We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
cover of episode 2020, China, Face Recognition, and DeepFakes

2020, China, Face Recognition, and DeepFakes

2020/8/14
logo of podcast Last Week in AI

Last Week in AI

AI Deep Dive AI Chapters Transcript
People
A
Andrey Kurenkov
S
Sharon Zhou
Topics
Andrey Kurenkov:2020年的诸多变化,例如口罩的普及和居家办公的增多,导致现有AI系统在人脸识别和场景分类等方面难以适应。AI在应对这些变化时表现出脆弱性,难以像人类一样灵活调整。此外,AI无法神奇地解决所有问题,例如研发新冠疫苗,它只能辅助部分流程,无法替代人类的实验和测试。在人脸识别领域,中国利用AI技术进行大规模监控,其数据收集能力强大,并与政府身份识别系统相结合,这引发了对隐私和社会控制的担忧。同时,针对人脸识别技术的对抗性研究也在不断发展,例如通过微小改动图像像素来欺骗人脸识别系统,以及将图像识别成其他人。 Sharon Zhou:AI系统对带口罩的人脸识别存在困难,这与数据集的偏差有关,需要改进数据集以提高鲁棒性,但这面临着数据收集的挑战。中国的人脸识别监控系统之所以强大,是因为手机等数据与个人身份紧密关联,且政府与企业合作密切。针对隐私保护,最好的方法不是删除信息,而是伪造信息,使其自相矛盾。此外,Algorithmic Justice League等组织致力于揭示和解决AI系统中的偏差和危害,并通过Algorithmic Vulnerability Bounty Project等项目鼓励人们报告AI系统中的偏差和危害。GPT-3等AI文本生成模型虽然可以生成逼真的人类文本,但其在处理复杂问题时仍存在局限性,容易被识别。

Deep Dive

Chapters
The episode discusses how AI is struggling to adapt to the changes brought by 2020, particularly with facial recognition and scene classification, and the challenges in creating unbiased datasets.

Shownotes Transcript

Stanford AI Lab PhDs Andrey Kurenkov and Sharon Zhou discuss this week's major AI news stories, relating to AI's struggles in 2020, Chinese AI surveillance, tools to fight facial recognition, and more!

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

Theme: Deliberate Thought Kevin MacLeod (incompetech.com)