Do I have to memorise which industry each example belongs to?

Yes — healthcare, agriculture, manufacturing and logistics each have their own example, and the exam asks you to pair them. The trap is predictive maintenance: it predicts, but it belongs to manufacturing, not healthcare.

What is the difference between the black-box problem and hallucination?

The black box is about the process: nobody can see how the judgement was reached. Hallucination is about the output: the content differs from the facts while sounding right. One hides the reasoning, the other gives you a wrong answer that looks like a correct one.

What does the 6-mark question want from me?

Not a list. It wants the underlying reason: AI is strong where the task is finding a pattern in data, and needs a human where the decision carries responsibility, ethics or privacy — and the higher the impact of an error, the more the human check matters.

Lesson 1-3

AI in Daily Life and Industry

Unit 1 — Information Technology and Society

Programming & Artificial Intelligence — Baccalaureate Year 2

Textbook pages pp. 19–25Published

Where AI already works — recommendations, voice assistants, translation, face recognition, and four industries — and where it needs a human: textbook pp. 19–25, with the caution list the exam is built on.

Learning objectives

01Give examples of how AI is used in daily life and in different industries, and say what the AI actually does in each.
02Explain what AI is good at and what requires caution when it is used.
03Match an example of AI use to the industry it belongs to, and name the term for it.
04Judge a specific use of AI and justify why human oversight is needed when the impact of an error is high.

Summary

The lesson opens with four services a student already uses without calling them AI: a video site that suggests what to watch next, an assistant that obeys a spoken command, an app that translates a sentence, a phone that unlocks at a glance. Each is named and defined in one line — a recommendation system predicts preferences from past behaviour, a voice assistant recognises a voice and executes a command, machine translation converts text between languages, face recognition detects and identifies a face in an image. Four examples, four different verbs; the exam swaps them to see whether the verb was learned or only the name.

Industry is the second half, in four fields. Healthcare reads scans: image-diagnosis AI detects disease from X-ray and CT images. Agriculture monitors fields and predicts harvests. Manufacturing inspects products and predicts failures before they happen, which has its own name — predictive maintenance. Logistics optimises delivery routes. The pairing of example to industry is asked directly, and predictive maintenance is the one students most often file under healthcare because it contains the word 'predict'.

Then the lesson divides everything in two. What AI is good at: finding and classifying patterns in complex data such as images and text, recognising and generating images, audio and text, and making probabilistic predictions from data. What requires caution: ethical judgements, which may lead to discrimination; anything touching personal information and privacy; final decision-making, because someone must answer for the result; and accuracy when the training data is biased. Two problems get their own names — the black-box problem, when it is unclear how an AI reached its judgement, and hallucination, when it generates content that differs from the facts while sounding convincing.

The closing idea is a scale, not a switch: the higher the impact of a decision, the more the human check matters. That sentence answers the 6-mark question, and it is also what separates a full answer from a list. The same recommendation system that suggests a video is a different matter when it suggests which students enter an advanced class — the technology has not changed, the stakes have.

Key terms

Recommendation system: A system that predicts preferences from past behaviour data and displays recommendations, as a video site or an online shop does.
Voice assistant: AI that recognises a voice, understands the command in it and executes that command.
Machine translation: AI that automatically translates text into a different language.
Face recognition: AI that automatically detects a human face in an image and identifies whose it is; unlocking a phone is the everyday example.
Image-diagnosis AI: Healthcare AI that detects disease from X-ray and CT images. It flags what a human might miss; the diagnosis itself stays with the doctor.
Predictive maintenance: In manufacturing, using data to predict a machine or product failure before it happens, so it can be dealt with in advance.
Black-box problem: When it is unclear how an AI reached its judgement, so nobody can check whether the result is right or ask why it was made.
Hallucination: When an AI generates content that differs from the facts while sounding convincing. Nothing looks broken — that is what makes it dangerous.

The lesson on video

Unit 1 · Part 3 — AI in Daily Life and Industry | Second Baccalaureate | English Schools

Watch on YouTube

Sample questions

01A factory fits sensors to its machines and uses AI to warn which machine will break down next week. Name the term for this use, say which industry it belongs to, and state which strength of AI it rests on.Apply
02A school library wants AI to recommend books to students from what they borrowed before. Explain how the recommendation would be produced, and name the one caution the school should weigh before switching it on.Analyse
03A hospital manager argues: "Our image-diagnosis AI is more accurate than our doctors, so asking a doctor to confirm every diagnosis wastes time." Judge the argument, and support your judgement with two reasons.Evaluate

Answers are on the portal, graded there.

Questions students ask

AI in Daily Life and Industry

Open this lesson on the portal

Back to the unit All units