Niloufar Salehi

Niloufar@hci.social

<p>Human Computer Interaction, Human Centered AI, health communication, restorative justice</p><p>Assistant professor at UC, Berkeley School of Information</p>

Posts

Post #673823
Ppl can&#39;t seem to stop talking about LLMs having their iPhone moment. But what if the LLM isn&#39;t the iPhone? More like the touch screen?
Post #673817
Many models produce outputs that are hard to verify for an end-user. Our new #emnlp2023 paper shows that providing a quality estimation model, can make a user better at deciding when to rely on the model. https://arxiv.org/pdf/2310.16924v1.pdf
Post #673816
@katestarbird really interesting, we make a similar argument here about harassment interventions: https://niloufar.org/wp-content/uploads/2023/03/sustainedonlineharmmuslims-CSCW2023.pdf
Post #673815
Many models produce outputs that are hard to verify for an end user. 🏆 Our new #emnlp2023 paper won an outstanding paper award for showing that a secondary quality estimation model can help users decide when to rely on the model output. We ran a controlled experiment showing that a calibrated quality estimation model can make physicians twice better at correctly deciding when to rely on a translation model output. Paper: https://arxiv.org/pdf/2310.16924v1.pdf
Post #673814
So the whole rich people buying social media the second they don&#39;t like what is on it is foreshadowing what&#39;s going to happen to all the big models isn&#39;t it? https://www.cnbc.com/2024/03/14/former-treasury-secretary-mnuchin-is-putting-together-an-investor-group-to-buy-tiktok.html
Post #673813
Colorado&#39;s star DNA analyst intentionally manipulated data and affected more than 650 cases. The U.S. government increasingly uses difficult to scrutinize, ML-based decision systems to convict and incarcerate people, such as facial recognition, gunshot detection, and probabilistic genotyping tools. Our #chi2024 paper asks what challenges do people face contesting these systems in court? https://www.wsj.com/us-news/colorados-star-dna-analyst-intentionally-manipulated-data-investigation...
Post #671833
What happens when online #counterpublics try to engage externally - meaning try to convince other publics that their ideas are valid? Our research found that they face severe harassment extended over time and space that weaponizes affordances, gender, and identity. New #CSCW2023 paper with @RoyaPak Nazita Lajevardi and Mariam Asad. https://niloufar.org/wp-content/uploads/2023/02/OurMuslimVoicesCSCW2023.pdf
Post #670941
Design is the art of being wrong safely https://spavel.medium.com/design-is-the-art-of-being-wrong-safely-7575b0c395c2