-
…in reply to @RealSaavedra
RealSaavedra Here we go, one of many examples: propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing AOC is 100% correct and there are many organizations across the world working on ways to mitigate bias in machine learning.
-
…in reply to @axbom
RealSaavedra AOC Algorithms are driven by computation, not by math. joanna-bryson.blogspot.com/2017/07/three-very-different-sources-of-bias-in.html
-
…in reply to @axbom
RealSaavedra AOC A video based on @mathbabedotorg's talk may help out here: vimeo.com/thersa/thetruthaboutalgorithms She wrote a book aptly named Weapons of Math Destruction.
-
…in reply to @axbom
RealSaavedra Hmmm, it seems even Amazon can't get it right... reuters.com/article/us-amazon-com-jobs-automation-insight/amazon-scraps-secret-ai-recruiting-tool-that-showed-bias-against-women-idUSKCN1MK08G
-
…in reply to @axbom
RealSaavedra Here's one where Google apologized: theguardian.com/technology/2015/jul/01/google-sorry-racist-auto-tag-photo-app
-
…in reply to @axbom
RealSaavedra Google says: "Machine learning models are not inherently objective. Engineers train models by feeding them data sets of training examples, and human involvement in the provision and curation of this data can make a model's predictions susceptible to bias." developers.google.com/machine-learning/crash-course/fairness/types-of-bias