PES
EN English 한국어 日本語 中文 हिन्दी বাংলা اردو Español Português Français Deutsch Русский العربية Bahasa Indonesia Tiếng Việt ไทย Türkçe

How to enable notifications

Notifications are unavailable in this browser or on this address. Try a supported browser on the secure PES site.

If notifications are blocked, allow them in your browser settings.

  1. Open the site controls beside the address bar, or your browser’s site settings. Set Notifications for this site to Allow.
  2. In your device settings, allow notifications for your browser and check Do Not Disturb or Focus.
  3. Return to PES, reload the page, and use the bell to enable notifications.

On iPhone or iPad, add PES to your Home Screen, open it there, then allow notifications.

My learning progress
  1. How to use AI
  2. Terminal commands
  3. Advanced AI
  4. Vibe coding
  5. AI certification Getting ready
  6. Keyboard
See details

Level 5

Know that an AI is a machine people build — not a living thing — and describe the one you want to build in your own words.

Who can apply

No age limit. Students under 12 can sit this level.

Required programming language

Required at every level, from Level 5 to Level 1.

Shell Scripting

The single most important language for working with AI.

The shell is useful for running programs, managing files, and automating development. AI can also run through browsers, apps, notebooks, and APIs. A terminal is one useful interface, not a requirement for every AI user.

Learn shell programming

The one question, at every level

You want to build an artificial intelligence completely from scratch. Design how you would build it.

Every level from 5 to 1 answers this same question. What changes is the depth expected of the answer, and the constraints attached to it.

What to learn

  1. Understand that an AI is logic made by people, not a creature that thinks or feels.
  2. Describe the AI you want to build, in words and in drawings.
  3. Name the kind of information your AI would need to work.

How you are marked

  • Did you say what you would build in one clear sentence?
  • Did you give examples of the information it needs?
  • Did you give a reason for building it that way?

Exam format

Stage 1 is a written paper at the venue; Stage 2 is an interview followed by a practical build. Both stages use the same question.

Stage 1 — Written, in person

Held at a designated exam venue. You write your answers on the exam paper by hand. This stage tests the design you can put into words.

Stage 2 — Interview and practical

An examiner interviews you about your design, then you build the for real. This stage tests whether the plan on paper actually works in your hands.

Practice questions

The official exam gives you just the one common question above. The 20 questions below are practice questions that check the words, names and facts you need to know before you write that design. Every one is a short-answer question with a single correct answer: read the description, write your answer, then open the answer to check. They are not official questions.

Basic AI words

  1. 1

    AI is short for two English words, taking the first letter of each. Write the two words in English.

    Show answer

    Answer Artificial Intelligence

    Artificial means "made by people" and Intelligence means "the ability to think and learn".

  2. 2

    What do we call a procedure that lists the steps for solving a problem in order, so that following those steps always leads to the same result?

    Show answer

    Answer Algorithm

    It is a procedure with fixed steps, like a recipe. A computer follows the algorithms people write, exactly as written.

  3. 3

    Instead of people writing every rule by hand, a computer is shown many examples and finds the rules by itself. What is this method called?

    Show answer

    Answer Machine learning

    For example: show a computer lots of spam and normal emails, and it finds the signs of spam by itself.

  4. 4

    What do we call the many examples (photos, text, sounds and so on) shown to an AI so that it can learn?

    Show answer

    Answer Training data

    An AI's answers resemble its training data. If it learns from wrong data, it gives wrong answers.

  5. 5

    Like writing "cat" on a cat photo and "dog" on a dog photo, what do we call the correct-answer tag attached to each piece of training data?

    Show answer

    Answer Label

    Learning to give the right answer from labelled data is called supervised learning.

Models, neural networks and vectors

  1. 6

    What do we call the finished result of training — it holds what was learned as numbers, and when it gets an input it calculates an output?

    Show answer

    Answer Model

    GPT, Claude and Gemini are all names of language models. Running a trained model to get an answer is called inference.

  2. 7

    It borrows its name and idea from the nerve cells (neurons) of the human brain, and connects simple calculating units in layers, like a net. What is this structure called?

    Show answer

    Answer Neural network (artificial neural network)

    Only the name comes from the brain. In reality it is a bundle of calculations that multiply and add numbers.

  3. 8

    What is the machine learning method that stacks very many layers of a neural network, deep, and learns with them?

    Show answer

    Answer Deep learning

    Deep means there are many layers. Most of today's AI is built on deep learning.

  4. 9

    What do we call a list of several numbers placed in a fixed order, like [0.2, -1.5, 3.0]?

    Show answer

    Answer Vector

    With 3 numbers it is a 3-dimensional vector. A computer cannot understand letters, pictures or sounds directly, so it turns them into lists of numbers like this to calculate with.

  5. 10

    What do we call turning the meaning of a word or sentence into a vector (a list of numbers), so that words with similar meanings, like "puppy" and "dog", get numbers that are close to each other?

    Show answer

    Answer Embedding

    The closer the meanings, the closer the vectors. That is why search, recommendations and language models use embeddings to compare meanings.

Language models and conversation

  1. 11

    What three-letter English abbreviation names a very large language model that has learned from a huge amount of text, understands what people say and continues sentences?

    Show answer

    Answer LLM (Large Language Model)

    Chat AIs such as ChatGPT, Claude and Gemini are all built on LLMs.

  2. 12

    When a language model processes text, it splits the text into small basic units — a whole word or a piece of a word. What is each unit called?

    Show answer

    Answer Token

    A language model looks at the tokens so far, calculates a score for the token that could come next, and writes the text one piece at a time.

  3. 13

    What do we call the question or instruction you type to get an AI, such as a chatbot, to do something?

    Show answer

    Answer Prompt

    The clearer your prompt, the easier it is to get the answer you want — but even a good prompt does not guarantee a correct answer.

  4. 14

    What do we call it when an AI makes up something that is untrue or has no basis, and says it as if it were true?

    Show answer

    Answer Hallucination

    It does not mean the machine really experiences hallucinations — it is just the name for output that sounds right but is wrong. Always check important things with another source.

  5. 15

    The GPT in ChatGPT is short for Generative Pre-trained T… The last T stands for the neural network design that most of today's language models are built on. Write its name in English.

    Show answer

    Answer Transformer

    The Transformer was first published in 2017 by Google researchers in the paper "Attention Is All You Need".

AI history, names and tools

  1. 16

    In 1950, which British mathematician asked "Can machines think?" and proposed a test (the imitation game) to tell whether a machine can hold a conversation like a person?

    Show answer

    Answer Alan Turing

    This test is called the Turing test. Talking like a person does not mean a machine has a mind.

  2. 17

    In March 2016, a Go-playing AI played five games against the Go player Lee Sedol and won 4 games to 1. What is its name?

    Show answer

    Answer AlphaGo

    It was made by Google DeepMind. Lee Sedol won the fourth game.

  3. 18

    ChatGPT was released in November 2022 and made chat AI known all over the world. Write the name of the company that made it.

    Show answer

    Answer OpenAI

    ChatGPT runs on OpenAI's GPT family of language models.

  4. 19

    Write the name of the company that made the AI model Claude.

    Show answer

    Answer Anthropic

    Every AI model has a company that made it. For example, Gemini was made by Google.

  5. 20

    In the shell (terminal), write the command that makes a new folder (directory).

    Show answer

    Answer mkdir

    It is short for make directory. For example, mkdir my-ai makes a folder called my-ai.

Try a practice exam The online AI knowledge exam is the quickest way to check where you stand before applying.

Guidelines

AI

AI Certification

Study

Vibe Coding

Game

Practice

Tutors

Chat

Forum

Create

Community

More

Glossary

Pick a word you don't know. You'll see a simple explanation and related terms.

No matching term yet.

Pick a term to see its explanation.