Data Mining (29)
Find narratives by ethical themes or by technologies.
FILTERreset filters-
- 15 min
- The App Solutions
Overview of recommender systems, which are information filtering algorithms design to suggest content or products to a particular user.
- The App Solutions
-
- 15 min
- The App Solutions
5 types of recommender systems and their impact on customer experience
Overview of recommender systems, which are information filtering algorithms design to suggest content or products to a particular user.
How do information filtering algorithms work and learn? Are some types of recommender systems more generally ethical than others?
-
- 8 min
- Kinolab
- 2019
Chris is a ride-share driver who has taken passenger Jaden hostage, with the conditions of release being that he is connected with Billy Bauer, the CEO of social media company Smithereen, for a conversation. While the London police attempt to deal with the situation through negotiation, the management team at Smithereen uses several data mining techniques, including analysis of Chris’s various social media pages and audio data streaming from his device, to provide the police with a valuable and complete profile on Chris.
- Kinolab
- 2019
-
- 8 min
- Kinolab
- 2019
Social Media Data and Cooperation with Law Enforcement
Chris is a ride-share driver who has taken passenger Jaden hostage, with the conditions of release being that he is connected with Billy Bauer, the CEO of social media company Smithereen, for a conversation. While the London police attempt to deal with the situation through negotiation, the management team at Smithereen uses several data mining techniques, including analysis of Chris’s various social media pages and audio data streaming from his device, to provide the police with a valuable and complete profile on Chris.
Are things said on social media fair game for law enforcement to use against a person? Does this include data that a user might not even know a company has gathered on them? How might “abstractions” of a user formed by a social media company be misused to make a bad judgement about a person? Should social media information and profiles of users made by companies be used in attempts to stop crime or criminals before any wrongdoings are committed? What are the dangers of big data companies having a close relationship with law enforcement?
Black Mirror: Smithereens and the FBI-
- 10 min
- MIT Technology Review
- 2020
This article explains the ethical warnings of Timnit Gebru against training Natural Language Processing algorithms on large language models developed on sets of textual data from the internet. Not only does this process have a negative environmental impact, it also still does not allow these machine learning tools to process semantic nuance, especially as it relates to burgeoning social movements or countries with lower internet access. Dr. Gebru’s refusal to retract this paper ultimately lead to her dismissal from Google.
- MIT Technology Review
- 2020
-
- 10 min
- MIT Technology Review
- 2020
We read the paper that forced Timnit Gebru out of Google. Here’s what it says.
This article explains the ethical warnings of Timnit Gebru against training Natural Language Processing algorithms on large language models developed on sets of textual data from the internet. Not only does this process have a negative environmental impact, it also still does not allow these machine learning tools to process semantic nuance, especially as it relates to burgeoning social movements or countries with lower internet access. Dr. Gebru’s refusal to retract this paper ultimately lead to her dismissal from Google.
How should models for training NLP algorithms be more closely scrutinized? What sorts of voices are needed at the design table to ensure that the impact of such algorithms are consistent across all populations? Can this ever be achieved?
-
- 5 min
- Inc
- 2021
On International Data Privacy Day, Apple CEO Tim Cook fired shots against Mark Zuckerberg and Facebook’s model of mining user data through platform analytics and web mining to serve up targeted ads to users. By contrast, Cook painted Apple as a privacy oriented company who wants to make technology work for its users by not collecting their data and manipulating them psychologically through advertising.
- Inc
- 2021
-
- 5 min
- Inc
- 2021
Tim Cook May Have Just Ended Facebook
On International Data Privacy Day, Apple CEO Tim Cook fired shots against Mark Zuckerberg and Facebook’s model of mining user data through platform analytics and web mining to serve up targeted ads to users. By contrast, Cook painted Apple as a privacy oriented company who wants to make technology work for its users by not collecting their data and manipulating them psychologically through advertising.
Are you convinced that Apple has a better business model than Facebook? Should users be responsible for taking steps to protect themselves against web mining, or should Facebook be responsible for adding in more guardrails? What are the consequences of both Facebook and Apple products being involved in larger architectures that extend beyond the singular digital artifact?
-
- 3 min
- CNN
- 2021
The prominence of social data on any given person afforded by digital artifacts, such as social media posts and text messages, can be used to train a new algorithm patented by Microsoft to create a chatbot meant to imitate that specific person. This technology has not been released, however, due to its harrowing ethical implications of impersonation and dissonance. For the Black Mirror episode referenced in the article, see the narratives “Martha and Ash Parts I and II.”
- CNN
- 2021
-
- 3 min
- CNN
- 2021
Microsoft patented a chatbot that would let you talk to dead people. It was too disturbing for production
The prominence of social data on any given person afforded by digital artifacts, such as social media posts and text messages, can be used to train a new algorithm patented by Microsoft to create a chatbot meant to imitate that specific person. This technology has not been released, however, due to its harrowing ethical implications of impersonation and dissonance. For the Black Mirror episode referenced in the article, see the narratives “Martha and Ash Parts I and II.”
How do humans control their identity when it can be replicated through machine learning? What sorts of quirks and mannerisms are unique to humans and cannot be replicated by an algorithm?
-
- 7 min
- Wired
- 2021
An anonymous college student created a website titled “Faces of the Riot,” a virtual wall containing over 6,000 face images of insurrectionists present at the riot at the Capitol on January 6th, 2021. The ultimate goal of the creator’s site, which used facial recognition algorithms to crawl through videos posted to the right-wing social media site Parler, is to hopefully have viewers identify any criminals that they recognize to the proper authorities. While the creator put safeguards for privacy in place, such as using “facial detection” rather than “facial recognition”, and their intentions are supposedly positive, some argue that the implications on privacy and the widespread integration of this technique could be negative.
- Wired
- 2021
-
- 7 min
- Wired
- 2021
This Site Published Every Face From Parler’s Capitol Riot Videos
An anonymous college student created a website titled “Faces of the Riot,” a virtual wall containing over 6,000 face images of insurrectionists present at the riot at the Capitol on January 6th, 2021. The ultimate goal of the creator’s site, which used facial recognition algorithms to crawl through videos posted to the right-wing social media site Parler, is to hopefully have viewers identify any criminals that they recognize to the proper authorities. While the creator put safeguards for privacy in place, such as using “facial detection” rather than “facial recognition”, and their intentions are supposedly positive, some argue that the implications on privacy and the widespread integration of this technique could be negative.
Who deserves to be protected from having shameful data about themselves posted publicly to the internet? Should there even be any limits on this? What would happen if a similar website appeared in a less seemingly noble context, such as identifying members of a minority group in a certain area? How could sites like this expand the agency of bad or discriminatory actors?