MIT Center for Constructive Communication
@cccatmit
Designing tools, methods, and systems to understand and address societal fragmentation.
ID: 2491957010
https://lnk.bio/cccxmit 12-05-2014 20:57:46
481 Tweet
4,4K Followers
124 Following
🧠New research from MIT Center for Constructive Communication: Even AI reward models trained on factual data can show political bias. Led by Suyash Fulay, Jad Kabbara, & Deb Roy, the study uncovers challenges in aligning truthfulness & neutrality in AI. Read more: news.mit.edu/2024/study-som… MIT Media Lab Massachusetts Institute of Technology (MIT)