Is generative AI the same thing as ChatGPT?

No. ChatGPT is the book's example of generative AI; generative AI is the category: AI built on deep learning that creates new data such as text, images, audio and programs. Image generators are the book's other example.

Why does the book say a neural network is modelled on the brain?

It is a comparison, not a claim that the machine works like a brain. The system connects many simple components so that, together, they learn from data and make complex judgements; the Arabic edition adds that the connections change during training.

What does the 6-mark question want from me?

Two halves: the relationship between AI, machine learning, deep learning and generative AI stated as one inside the other, and one example at each level. A list of four definitions side by side has not answered the question.

Lesson 1-2

How AI Works

Unit 1 — Information Technology and Society

Programming & Artificial Intelligence — Baccalaureate Year 2

Textbook pages pp. 12–18Published

What AI is, and how machine learning, deep learning and generative AI nest inside it: textbook pp. 12–18, with the 6-mark exam question explained.

Learning objectives

01Explain what artificial intelligence is and give examples of what it does on a computer.
02Explain how generative AI is positioned within AI technologies, using the idea of nested categories.
03Distinguish rules written by a person from patterns learned from data, and use the distinction to classify an application.
04Identify the risks of generative AI, including hallucination, and recommend a rule for its use with reasons.

Summary

The lesson starts from an ordinary phone: a spam filter sorts your mail, a shop recommends products, an app translates, a chatbot writes text from a prompt. All four are AI, but not the same kind, and telling them apart is the whole lesson. Artificial intelligence is the general term for technologies that reproduce intelligent human behaviour (learning, reasoning, judgement) on a computer. The book's margin adds a caution: today's AI is narrow, expert at one defined task.

The core is one figure of four nested boxes. Machine learning is one of the technologies that make AI work: it learns patterns from data instead of following rules a person wrote (spam filters, product recommendations). Deep learning is an advanced technology within machine learning that uses neural networks and needs large-scale data (image analysis for autonomous driving, speech synthesis). A neural network is modelled, in a simplified way, on the brain's connected nerve cells. Generative AI is built on deep learning and produces new data: text, images, audio or programs (ChatGPT, image generators).

The exam tests this nesting in every direction. Almost every wrong statement pulls one box out of the box that holds it, or collapses two into one: deep learning as separate from machine learning, or AI as another name for machine learning. The 6-mark exam-style question wants nested categories with one example at each level; an example at the wrong level earns nothing.

Two cautions close the lesson. Generative AI can hallucinate, producing text that sounds plausible and is factually wrong, so its output cannot be handed in unchecked. And a system trained on little data struggles with what it has rarely seen, which is why the farmer's crop-disease task asks what the AI must learn from. The Think-as-an-Engineer task then asks whether generative AI belongs in homework.

Key terms

Artificial intelligence (AI): The general term for technologies that reproduce intelligent human behaviour, such as learning, reasoning and judgement, on a computer; examples include speech recognition, image recognition and translation.
Machine learning (ML): One of the technologies that make AI work: it learns patterns from data to make predictions and judgements, instead of following rules programmed by hand.
Deep learning (DL): An advanced technology within machine learning that uses neural networks and learns complex patterns from large-scale data.
Neural network: A system modelled, in a simplified way, on the connected nerve cells of the brain: many simple components, connected together, that learn from data and make complex judgements.
Generative AI: AI built on deep learning that generates new data, such as text, images, audio or programs, from the patterns it learned in training.
Hallucination: Output from a generative AI that sounds plausible but is factually incorrect.
Narrow AI: The kind of AI that exists today: designed for one defined task or a limited set of tasks, not for everything a person can do.

The lesson on video

Unit 1 · Part 2 — How AI Works | Second Baccalaureate | English Schools

Watch on YouTube

Sample questions

01A classmate revises with the sentence: "Deep learning is a completely different technology from machine learning." Decide whether this sentence keeps the book's nesting of AI technologies, and if it does not, rewrite it correctly.Understand
02A school's attendance app marks a student late after 07:45, absent if they never arrive, and excused if a parent has sent a message. A programmer wrote each of these rules by hand. Is this app machine learning? Answer yes or no and justify your answer from the definition.Apply
03A supplier offers a road-sign recognition system to Egypt and says: "It reads signs with 99% accuracy. It was trained on two million photographs of European roads, so it will work just as well here." Judge this claim. State whether you accept it, and support your judgement with two ideas from this lesson.Evaluate

Answers are on the portal, graded there.

Questions students ask

How AI Works

Open this lesson on the portal

Back to the unit All units