Skip to content
BENIBIX Learn

Artificial Intelligence (AI)

Level 1 · What AI is

What AI means, and why most AI today learns from examples instead of rules. What a model is, and the guess it hands back.

Course contentsLesson 1 of 10

After this lesson you can

  • Say what AI is in one sentence
  • Tell written rules from learned rules
  • Explain what a model gives back

10 mincovers 4 glossary entries

What artificial intelligence means

Artificial intelligence, or AI, is a computer doing a job that used to need a human mind. Reading, talking, seeing and planning are all such jobs. The name was first used in 1955, in a plan for a summer study of thinking machines. Today AI unlocks phones by face and turns speech into text.

There are two ways to build it. In the first, people write every rule by hand: if this, then that. In the second, the computer finds the rules itself, from lots of examples. Almost all AI in use today is built the second way. That way is called machine learning.

Finding the rules in examples is called pattern recognition. Show a program 10,000 photos marked cat or dog. It learns what the cats have in common, and what the dogs share. Then it can name the animal in a photo it has never seen. Spam filters and face unlock work the same way.

What learning leaves behind is a model. A model is a file full of numbers that hold the patterns it found. You give it an input, like a photo or a sentence. It gives back an output, like a label or a reply. The same model can be copied to a million phones.

The output is called a prediction, even when it is not about the future. A model that labels a photo cat is predicting the label. Most predictions come with a score for how sure the model is. A score of 0.97 means very sure. It does not mean right.

A model finds what goes together, not why. If every cow photo it saw stood on grass, it may learn that grass means cow. Then a cow on a beach can fool it. That is the main way pattern finding goes wrong, and it is why people still check the answers.

People often think AI understands the way a person does. It does not. It is very good at spotting patterns, and it can be wrong in a calm, sure voice. This course stays with the big picture. How a chat model reads and writes is a subject of its own, with its own words.

Photos with labels and lots of chip power give a pattern. It works on new photos too.

Worked example

Emma's phone sorts her photos

Emma has 4,000 photos on her phone. The photo app groups them by face, with no help from her.

The model inside never got a rule for her sister's face. It learned what faces look like from a huge pile of photos, then matched the ones that look alike.

The words for this lesson4

Check yourself

Does AI understand things the way people do?

Finished this lesson?

Marking it complete also marks its 4 glossary entries learned, which updates your roadmap and your saved page. 0 are already marked.