LANDMARK · 5 MIN · ORIENT
Machine learning
After this landmark, you can explain why a machine-learning system is powerful and fragile for the same reason — it only knows what was in its data.
In traditional software, a person writes the rules: if this, then that. Machine learning flips that around — you show a system many examples, and it learns the patterns itself. That’s powerful, because it can pick up regularities no one could write down by hand, like what a cat looks like or which emails are spam. It’s also fragile, because the system only knows what was in its examples, and it learns correlations, not understanding. Change the world it was trained on, and it can be confidently wrong.
This is the canonical concept. It stays the same across learner lenses so personalization never changes the underlying facts.
◌ EVERYDAY LENS
What this looks like for you
When a tool “learns your taste” or flags a photo, it’s matching patterns from past examples, not reasoning. If your situation is unusual, expect it to miss — it has probably never seen your case.
MAKE A DECISION
A hiring model trained on your past hires keeps favouring one narrow profile. What’s the most useful reading?
CARRY THISPick one “smart” tool you use and name the examples it must have learned from. Then name a case those examples probably missed.