Posted by & filed under AI/Artificial Intelligence, Ethical issues.

ImageNetRoulette

An artificial-intelligence art project has been criticised for using racist and sexist tags to classify its users.

When they share a selfie with ImageNet Roulette, the web app matches it to the ones it most closely resembles from an enormous library of profile photos.

It then reveals the most popular tag, assigned to the matching pictures by human workers using data set WordNet.

These include racial slurs, “first offender”, “rape suspect”, “spree killer”, “newsreader”, and “Batman”.

Source: BBC Technology News

Date: September 17th, 2019

Link: https://www.bbc.com/news/technology-49726652

Discussion

  1. The title of this article suggests that AI is being racist. However, it turns out that the photos/art are being labelled by humans, sourced from Amazon’s Mechanical Turk. What is the problem with using humans to label things for AI?
  2. How can AI be trained to be correct without human input?
  3. Can AI that has trained itself still be racist?

Leave a Reply

Your email address will not be published. Required fields are marked *