Unable to load image
/h/femaledatingstrategy House Vampire
asshole there/then (>^.^(>O.○)>  5mo ago (thepinknews.com) 465 thread views #273916

Lesbian :marseyrowling:TERF's app to 'use facial recognition' to exclude trans :marseypass: users

https://www.thepinknews.com/2024/06/03/lesbian-app-facial-recognition-technology-trans/

Actual womxn malding

https://old.reddit.com/r/TwoXChromosomes/comments/1d71oa9/in_todays_terrible_news_a_new_lesbian_dating_app/

72
New Old Top Bottom Controversial Random
Jump in the discussion.

No email address required.

View entire discussion

A lot of biofoids are gonna be taking the L to end all Ls because of this.

:#marseysigh:

:#marseydisintegrate: :!#marseyflamewar::space::!marseyagree:

Jump in the discussion.

No email address required.

this happened already in straya

In Australia, women-only app becomes latest front in war over trans rights

The founder of a women-only social networking app is being sued for denying her service to a trans woman.

https://www.aljazeera.com/news/2024/4/22/in-australia-a-women-only-app-is-latest-front-in-the-war-over-trans-rights

skip to min 20 for the ai scanning !transphobes

Jump in the discussion.

No email address required.

AI might actually kill the "right to service" BS because you can deny people and then say the AI is a blackbox and you just trained it on an unrelated parallel. How can they ever prove that you intended to discriminate

Jump in the discussion.

No email address required.

https://www.britannica.com/topic/disparate-impact

Jump in the discussion.

No email address required.

yeah but literally first paragraph in ur link:

However, civil rights advocates have been disappointed as federal courts have increasingly limited how and when plaintiffs may file disparate-impact claims. As a result, disparate-impact suits have become less successful over time.

Jump in the discussion.

No email address required.

How can they ever prove that you intended to discriminate

if you hurt someones feelings then you discriminated

https://media.giphy.com/media/b9w56508q1nZxznfnV/giphy.webp

Jump in the discussion.

No email address required.

Someone can very easily test for this by simply feeding a bunch of people in "in-group" vs "out-group" to the AI and see how it performs to prove its bias against something though. People already do this (prior to AI release) in stuff like medical AI to make sure that it doesn't perform particularly poorly for minorities and stuff.

Legislation may not be here yet for such a case but I'm sure it would make a pretty easy discrimination suit.

Jump in the discussion.

No email address required.

Link copied to clipboard
Action successful!
Error, please refresh the page and try again.