AI & Machine LearningAI Ethics

The Biased Mirror

AI learns the unfair patterns hidden in its training data

A computer has no grudges and no favorites. So how can a hiring program end up treating two equal people differently?

You'll find out

  • Where a model gets its habits
  • Why deleting a column isn't enough
  • Who face scanners got wrong

Why it matters

Programs now help pick who gets a job, a loan or a second look, so their blind spots land on real people.

About 3 min6 interactive slides · 5 questionsLesson 210 · Zone 24

Play this lesson

Free to play in your browser or on iPhone

More in this zone

  1. 208Digital Audio44,100 Snapshots a SecondDigital audio is a string of measurements, not a wave
  2. 209Internet HistoryWhy Email Uses @The first email's sender picked the one key nobody used
  3. 211Perception ScienceThe Rubber Hand IllusionYour brain can adopt a fake hand in minutes
  4. 212Orbital MechanicsSpace JunkOne collision can create thousands of bullets in orbit
  5. 213Image CompressionHow JPEG Shrinks PhotosPhotos shrink by throwing away detail your eyes barely notice
  6. 214TopologyTwo Opposite Places With the Same WeatherRight now two opposite points on Earth share temperature and pressure
  7. 215Web ArchivingThe Internet's AtticHundreds of billions of old web pages are saved in one archive
  8. 216Wireless SecurityThe Evil TwinA fake hotspot can borrow the name of one you trust