A computer has no opinions, no upbringing, and no feelings about anyone. So it should be the fairest judge imaginable — and yet researchers keep documenting the opposite: face-recognition systems that misidentify darker-skinned faces far more often than lighter ones, voice assistants that understand some accents easily and others barely at all. Nobody wrote a line of code that says “work worse for these people.” The bias got in anyway, through three doors: the data the system learned from, the design choices about what to build and test, and the deployment — where the system gets used, and on whom.
Questions worth arguing about:
- A face-recognition system fails most often on the faces least like the ones in its training photos. Nobody meant it. Does “nobody meant it” matter to the person misidentified — and should it matter to the people who shipped it?
- When a voice assistant understands one accent better than another, somebody’s speech was treated as the default. Who decided whose voice was “normal” — and how would the builders even notice, if everyone on the team spoke the default?
- Suppose a biased system still makes fewer mistakes, on average, than the humans it replaces. Use it or not? Does your answer change depending on which group pays for the mistakes it still makes?
- What does “fixing it” actually involve — more diverse data, more diverse builders, testing on people unlike yourself, rules with teeth? Which of those can a Grade 10 programmer already practise?
- A tool that a blind or deaf user simply cannot operate — is that bias, or oversight? Is there a difference, from the outside?
The practical edge: Bias and Accessibility in Technology digs into how these failures work and what accessible design does about them. And when you design anything in this course, the habit starts now — before asking “does it work?”, ask “who does it work for, and who did I never test?” Per Our Classroom Norms, we argue this one with ideas, never at people — the point is to build better, not to find villains.
Curriculum connection
A2.4
investigate how to identify and address bias involving digital technology
Link to original
A2.5
analyze accessibility issues involving digital technology, and identify measures that can improve accessibility
Link to original