🐱 Good Examples, Bad Examples
What you show it is what it learns!
🌟 The Big Idea
Remember how an AI helper learns by seeing lots of examples? Well, here is the next secret: those examples need to be good ones!
If you show the helper wrong examples, it learns the wrong thing. If you only show it one kind of example, it only knows that one kind.
Think of it like this: if someone only ever showed you orange cats, you might think ALL cats are orange! A black cat would confuse you. The same thing happens to an AI helper.
There is a famous saying in computers: "Garbage in, garbage out." It means: bad stuff in → bad stuff out. Good, fair, mixed examples make a smart helper. Bad or one-sided examples make a confused one.
Illustrator brief: Draw a friendly cartoon robot sitting at a desk. On its left is a neat, colourful pile of cat photos (orange, black, white, stripy). On its right is a messy pile where some photos are upside-down, one dog is labelled "cat," and every cat is orange. The robot looks at the messy pile with a big question mark over its head and swirly eyes. Gentle, funny, not scary. Bright pastel colours.
Alt text: A cartoon robot looks confused at a messy pile of one-sided and mislabelled animal pictures.