Dr Abeba Birhane

Dr Abeba Birhane

@abeba.bsky.social

Founder & PI @aial.ie. Assistant Professor of AI, School of Computer Science & Statistics, @tcddublin.bsky.social AI accountability, AI audits & evaluation, critical data studies. Cognitive scientist by training. Ethiopian in Ireland. She/her

51,661 Followers 864 Following 4,203 Posts Joined May 2023
13 hours ago
it just feels so horrible
we feel powerless
we've tried fighting, 
we've tried surrendering
we've tried ignoring
we've tried building
israel doesn't let you exist
that's it, our choice is fast death 
or faster death

From a friend in Beirut this morning, who in the last few days has seen several Israel missile strikes within walking distance of their home.

1,342 524 12 8
6 hours ago

That's one awful feature of Grammarly gone - plenty more terrible features still to go www.theguardian.com/books/2026/m...

22 13 1 0
7 hours ago
Preview
Meta to Shut Down Instagram End-to-End Encrypted Chat Support Starting May 2026 Meta will end Instagram E2EE chats May 8, 2026, reversing a 2021 privacy test and reigniting debate over encrypted messaging oversight.

In this golden age of surveillance, making E2EE chat normal and expected has been privacy's biggest win. I hate to see it getting rolled back. thehackernews.com/2026/03/meta...

135 67 3 9
6 hours ago
Preview
Musk's X submits remedies relating to its blue check mark, EU spokesperson says Elon Musk's social media platform X has submitted remedies in relation to its blue checkmark ​verification feature following a fine of 120 ‌million euros ($137.63 million), a European Commission spoke...

X has offered to make changes to its verification system after a multi million euro fine by the EU Commission under the DSA.

Without knowing exactly what X is proposing it’s hard to form an opinion, but a deeply cynical person might think this is a stalling tactic by X to delay full enforcement.

23 11 1 0
8 hours ago

Chilling. These people know that their ideas are so unpopular that people would never agree to them by consent, so they use algorithms.

47 22 2 1
9 hours ago
A picture of me holding a placard that says Free Palestine, Staying silent is not an option

#FreePalestine #Dublin

996 178 16 6
11 hours ago
Preview
Dark money group offers influencers $1,500 for posts attacking Chicago Democratic primary candidate Progressive House candidate Kat Abughazaleh told MS NOW the secretive campaign is “filled with false and defamatory claims.”

HOLY SHIT.

A dark money group, hidden behind a bunch of shell companies, is paying influencers thousands of dollars PER POST to attack @katmabu.bsky.social on social media a month before her election.

Thank you to the people who declined the money and went forward with this. So who's behind this?

17,732 7,170 724 523
1 day ago
Post image

I can't even find the words to express how repugnant and despicable this post is.

People who spew such Islamophobic bile don't belong in the United States Congress. The job of elected officials is to combat hate, not fan its flames. (1/2)

18,820 3,416 1,071 337
15 hours ago

Más allá de la barbaridad a todos los niveles (y que siguen usando 15 años más tarde), la tecnología es política, claro. Obvio que cuando te dicen que "la tecnología nos permite alejarnos de los sesgos, de la ideología, del partidismo" en realidad quieren decir "me permiten hacer lo que YO quiera"

99 32 1 0
16 hours ago

This parasite's company has billions in UK government contracts. I would very much like to know why.

55 19 1 1
16 hours ago

What I find shocking is that the UK government has given Palantir so much control over vital UK infrastructure.

45 16 1 0
16 hours ago

Palantir CEO Karp thinks his AI tech will lessen the power of “highly educated, often female voters, who vote mostly Democrat” while increasing the power of working-class men. Meanwhile @vonderleyen.ec.europa.eu @eppgroup.bsky.social diligently working to dismantle AI regulation. 🤦‍♀️

36 17 1 0
15 hours ago

i can’t tell what’s more painful: grant writing or progress/impact/financial report writing

27 2 2 1
16 hours ago
Video thumbnail

this was Thiel making it very explicit in 2010

131 64 4 12
16 hours ago
Preview
Palantir CEO Makes Shocking Confession on Disrupting Democratic Power They’re saying the quiet part out loud now.

this is shocking only if one is not remotely familiar with the history of Palantir, Thiel and Karp

newrepublic.com/post/207693/...

259 91 6 12
1 day ago
View Boris Gamazaychikov’s  graphic link
Boris GamazaychikovBoris Gamazaychikov
   • 1stVerified • 1st
Head of AI Sustainability @ SalesforceHead of AI Sustainability @ Salesforce
32m • Edited •  32 minutes ago • Edited • Visible to anyone on or off LinkedIn

The gas-powered data center craze has come to Europe.

Yesterday, a 110MW off-grid AI data center outside of Dublin was switched on - and it's powered entirely by fossil fuels.

Developers went with natural gas engines (specifically RICE) to bypass the grid connection queue. They claim the engines "could" be switched to HVO, and that the project will attempt to connect to the grid "when possible".

What does this mean for Ireland's climate goals? According to my estimates, this single project will increase the entire country's annual GHG emissions by ~1%.

Assumptions:
- The project seems to be using Rolls Royce (mtu) 4000 Series RICE which have an electrical efficiency of 44% https://lnkd.in/eRX4PRuf
- Modifying the NREL life cycle gas emissions factor with this efficiency yields 656.75 gCO2e/kWh https://lnkd.in/gfgs2X6P
- Assuming a 90% capacity factor because AI hyperscale data centers run near 24/7 to recoup massive GPU costs
- Ireland’s 2024 GHG emissions = 53.8 million tCO2e https://lnkd.in/ejjRgRpE

hashtag#DataCenters hashtag#Sustainability hashtag#ArtificialIntelligence Royce (mtu) 4000 Series RICE (natural gas)
Al Data Center

Ireland's new fossil-fuelled #Petrotech data center will increase Ireland's **total** (everything, not just the power sector) emissions by ~1%

Absolutely nuts. A 1% national increase - for what? Come on boosters, explain why this is worth it 🫠

www.linkedin.com/posts/bgamaz...

255 122 11 17
1 day ago

the “intelligence” he’s promising is just a suicide-instructing, porn-generating, plagiarism machine

85 17 3 0
1 day ago

👌🏾

4 0 0 0
1 day ago
Preview
Keir Starmer says what Kneecap stand for is ‘completely intolerable’ The Prime Minister made the statement during a trip to Belfast

Fuck Keir Starmer

- A man who couldn't get us cancelled from Glasto

- A man who couldn't beat us in court

- A man who has destroyed the Labour Party

- A man whose government supports war criminals in Israel with weapons and intelligence

www.independent.co.uk/news/uk/home...

1,106 307 34 27
1 day ago
Preview
Reframing Impact: AI Summit 2026 The 2026 AI Impact Summit in India is the latest iteration of an event that has become a bellwether for global discourse around the AI industry, especially the question of whether, and how, it can be ...

. @meredithmeredith.bsky.social, @grohmannrafael.bsky.social, Chenai Chair, @chinasa.bsky.social, Nikhil Dey, @timnitgebru.bsky.social, @audreyt.org, @abeba.bsky.social, Joan Kinyua, Usha Ramanathan, @naomiaklein.bsky.social, @karenho.bsky.social, Sumedha Deshmukh, Mila T Samdub

shorturl.at/pCIlv

15 6 0 0
1 day ago
Preview
Beyond Impact Lingo: Questioning, Concretizing, Building - AI Now Institute In the lead-up to this year’s India AI Impact Summit, we attempted to pre-bunk a new kind of AI hype that was circulating. We observed that the “right” words were being used to have the wrong conversa...

Last month, we proposed ways to "reframe" the AI Summit conversation to bring meaning back to co-opted terms like sovereignty, democratization, and accountability.

How can we turn the seeds of the alternative agenda into action?

Read our final essay: ainowinstitute.org/general/beyo...

19 7 2 0
1 day ago

*waves in civic media*

The fact most fo us who do that are unemployed and we’re pivote that way because of AI ……

Means it’s probably correct

16 3 0 0
1 day ago
Post image

what kind of overleaf error is this

11 0 3 0
1 day ago

amen!!!

4 0 0 0
2 days ago

Just a couple of weeks ago I was trying to think of non-evil uses for AI, and I thought, "well Grammarly just checks your grammar, that's not too bad, right?"

lesson learned I guess.

40 8 0 0
2 days ago

I mean isn't there one .. just one
.. person in the company who said "folks this isn't right or legal and people are going to be pissed"? Or is it management overruling every such voice raised?

14 3 5 0
2 days ago
Preview
Grammarly Is Facing a Class Action Lawsuit Over Its AI ‘Expert Review’ Feature The feature, which Grammarly shut down Wednesday, presented editing suggestions as if they came from established authors and academics—without their consent.

I'm suing Grammarly over its paid AI feature that presented editing suggestions as if they came from me - and many other writers and journalists - without consent.

State law requires consent before someone's name can be used for commercial purposes.

www.wired.com/story/gramma...

7,963 2,418 120 207
2 days ago

this thread! 🫡

7 1 0 0
2 days ago

let's fucking go

61 5 3 0