Week 1: Introduction to AI Literacy

Week 1: AI Is Already Around Us — Introduction to AI Literacy

Welcome to Week 1 of the AI Literacy Course.

This week introduces the meaning of artificial intelligence, why AI literacy matters, and how AI is already part of everyday life, education, work, and society.

Learning goals

By the end of this lesson, learners should be able to:

• Recognise examples of AI-supported systems in everyday life.
• Distinguish ordinary automation from systems that recognise, predict or generate.
• Explain why recommendations and AI outputs aren’t necessarily neutral.
• Identify the purpose, possible data, benefits and risks of an AI-supported system.
• Adjust the level of verification to the possible consequences of an error.
• Explain why human judgement and responsibility remain necessary.

1. Not Every Digital System Is AI

Not every digital system uses artificial intelligence. A timer, calculator or traditional washing machine usually follows fixed instructions. This is automation, but it isn’t necessarily AI.

AI-supported systems typically use data and trained models to:

• Recognise patterns, such as speech or spam.
• Predict outcomes, such as travel time or recommendations.
• Generate content, such as text, images or music.

Some products combine automation and AI, so the difference isn’t always obvious. Instead of trusting an “AI-powered” label, ask: What does the system actually do?

2. What Is Artificial Intelligence?

Artificial intelligence, usually called AI, refers to computer systems that use data and learned patterns to perform tasks such as:

• Recognising images, speech or other patterns
• Making predictions and recommendations
• Understanding or generating text
• Translating languages
• Producing images, music or video
• Supporting decisions

AI doesn’t think or understand the world in the same way a person does. For example, a text-generating AI can produce a convincing answer because it has learned patterns from large amounts of language. This doesn’t mean it understands the answer or that the information is necessarily correct.

3. Where Do We Meet AI?

AI-supported systems are already present in many parts of everyday life. Examples include:

• Search engines and translation tools

• Navigation and travel-time predictions

• Spam and fraud detection

• Social media and streaming recommendations

• Voice assistants and chatbots

• Image, audio and video tools

• Online shopping recommendations

Many people use these systems without noticing that AI may be involved.

Discussion question: Which AI-supported systems have you used today?

4. Recognition, Prediction, and Generation

Many AI-supported systems perform one or more of these functions:

• Recognition: identifying patterns, such as speech, faces or spam.

• Prediction: estimating an outcome, such as travel time or what a user may prefer.

• Generation: producing content, such as text, images, music, video or code.

A single service may combine several functions. For example, a social media platform may recognise content, predict what interests a user and generate automatic captions.

5. What Is AI Literacy?

AI literacy means having the knowledge and skills needed to understand, use, question and evaluate AI-supported systems.

An AI-literate person should ask:

• What does this system do?

• What data might it use?

• Who benefits from it?

• What could go wrong?

• Can I trust its output?

• Who is responsible for the final decision?

AI literacy isn’t only about knowing how to operate AI tools. It also involves critical thinking, source criticism, privacy awareness and human judgement.

6. Data and Recommendations Aren’t Neutral

AI systems learn patterns from data, but data can be incomplete, outdated or unevenly represent different groups. This means a system may work better for some people or situations than for others.

Recommendations aren’t neutral either. A platform may select content to increase clicks, purchases or time spent online. By repeatedly showing some options and hiding others, it may influence what users notice and prefer.

This is why we should ask what goal the system is designed to achieve and whether that goal matches our own needs.

7. Use Calibrated Trust

AI can be helpful, but usefulness isn’t the same as reliability. A chatbot may produce a useful outline and still invent a source. A translation tool may communicate the general meaning while missing an important detail.

Calibrated trust means adjusting verification to the purpose, available evidence and possible harm.

For example:

• A music recommendation usually requires little checking.

• A study summary should be compared with the original source.

• Medical, legal or financial information requires authoritative sources and qualified human support.

The higher the possible consequences of an error, the stronger the verification should be.

8. Notice, Question, Evaluate, and Act

A practical AI-literacy habit has four steps:

1. Notice: Identify where AI may be involved.

2. Question: Ask what the system does and what data it may use.

3. Evaluate: Consider the benefits, risks, evidence and people affected.

4. Act: Verify the output, change a setting, seek human support or challenge a decision.

The goal isn’t to trust or reject every AI system. The goal is to make an informed decision based on the situation.

9. Classroom Activity: Map One Ordinary Day

Choose five digital services you used during an ordinary day. For each service, answer:

1. What does the system do?

2. Does it recognise, predict or generate?

3. What data might it use?

4. Who benefits from it?

5. What could go wrong?

6. Who is responsible for its use or decisions?

You may include services such as YouTube, Netflix, Google Maps, social media, online shopping, translation tools or chatbots.

Don’t include passwords, identification numbers, health information or other sensitive personal data.

10. Reflection Questions

Write short answers to the following questions:

1. Which AI-supported system do you use most often?

2. How is AI different from ordinary automation?

3. Can a useful AI output still be wrong? Give an example.

4. How can recommendations influence what people see or choose?

5. When should an AI output receive stronger verification?

6. What responsibility remains with the person using AI?

11. Key Vocabulary

Automation:
A system carrying out defined instructions without a person repeating every step.

Artificial intelligence:
Computer systems that use data and learned patterns to recognise, predict, generate or support decisions.

Recognition:
Identifying or classifying patterns in data.

Prediction:
Estimating a possible outcome based on data and patterns.

Generation:
Producing new content such as text, images, audio, video or code.

Recommendation system:
A system that selects and ranks content or products it predicts a user may engage with.

Calibrated trust:
Adjusting verification according to the purpose, evidence and possible consequences of an error.

Human oversight:
A person understanding the system’s role, checking important outputs and having the ability to reject or change a result.

12. Optional Homework

During the week, identify three AI-supported systems you encounter.

For each example, write:

• What the system does

• Whether it recognises, predicts or generates

• One possible benefit

• One possible risk

• How carefully its output should be checked

13. Summary

AI is already part of everyday life, but not every digital system is AI. AI-supported systems commonly recognise patterns, make predictions or generate content.

These systems can be useful, but their data and outputs aren’t necessarily neutral or correct. Responsible use means noticing where AI is involved, asking how it works, evaluating its benefits and risks, and deciding what action or verification is needed.

The central lesson is simple:

Use AI as a tool, not as a replacement for human judgement.

Lesson 1 Interactive Quiz: AI Is Already Around Us

Choose one answer for each question. Then click Check my answers. This self-checking quiz does not collect names or scores.

1. Which statement best describes ordinary automation?

A. A system learns a new task independently every time it is used B. A system carries out defined instructions without a person repeating every step C. A system always predicts what will happen in the future D. A system generates new content from training data

2. Which question is most useful when a product is described as “AI-powered”?

A. Does the product connect to the internet? B. Is the product newer than similar products? C. What does the system actually do? D. Does the system complete its task quickly?

3. Which example mainly involves prediction?

A. A fixed alarm rings every morning at 07:00 B. A navigation app estimates tomorrow’s travel time C. An image generator creates a picture from a description D. A calculator follows a rule to add two numbers

4. Why can recommendation systems influence people’s preferences?

A. They provide a neutral list of every available option B. They repeatedly make some options more visible than others C. They show content without using information about user behaviour D. They always select what is most useful for the user

5. Why should AI data and outputs not automatically be considered neutral?

A. Digital data cannot contain information about people B. AI systems ignore all patterns in their training data C. Data may be incomplete, outdated or unevenly represent different groups D. Every AI output is deliberately designed to be unfair

6. What does calibrated trust mean?

A. Trusting an AI system after it has produced one correct answer B. Adjusting verification to the purpose, evidence and possible harm C. Using the same level of checking for every AI-supported task D. Avoiding AI whenever an answer cannot be completely guaranteed

7. Which situation requires the strongest verification?

A. Asking for ideas for a party game B. Creating a personal music playlist C. Seeking information about an important legal right D. Comparing colour suggestions for a poster

8. Which sequence represents the four-step AI-literacy habit used in this lesson?

A. Search, copy, submit and forget B. Accept, improve, publish and defend C. Predict, generate, automate and recommend D. Notice, question, evaluate and act

9. Which example demonstrates meaningful human oversight?

A. A person approves every AI suggestion without examining it B. A person understands the system’s role, checks important outputs and can reject them C. A person assumes that responsibility belongs entirely to the developer D. A person checks only whether the output is clearly written

10. What is the main purpose of creating an everyday AI map?

A. To decide whether all AI-supported services should be avoided B. To create a complete technical description of every AI model C. To examine purpose, possible data, benefits, risks and responsibility D. To rank digital services according to how modern they appear

Reflection: After checking your answers, identify one AI-supported system you now notice, one reason its output may require verification and one responsibility you have when using it.