What if Algorithms Worked For Accused People, Instead of Against Them? | News & Commentary | American Civil Liberties Union
We created an algorithmic tool to find out what risks the criminal legal system poses to the people entering it, rather than their risk to "public safety."
The Algorithmic Imprint | Upol Ehsan, Ranjit Singh, Jacob Metcalf, Mark Riedl @ FAccT 2022
When an algorithm causes harm, is discontinuing it enough to address its harms?
This paper introduces the concept of the **The Algorithmic Imprint** to show how algorithmic harms can persist long after the algorithm is discontinued. It chronicles the 2020 Ofqual Algorithmic Grading Scandal, not from the UK, but from a Bangladeshi perspective. The concept of the Algorithmic Imprint helps us understand how the algorithm's impact lives on in the algorithm's afterlife much the remnants of palimpsest remain. Critically examining our current conception of algorithmic impact, it expands how we may view algorithmic impact, especially in the algorithm's afterlife (after being discontinued). It also offers practical and actionable guidance on how an imprint-aware mindset can inform algorithmic design.
This is a presentation of the paper "The Algorithmic Imprint" to be presented at the ACM Fairness, Accountability, and Transparency (FAccT) conference.
Here is a quick tweetorial covering the key points: https://bit.ly/AlgorthmicImprint_Tweetorial
Elon Musk's regulatory woes mount as U.S. moves closer to recalling Tesla's self-driving software
Traffic safety officials escalate and widen probe to 830,000 Tesla cars as they discover patterns suggesting Tesla's Autopilot feature can encourage dangerous driving behavior.
This is Fine: Optimism & Emergency in the P2P Network - A New Design Congress Essay
Centralised power and decentralised communities are on the verge of outright conflict for the control of the digital public space. The resilience of centralised networks and the political organisation of their owners remains significantly underestimated by protocol activists. At the same time, the peer-to-peer community is dangerously unprepared for a crisis-fuelled future that has very suddenly arrived at their door.
Leaked Audio From 80 Internal TikTok Meetings Shows That US User Data Has Been Repeatedly Accessed From China
“I feel like with these tools, there’s some backdoor to access user data in almost all of them,” said an external auditor hired to help TikTok close off Chinese access to sensitive information, like Americans’ birthdays and phone numbers.
Is "acceptably non-dystopian" self-sovereign identity even possible?
In the recent paper about "soulbound" tokens, Vitalik Buterin et al. write that they aim to build technology that is "acceptably non-dystopian". Do any of today's self-sovereign identity projects fit that bill?
Police VR Training: Empathy Machine or Expensive Distraction?
The maker of the Taser believes its new virtual reality firing range and community policing simulator can make cops less violent. Critics are less than sure.
Google collects almost 40 data points per user - most out of top tech giants - Atlas VPN
Data analyzed by Atlas VPN reveals that Google collects the most information about its users out of five selected tech giants - Google, Facebook, Twitter, Amazon, and Apple.
🇩🇪 2022(“Dear bathers, if your ticket does not come out of the machine, use the rubber mallet and hit the marked cross.”) https://t.co/uVhkiIsNHl pic.twitter.com/LjiUxz3epv— Stefan Plattner (@splattne) May 29, 2022
DuckDuckGo browser allows Microsoft trackers due to search agreement
The privacy-focused DuckDuckGo browser purposely allows Microsoft trackers on third-party sites due to an agreement in their syndicated search content contract between the two companies.