The Daily UbuntuFeatured
Errors that do not fall at random
When an algorithm gets certain faces wrong more often, it is not a neutral bug: it is an incomplete "us."
Read the articleThread 02 / 07 · Stories
Technology and dehumanization.
A machine learns from what we show it. And from what we forget to show it.
The thread
Algorithms, facial recognition, automated decisions: when technology reduces a person to data, and how to make it fairer.
14 articles · each in four movements: a scene, a documented case, a voice, a question
All articles · page 1 of 2
health
One shared breath
From Morton's skulls to spirometers "corrected" by race, science long served hierarchy; today it is learning to undo it.
Readtechnology
Robodebt: when the machine accuses
In Australia, an automated calculation wrongly claimed hundreds of thousands of debts from welfare recipients. A royal commission found it cruel and unlawful.
Readtechnology
A compass written together
In 2018, hundreds of Montrealers helped write ten principles so that artificial intelligence serves everyone, without discrimination.
Readtechnology
A resemblance is not proof
In Detroit, a Black father was wrongly arrested in front of his daughters because of software; his courage changed the police's rules.
Readtechnology
Teaching one another, even in code
In Montreal, a Mila program trains young Indigenous people in artificial intelligence so that tomorrow's technology carries their voices too.
Readtechnology
Measuring illness in dollars
A U.S. health algorithm judged patients' needs by their past spending; Black patients, who received less care, therefore seemed less sick.
Readtechnology
The machine does not refuse alone
Ottawa requires its departments to assess the risks of their algorithms before using them, and to keep a human in the loop when the answer is no.
Readtechnology
A tool too good to be honest
The RCMP used facial recognition software built on photos that the commissioners found had been collected illegally; even the police must know where their tools come from.
Readtechnology
The right to talk to a human
In Quebec, Law 25 requires businesses to say when a machine alone has made a decision, and to let the person explain themselves to someone.
Readtechnology
Three billion faces without permission
A company copied billions of faces online to sell them to police; four Canadian commissioners said no together.
Readtechnology
Draw me a judge
In 2023, a Bloomberg analysis showed that an image generator amplified gender and skin-colour stereotypes. A tool that makes images also makes places for people.
Readtechnology
The crop that chose a face
In 2020, users showed that Twitter's automatic cropping favoured certain faces; the company measured the bias, then gave the choice back to people.
Read