbender
will be joined by Kerry McInerney & Eleanor Drage to talk biometrics. Join us for the live stream at 9am pacific time at .
As journalist Antony Loewenstein noted in The Palestine Laboratory: How Israel Exports the Technology of Occupation Around the World: โPalestine is Israelโs workshop, where an occupied nation on its doorstep provides millions of subjugated people as a laboratory 5/17
Our 1 year anniversary celebration is happening in ~2 weeks! Sign up here for virtual panels, talks & breakout sessions with us:
In our 2nd panel
@alexhanna
@meronina
@dylnbkr
&
@MilagrosMiceli
will discuss non-exploitative community rooted research.
Looking for more Mystery AI Hype Theater 3000? Tune in live on Monday Oct 30 at 12pm pacific, when
@alexhanna
&
@emilymbender
are joined by
@SashaMTL
&
@strubell
to take on hype around AI and the environment.
As researchers, activists and organizers working in the field of AI, we know that Gaza and the West Bank have been the testing ground for some of the most advanced automated surveillance and warfare technologies before they are exported around the world. 4/17
We have now setup the registration page for our event where you can directly register below. While the event is free, you have an option to donate with proceeds going to
@pal_legal
@MedicalAidPal
&
@RespondCrisis
Congratulations to our very own
@Zebra_023
&
@bonjora
for being named Rising Stars in AI Ethics & for
@milamiceli
being named on of the โ100 Brilliant Women in AI Ethicsโ by Women in AI Ethics.
with data annotators, collectors, and scientists in different sites around the world.More broadly, she is interested in question of legitimacy and knowledge production in data, worker-led research, labor organizing, and the intersection of Critical Data Studies and data activism.
"Google & Facebook overhype that they have built one single giant model to solve machine translation for hundreds of languages. If you put it on its face & compare it with smaller startups, like Lesan or
@GhanaNLP
, the quality is...low."
Interview with our fellow
@asmelashteka
You can hear more from Richard on our Stochastic Parrots Day panel where he discussed worker exploitation with
@safiyanoble
,
@MilagrosMiceli
and
@timnitGebru
which you can watch here:
๐งต1/5
Open Letter: The Responsible AI Community in Solidarity with Gaza and Palestine
Sign on and Share (list of signatories will be updated once a day):
"For any automated system, we need to know what it was trained to do, what training data was used, who chose that data and for what purpose...mimicking human behaviour is a โbright lineโ โ a clear boundary not to be crossed โ in computer software development."
If you missed DAIR fellow
@adrienneandgp
and founder
@timnitGebru
discussing their Noema Magazine article co-written with
@MilagrosMiceli
on KQED's Forum live radio program, you can listen to the recording here, .
Our advisory committee member
@safiyanoble
is one of the signatories of this letter warning that "those who advocate for even the most basic rights for Palestinians" are facing "doxxing, death threats, deplatforming, censorship, & violent intimidation."
"A better future for AI is possibleโan AI future that isnโt motivated by vast accumulations of data gathered through the poorly regulated extraction of personal information or profit motives that are centered around maximizing and appealing to shareholders quarter after quarter."
Check out
@alexhanna
's and my new piece out today for the
@SSIReview
x
@FordFoundation
series on technology in the public interest! We write about why institutes like
@DAIRInstitute
are more necessary than ever in the current landscape of AI ethics:
2.
@marwasf
who leads Access Now's policy & advocacy work on digital rights in MENA region. Take a look at this publication of hers from yesterday. Read
"Palestine unplugged: how Israel disrupts Gazaโs internet"
Congratulations to all
@FAccTConference
best paper awardees, including our fellow
@MilagrosMiceli
for co-authoring Algorithmic Tools in Public Employment Services: Towards a Jobseeker-Centric Perspective,
The Israeli military has killed an estimated 14 Palestinians every hour, and dropped the equivalent of a quarter nuclear bomb just in the first 10 days of its bombardment of Gaza. We call on our peers in the field of AI, 2/17
Solidarity with our fellow
@meronina
and other human rights defenders calling on the Human Rights Foundation to actually stand for human rights and against apartheid, occupation and genocide. Read their letter here:
The
@OsloFF
is coming up and as former speakers/participants we're asking everyone to
#BoycottOFF
. You can read the letter at and I hope you'll join us in signing on.
You either stand for justice everywhere or not at all...
"Giving the example of when she had to label aerial photos of border crossings, Kauffman says one of the major issues...is the lack of transparency about what theyโre working on, & who for."
Interview with our fellow &
@turkopticon
organizer
@Zebra_023
Our article by
@adrienneandgp
@MilagrosMicel
&
@timnitGebru
is one of the 10 most read pieces in 2022 on Noema Magazine.
You can read the piece here & continue the conversation during our 1st year annivesary event.
Sign up here:
Weโre excited to announce our best pieces from 2022!
From โdeep timeโ to the sophisticated minds of insects, thereโs plenty here for the intellectually curious.
3.
@Monashtayya
: Campaigns & Partnerships Manager (MENA) & Corporate Engagement Lead at Digital Action. She was also advocacy & communications manager at
@7amleh
. See 7amleh's letter on silencing of Palestinian voices on social media here:
In this op-ed for
@techpolicypress
"Policy Is Urgently Necessary to Enable Social Media Research"
@dylnbkr
outlined the hurdles that social media companies place on those doing research that holds them accountable.
6.
@DocMattMoudi
who is a Researcher/Adviser for
@AmnestyTech
& Lecturer at University of Cambridge. Take a look at this must read report he worked on "Israel and Occupied Palestinian Territories: Automated Apartheid:
This week, we welcome
@dylnbkr
to DAIR. Dylan's engineering and research work is broadly focused on ethics in machine learning, particularly how it relates to data labor, classificatory systems, and identity. 1/4