I'm a principal applied scientist at Microsoft, a senior scientist at UMN and a Wikipedia volunteer. I study socio-technical dynamics to build better tech.

Minnesota, US
Joined October 2009
❤️This is a collaborative work with a wonderful team of co-authors @halfak, @Zirui_Cheng_, @jordynjkim, Meng-Hsin, @tongshuangwu, and my advisors @d19fe8 and @Haiyi_Zhu. Please check our paper for more details: arxiv.org/abs/2402.14147 (8/9)
1
1
9
1,051
Aaron Halfaker retweeted
Observations and speculations @WikiEducation speaker series “Wikipedia in a generative ai world” today @ragesoss @halfak @tnbob @harrisonstephen #jojodoodles
1
3
7
724
"This meeting could have been an email." Soon enough, every meeting is going to be an email.
Replying to @jteevan
#Slide52: Too many meetings this week? Large language models can transform some of your synchronous meetings into async interactions. Listen to @halfak bring today's and yesterday's slides to life in <90 seconds: microsoft.com/en-us/research… aka.ms/nfw2022
1
It's currently pi-o-clock on May the 4th. :)
2
"Automatically Labeling Low Quality Content on Wikipedia by Leveraging Patterns in Editing Behavior" improves detection of sentences that need citations, NPOV, and clarifications. Work by @sumit_ast, Sabrina Tobar Thommel, @halfak, and @nikola_banovic was accepted to #CSCW2021.
2
4
1
28
One of my colleagues (@julia_kiseleva) is an organizer of the #igluContest, an open ML dataset competition for building agents that learn from natural language instruction. Check it out! iglu-contest.net/
We are happy to announce that the submission system is up and running! So #igluContest has officially started: Architect Task: competitions.codalab.org/com… Builder Task: competitions.codalab.org/com… Please register and try it out! #NeurIPS2021
2
Aaron Halfaker retweeted
Looking for a job that combines the benefits of industry and academic positions? Have we got an opening for you! The Microsoft Office of Applied Research is seeking researchers who will lead projects that advance the state-of-the-art and Microsoft’s products at the same time.
1
14
5
66
I rode my bike a little bit this year. COVID be damned.
3
This fantastic graphic summarizes a ton of useful information for cyclists.
This Post is from an account that no longer exists. Learn more
1
"How do we align our AI systems to the kind of world we would want to live in?" I enjoyed read this nice view of Human-in-the-loop AI by @gewang hai.stanford.edu/blog/humans… I'm looking for thoughts on design of AI-powered UX. Any other suggestions?
4
Getting up at 5AM was worth it to see some of my work with @staeiou show up in the @CSCW2020 celebration of systems research panel -- our work measuring labor hours in Wikipedia. stuartgeiger.com/cscw13-labo…
13
#CSCW2022 will be in Taiwan! #CSCW2023 will be "back" in Minneapolis! (This is the one where I am general co-chair!) And... 🥁🥁🥁 #CSCW2021 will be virtual again, with the hopes this will result in a strong, accessible conference. Likely 10/23-10/27.
2
14
3
51
I'm setting up a Minneapolis bike tour live stream for #CSCW2020. A couple reflections: (1) Minneapolis is beautiful this time of the year and (2) GoPro's facilities for live streaming through a phone's hotspot is abysmal UX. Oof.
1
4
Hey folks. I'm looking to put together some "local" activities for #cscw2020 (which will be virtual). Have you had any virtual conference activity experiences (virtual hallway, meetups, conversation matches, etc.) that you thought worked well?
1
4
Aaron Halfaker retweeted
"ORES: Lowering Barriers with Participatory Machine Learning in Wikipedia", discussing the theoretical mechanisms of social change enabled by ORES, 5 years after its creation. (@halfak and @staeiou, @ACM_CSCW 2020 ) arxiv.org/pdf/1909.05189.pdf
2
12
2
24
I just started my new gig working at Microsoft with @bhecht and @jteevan. Still reviewing projects to find a place where I can contribute. This means I've taken the bridge to #Wikipedia volunteer land. It's going to be a big adjustment after 9 years as @Wikimedia staff.
19
1
1
100
Aaron Halfaker retweeted
"The effects of algorithmic flagging on fairness: quasi-experimental evidence from Wikipedia" algorithmic ƒflagging systems can lead to increased fairness but that the relationship is complex and contingent (TeBlunthuis et al, 2020) arxiv.org/pdf/2006.03121.pdf