Siobhan Mackenzie Hall
@smhall97.bsky.social
680 followers 190 following 95 posts
🇿🇦 | Fairness in AI | University of Oxford | Deep Learning Indaba | Internships: Google DeepMind || Microsoft Research
Posts Media Videos Starter Packs
Pinned
smhall97.bsky.social
We are so proud of our ✨Honourable Mention✨ for the #FAccT2025 Best Paper Awards! This work was borne out of (com)passion, determination & a commitment to challenging norms in dataset collection & evaluation. Read more ➡️ arxiv.org/abs/2406.09496

Catch the presentation on Thurs, 11:45 local time! 🇬🇷
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
Reposted by Siobhan Mackenzie Hall
We need new rules for publishing AI-generated research. The teams developing automated AI scientists have customarily submitted their papers to standard refereed venues (journals and conferences) and to arXiv. Often, acceptance has been treated as the dependent variable. 1/
Reposted by Siobhan Mackenzie Hall
morganklauss.bsky.social
Today is the deadline! But we don't mind if you submit through the weekend, no one is checking ;)
morganklauss.bsky.social
Once more sharing our #CSCW2025 (NOW HYBRID) workshop on how to responsibly curate data for massive AI models. The deadline is September 5 (next Friday)! Note that the application form is simple and minimal :) responsible-data-workshop.github.io/cscw2025/
Workshop on Responsibly Training Foundation Models @ CSCW2025
responsible-data-workshop.github.io
smhall97.bsky.social
Fingers crossed you made /
will make the deadline! 🤞
Reposted by Siobhan Mackenzie Hall
piotrmirowski.bsky.social
Research prototype, early access announcement and participatory AI! I am really excited to invite you to try Fabula, an experimental app designed to serve as a writing tool for screenwriters and playwrights of all skill levels. We invite critical feedback: read the thread below!
youtu.be/YH7sZ2GN_m0
Announcing early access to Fabula, a prototype AI app for writing by Google DeepMind researchers
YouTube video by Piotr Mirowski
youtu.be
Reposted by Siobhan Mackenzie Hall
pcastr.bsky.social
want to help shape the future of AI in latin america? then don't miss out on @lasalaai.bsky.social next march!
the past instances of this event have been truly fantastic, and i'm super pumped for the next one next year. don't miss out and apply today!
hope to see you in quito!! ♥️🇪🇨
lasalaai.bsky.social
📢MARK YOUR CALENDARS!🗓️
The Summit of AI in Latin America (SALA, formerly RIIAA) is happening in Quito, Ecuador, March 9-12, 2026, at USFQ!

Want to attend and apply for travel grants? Do so here before October 1st:
forms.gle/PgYUQ2urGNTs...

More details in thread 👇🏾🧵
Reposted by Siobhan Mackenzie Hall
lastpositivist.bsky.social
Sinners was fun. I enjoyed that film. Good stuff sure.
Reposted by Siobhan Mackenzie Hall
mariaa.bsky.social
We have a really strong position paper that has been rejected multiple times from arxiv, both CS.AI and CS.CY, because our "submission does not contain sufficient original or substantive scholarly research and is not of interest to arXiv." Appealing has not helped. Any ideas of what we can do?
Reposted by Siobhan Mackenzie Hall
jeffdean.bsky.social
Honored to be named in TimeAI 100 list, along w/my awesome Google colleagues Josh Woodward & Hartmut Neven & many awesome people.

My recognition is due to collaboration with many great colleagues, so thank you to everyone I've been fortunate to work with over many years! 🙏

time.com/collections/...
time.com
Reposted by Siobhan Mackenzie Hall
vukosi.bsky.social
We gathered in Kigali 🌍 for the 2025 Deep Learning Indaba. Nearly 3,000 applied, 1,294 joined in person, 400 virtually. 6 days of AI research, debate, teaching & learning. From 300 in 2017 → to a pan-African Baobab tree. 🌳 #DLI2025 @DeepIndaba 1/
smhall97.bsky.social
An absolute 🔥🔥panel today at the Centring Data in African AI workshop at #DLI2025.
smhall97.bsky.social
Joy Victor sharing the scary reality of (the lack of) data protection on the continent - social security numbers sold for as little as 7c.
smhall97.bsky.social
“A high profile individual’s social security number was sold for as little as 7c.
smhall97.bsky.social
⚡️⚡️AI Ethics

💡It's not enough to create data, we need to think about ethics. And for ethics we need to think about human understanding

💡Performance of the model is not enough. Also need interpretability.
smhall97.bsky.social
Lightning talks continued

❓How do you crowdsource when there is no crowd?

We need to rethink the "conventions" on how we create data.
smhall97.bsky.social
💡 An appropriate answer to the right problem is worth a good deal more than an exact answer to an approximate problem."
John Wilder Tukey
smhall97.bsky.social
Next up we have another amazing keynote from Girmaw Abebe Tadesse, PhD from Microsoft’s AI for Good lab presenting his talk “Problem-driven and data-centric AI for Africa”
smhall97.bsky.social
What an amazing community to be a part of
smhall97.bsky.social
Next up we have our breakout sessions, where we look to our workshop attendees to share their insights across five prepared themes

⭐️ Data curation
⭐️ Who benefits from our data
⭐️ Beyond data
⭐️ Diversity
⭐️ Governance
smhall97.bsky.social
✨African data has helped train LLMs to scale this support

✨ This data also helps create culturally and economically relevant AI - avoiding recommending "salmon" to an expecting mother living in extreme poverty
Reposted by Siobhan Mackenzie Hall
willie-agnew.bsky.social
I often hear that critical AI researchers should start providing more solutions, but I rarely hear that uncritical AI researchers should start causing fewer problems.
smhall97.bsky.social
Lyvia Lusiji: Data to Delivery: Designing LLMs for Maternal Health Impact in Low-Resource Settings

📣 Mums do not have access to the Internet
📣 Appointment reminders
📣 Q&A systems to support the pregnancy journey
📣 Triaging, with a human health desk when neededd
smhall97.bsky.social
Lightning talks ⚡

📣 Who is benefitting from the data being collected?

A strong theme coming through even in the early hours at the workshop