We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
cover of episode Hate Speech, Applied AI, NYPD, & Grades

Hate Speech, Applied AI, NYPD, & Grades

2020/8/26
logo of podcast Last Week in AI

Last Week in AI

AI Deep Dive AI Chapters Transcript
People
A
Andrey Kurenkov
D
Danielle Bashir
S
Sharon Zhou
Topics
Danielle Bashir: 本周新闻主要关注 Facebook 利用 AI 检测仇恨言论的挑战,包括图片、模因等混合媒体内容的识别难度;AI 研究社区对现实世界问题的关注不足,以及算法偏差导致的社会不公平问题;纽约市警察局使用面部识别技术引发争议;英国政府使用算法决定 A-level 和 GCSE 考试成绩的失败案例。 Andrey Kurenkov: Facebook 在仇恨言论检测方面取得了显著进展,但仍面临挑战,尤其是在处理模因等混合媒体内容方面。同时,许多 AI 研究人员更关注理论问题,而忽视了现实世界问题的解决。算法偏差问题也需要引起重视。英国政府使用算法决定考试成绩的案例说明,在大型 AI 应用中,需要考虑备选方案和公众意见。 Sharon Zhou: Facebook 的 AI 系统在检测针对少数群体的仇恨言论方面表现较差,这凸显了算法偏差问题。模因是 AI 检测仇恨言论的一大挑战,因为需要同时处理图像和文本信息。纽约市警察局使用面部识别技术搜捕 Black Lives Matter 活动家的事件,其动机和面部识别技术的作用尚不明确。英国政府使用算法决定考试成绩的决定是错误的,因为它忽视了学生的个体情况,并导致了不公平的结果。

Deep Dive

Chapters
This chapter discusses Facebook's efforts to use AI for detecting and removing hate speech and misinformation, highlighting challenges such as the complexity of mixed media content like memes and the AI's uneven performance across different demographics.

Shownotes Transcript

Our latest roundup of last week's big AI news!

Plus, Stanford AI Lab PhDs Andrey Kurenkov and Sharon Zhou discuss the news and offer their thoughts.

We are trying out this combination of formats for the first time - please fill out the listener survey to let us know what you think about it: bit.ly/ltasurvey

Check out all the stories discussed here and more at www.skynettoday.com)

Theme: Deliberate Thought Kevin MacLeod (incompetech.com)