filmov
tv
Technologies of Speculation: Data, AI and the Inequality of Facts - Sun-ha Hong (Simon Fraser Univ.)
Показать описание
BK FOUR "자유롭고 책임있는 AI 미디어" 교육연구단과 언론정보연구소에서 새로운 시리즈인 "Emerging Scholars Series"를 진행합니다. 첫번째 강연은 Simon Fraser University의 홍선하 교수님께서 맡아주셨습니다.
●주제: Technologies of Speculation: Data, AI, and the Inequality of Facts
●발표자: Sun-ha Hong 교수 (Simon Fraser University)
●초록: What counts as knowledge in the age of big data, AI and smart machines? From facial recognition to automated hiring systems, technology is reshaping what counts as suspicious, productive, valuable, in its own image. Yet in practice, data generates speculation as much as it does information. Landmark datasets are incomplete and partial, resulting in massively scaled reproduction of bias. Across schools to prisons to workplaces, algorithmic decision-making systems are often being used to intensify existing power asymmetries and prejudices.
In Technologies of Speculation (NYU Press, 2020), I trace these shifts by looking at large-scale government surveillance for counter-terrorism, and the ongoing rise of self-tracking and smart home systems. While data-driven systems often appear objective and impersonal from the bird’s eye view, a closer look often reveals new backdoors by which prejudice and inequality seep into the process. The ongoing intensification of data-driven surveillance further insists that a healthy, productive, and valuable subject is one that lives in machine-readable ways. For some, this abundance of data can seem an empowering source of knowledge; for others, to appear correctly in flawed databases can be the unhappy obligation on which their lives depend.
Normatively, I argue that this primacy of data-driven knowledge fundamentally undermines the mechanisms we have traditionally relied on to combat systemic harms. The emphasis on inhumanly large datasets and opaque algorithms cripples the effectiveness of transparency, for instance, and forces us to rethink the ideal of the informed, rational subject as the centerpiece of a good society. In the talk, I will end by connecting these issues to current conversations around AI ethics and Big Tech, and the limitations of ‘bias’ as a critical frame.
● 주최: 서울대학교 언론정보학과 BK21 FOUR 자유롭고 책임 있는 AI미디어 교육연구단, 서울대학교 언론정보연구소
●주제: Technologies of Speculation: Data, AI, and the Inequality of Facts
●발표자: Sun-ha Hong 교수 (Simon Fraser University)
●초록: What counts as knowledge in the age of big data, AI and smart machines? From facial recognition to automated hiring systems, technology is reshaping what counts as suspicious, productive, valuable, in its own image. Yet in practice, data generates speculation as much as it does information. Landmark datasets are incomplete and partial, resulting in massively scaled reproduction of bias. Across schools to prisons to workplaces, algorithmic decision-making systems are often being used to intensify existing power asymmetries and prejudices.
In Technologies of Speculation (NYU Press, 2020), I trace these shifts by looking at large-scale government surveillance for counter-terrorism, and the ongoing rise of self-tracking and smart home systems. While data-driven systems often appear objective and impersonal from the bird’s eye view, a closer look often reveals new backdoors by which prejudice and inequality seep into the process. The ongoing intensification of data-driven surveillance further insists that a healthy, productive, and valuable subject is one that lives in machine-readable ways. For some, this abundance of data can seem an empowering source of knowledge; for others, to appear correctly in flawed databases can be the unhappy obligation on which their lives depend.
Normatively, I argue that this primacy of data-driven knowledge fundamentally undermines the mechanisms we have traditionally relied on to combat systemic harms. The emphasis on inhumanly large datasets and opaque algorithms cripples the effectiveness of transparency, for instance, and forces us to rethink the ideal of the informed, rational subject as the centerpiece of a good society. In the talk, I will end by connecting these issues to current conversations around AI ethics and Big Tech, and the limitations of ‘bias’ as a critical frame.
● 주최: 서울대학교 언론정보학과 BK21 FOUR 자유롭고 책임 있는 AI미디어 교육연구단, 서울대학교 언론정보연구소
0:46:14
3:40:34
0:00:43
0:00:53
0:30:06
2:29:05
0:06:39
0:10:23
1:00:29
1:02:43
0:09:59
0:35:52
0:12:34
0:00:16
0:03:06
0:35:31
0:08:04
0:22:38
0:09:08
0:12:42
0:49:31
0:51:57
0:26:40
0:10:13