Summary of From Experts to the Public: Governing Multimodal Language Models in Politically Sensitive Video Analysis, by Tanusree Sharma et al.
From Experts to the Public: Governing Multimodal Language Models in Politically Sensitive Video Analysis
by Tanusree Sharma, Yujin Potter, Zachary Kilhoffer, Yun Huang, Dawn Song, Yang Wang
First submitted to arxiv on: 15 Sep 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper explores how individuals and groups make decisions about multimodal large language models (MM-LLMs) when analyzing videos with political content. The authors conducted two studies: first, they interviewed 10 journalists to understand expert video interpretation, and second, they had 114 people from the general public engage in deliberation using a platform that facilitates democratic decision-making through decentralized autonomous organization (DAO) mechanisms. The findings show that experts emphasized emotion and narrative, while the general public prioritized factual clarity, objectivity, and emotional neutrality. The paper also examines how different governance mechanisms, such as quadratic vs. weighted ranking voting and equal vs. 20-80 power distributions, affect users’ decisions on how AI should behave. The results suggest that applying DAO mechanisms can help democratize AI governance. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about how people make decisions when they use special computer models to analyze videos with important political content. The authors talked to 10 journalists first and then had 114 regular people work together to decide what the models should do. They found that experts are good at understanding emotions and stories in the videos, but regular people want facts and objectivity. The paper also shows how different ways of making decisions can affect what people think about democracy and equality. |
