hhmx.de
timnitGebru@dair-community.social
timnitGebru@dair-community.social

Timnit Gebru (she/her)

(@timnitGebru@dair-community.social)

Fr 04.11.2022

Beiträge: 4.233Folgt: 460Folgende: 33.763

Fired from Google for raising issues of discrimination in the workplace and writing about the dangers of large language models: wired.com/story/google-timnit-.

Founded The Distributed AI Research Institute (dair-institute.org/) to work on community-rooted AI research.

dair-community.social · mastodon · 2025-01-17 14:30:04

Timnit Gebru (she/her)

Föderation EN Mi 12.03.2025 22:33:16

Good. Lets replace search with "AI" then! Totally logical if you ask me. Even more worth it when you know they're exponentially overtaking the airline industry in their carbon footprint.

Study: cjr.org/tow_center/we-compared

Tweet by Gisele Navarro @ichbinGisele

"“Collectively, the AI search engines provided incorrect answers to more than 60% of queries.” ~ @CJR

Screenoshot of an image showing "Generative search tools were often confidently wrong in our study" showing baplots of different offerings like ChatGPT, Perplexity and Deepseek, and showing the percentage of wrong answers in red and right ones in green.

(Medien: 1)

Timnit Gebru (she/her)

Föderation EN Mi 29.01.2025 06:14:51

Why does everyone get so excited about performance on some benchmarks when we don't know what the training data is? How do you know that they're not testing on the training set? How do you reject papers that don't tell you the splits but get so excited by these PR announcements?

Timnit Gebru (she/her)

Föderation EN Mi 29.01.2025 06:00:40

Friends, for something to be open source, we need to see

1. The data it was trained and evaluated on

2. The code

3. The model architecture

4. The model weights.

DeepSeek only gives 3, 4. And I'll see the day that anyone gives us #1 without being forced to do so, because all of them are stealing data.

Timnit Gebru (she/her)

Föderation EN Fr 25.10.2024 23:16:03

What's even the point of doing this type of work to show these issues at this point, since the conversation has been stuck at this stage while these companies are handed billions to "save humanity"?

arstechnica.com/ai/2024/10/goo

Timnit Gebru (she/her)

Föderation EN Do 29.08.2024 04:26:55

Humanity saving at work.

Megan Fritts
@freganmitts
Second week of the semester and I've already had students use (and own up to using) ChatGPT to write their first assignment: "briefly introduce yourself and say what you're hoping to get out of this class". They are also using it to word the *questions they ask in class*.

(Medien: 1)

Timnit Gebru (she/her)

Föderation EN Mo 05.08.2024 17:18:14

Let me tell you something @freedomtux, saying that people bombarded by the N word, Nazis and such "whining about racism" is racist.

And you're not the one who gets to define what is racist and not.

And if that is the "platform of the people" well then this place has curated a racist as hell community.

I'm blocking all y'all but not before exposing you.

Jushua M @freedomtux@fosstodon.org. 

Mastodon is the voice of the people and doesn't need defending. But I hear 50x more whinging about racism than actual racism on the #fediverse.

(Medien: 1)

Timnit Gebru (she/her)

Föderation EN Fr 26.07.2024 22:02:05

Y’all out here talking about our instance moderation when people talk about racism on Mastodon, wanting to believe that all the racists hang out at IamANazi.social.

Maybe mastodon isn’t the “good people” platform and is filled with racists like everywhere else?

Timnit Gebru (she/her)

Föderation EN Mi 10.04.2024 03:47:13

You have refugees who escaped the most ruthless totalitarian regime in the world, which co-committed the worst genocide of the 21st century that killed 1M people, a regime that disappears people and puts them in shipping containers where their heads explode from the heat. These refugees get human trafficked and tortured trying to escape, and find this SAME regime operating with impunity in Europe. And the police act like its bouncer opening water cannons on protestors???

swissinfo.ch/eng/foreign-affai

Timnit Gebru (she/her)

Föderation EN Fr 29.03.2024 19:06:49

$660M crypto money to “save humanity from the existential risk of AI.” Same people wrote a few years ago how “without AI we are doomed.” Selling utopia and apocalypse for billions. No one knows how to circle money amongst themselves like these people.

politico.com/news/2024/03/25/a

Timnit Gebru (she/her)

Föderation EN Do 28.03.2024 03:04:32

I am going to have a lot to say about this article given that 1) I am an Eritrean who understands the level of transnational repression by the government & 2) My research in bias in AI, such as the face recognition systems they talk about here.

faz.net/aktuell/rhein-main/gew