AI (143)
Find narratives by ethical themes or by technologies.
FILTERreset filters-
- 6 min
- Wired
- 2019
Spreading of harmful content through Youtube’s AI recommendation engine algorithm. AI helps create filter bubbles and echo chambers. Limited user agency to be exposed to certain content.
- Wired
- 2019
-
- 6 min
- Wired
- 2019
The Toxic Potential of YouTube’s Feedback Loop
Spreading of harmful content through Youtube’s AI recommendation engine algorithm. AI helps create filter bubbles and echo chambers. Limited user agency to be exposed to certain content.
How much agency do we have over the content we are shown in our digital artifacts? Who decides this? How skeptical should we be of recommender systems?
-
- 7 min
- New York Times
- 2018
This article details the research of Joy Buolamwini on racial bias coded into algorithms, specifically facial recognition programs. When auditing facial recognition software from several large companies such as IBM and Face++, she found that they are far worse at properly identifying darker skinned faces. Overall, this reveals that facial analysis and recognition programs are in need of exterior systems of accountability.
- New York Times
- 2018
-
- 7 min
- New York Times
- 2018
Facial Recognition Is Accurate, if You’re a White Guy
This article details the research of Joy Buolamwini on racial bias coded into algorithms, specifically facial recognition programs. When auditing facial recognition software from several large companies such as IBM and Face++, she found that they are far worse at properly identifying darker skinned faces. Overall, this reveals that facial analysis and recognition programs are in need of exterior systems of accountability.
What does exterior accountability for facial recognition software look like, and what should it look like? How and why does racial bias get coded into technology, whether explicitly or implicitly?
-
- 10 min
- The Washington Post
- 2019
After prolonged discussion on the effect of “bots,” or automated accounts on social networks, interfering with the electoral process in America in 2016, many worries surfaced that something similar could happen in 2020. This article details the shifts in strategy for using bots to manipulate political conversations online, from techniques like Inorganic Coordinated Activity or hashtag hijacking. Overall, some bot manipulation in political discourse is to be expected, but when used effectively these algorithmic tools still have to power to shape conversations to the will of their deployers.
- The Washington Post
- 2019
-
- 10 min
- The Washington Post
- 2019
Are ‘bots’ manipulating the 2020 conversation? Here’s what’s changed since 2016.
After prolonged discussion on the effect of “bots,” or automated accounts on social networks, interfering with the electoral process in America in 2016, many worries surfaced that something similar could happen in 2020. This article details the shifts in strategy for using bots to manipulate political conversations online, from techniques like Inorganic Coordinated Activity or hashtag hijacking. Overall, some bot manipulation in political discourse is to be expected, but when used effectively these algorithmic tools still have to power to shape conversations to the will of their deployers.
How are social media networks architectures that can be manipulated to an individual’s agenda, and how could this be addressed? Should any kind of bot accounts be allowed on Twitter, or do they all have too much negative potential to be trusted? What affordances of social networks allow bad actors to redirect the traffic of these networks? Is the problem of “trends” or “cascades” inherent to social media?
-
- 7 min
- Farnam Street Blog
- 2021
Discusses the main lessons from two recent books explaining how algorithmic bias occurs and how it may be ameliorated. Essentially, algorithms are little more than mathematical operations, but their lack of transparency and the bad, unrepresentative data sets which train them mean their pervasive use becomes dangerous.
- Farnam Street Blog
- 2021
-
- 7 min
- Farnam Street Blog
- 2021
A Primer on Algorithms and Bias
Discusses the main lessons from two recent books explaining how algorithmic bias occurs and how it may be ameliorated. Essentially, algorithms are little more than mathematical operations, but their lack of transparency and the bad, unrepresentative data sets which train them mean their pervasive use becomes dangerous.
How can data sets fed to algorithms be properly verified? What would the most beneficial collaboration between humans and algorithms look like?
-
- 5 min
- Time
- 2021
In 2021, former Facebook employee and whistleblower Frances Haugen testified to the fact that Facebook knew how its products harmed teenagers in terms of body image and social comparison; yet because of their interest in their profit model, they do not significantly attempt to ameliorate these harms. This article provides four key lessons to learn from how Facebook’s model is harmful.
- Time
- 2021
-
- 5 min
- Time
- 2021
4 Big Takeaways From the Facebook Whistleblower Congressional Hearing
In 2021, former Facebook employee and whistleblower Frances Haugen testified to the fact that Facebook knew how its products harmed teenagers in terms of body image and social comparison; yet because of their interest in their profit model, they do not significantly attempt to ameliorate these harms. This article provides four key lessons to learn from how Facebook’s model is harmful.
How does social quantification result in negative self-conception? How are the environments of social media platforms more harmful in terms of body image or “role models” than in-person environments? What are the dangers of every person having easy access to a broad platform of communication in terms of forming models of perfection? Why do social media algorithms want to feed users increasingly extreme content?
-
- 6 min
- CBS News
- 2021
In light of the recent allegations of Facebook whistleblower Frances Haugen that the platform irresponsibly breeds division and mental health issues, AI Specialist Karen Hao explains how Facebook’s “algorithm(s)” serve or fail the people who use them. Specifically, the profit motive and a lack of exact and comprehensive knowledge of the algorithm system prevents groundbreaking change from being made.
- CBS News
- 2021
-
- 6 min
- CBS News
- 2021
Facebook algorithm called into question after whistleblower testimony calls it dangerous
In light of the recent allegations of Facebook whistleblower Frances Haugen that the platform irresponsibly breeds division and mental health issues, AI Specialist Karen Hao explains how Facebook’s “algorithm(s)” serve or fail the people who use them. Specifically, the profit motive and a lack of exact and comprehensive knowledge of the algorithm system prevents groundbreaking change from being made.
Do programmers and other technological minds have a responsibility to understand exactly how algorithms work and how they tag data? What are specific consequences to algorithms which use their own criteria to tag items? How do social media networks take advantage of human attention?
: