Artificial intelligence (AI) is everywhere: it is in our phone apps, in the websites we visit on the Internet, in many work tools... and it will be even more present in the future of the new generations.
Therefore, through the following activities, we will delve into basic concepts about AI, its applications to enhance the learning of our pupils and the knowledge they need to be able to use it correctly.
As we know, AI is the simulation of human intelligence by machines.
This allows them to perceive their environment, understand it, learn from it and act accordingly.
There are many types of AI and they can specialise in doing different tasks, for example, automating processes, recognising texts, images or people, or generating new content.
Machine learning: how does an artificial intelligence think?
Machine learning AIs categorise information by detecting patterns in text or images and follow the "instructions" that they have been trained with to think autonomously and to recognise objects or language in their environment.
In order for our pupils to understand that this mechanism is similar to our own cognitive processes, let's start with an activity that, without using any device, shows us how an AI thinks!
What planet are you from?
The pupils learn to work as artificial intelligence would in order to find out the planet of origin of different extraterrestrial creatures and help them to return to their home planet.
50 - 55 minutes
9 - 12 years old
Difficulty: intermediateFlowcharts: the logical thinking behind artificial intelligence
In so-called machine learning, the AI compares the data we offer it with the data it was trained with and it makes a decision based on the result. For example, if we train an AI to recognise tomatoes and we show it an orange, it will ask itself: is it rounded? yes; is it red? no; then it is not a tomato.
The basis of this process is a flowchart, but what is this and how can we teach it to our pupils? Try this activity!
Diamond adventures
The pupils discover what a flowchart is and go on an adventure through a video game where they learn how to read flowcharts in order to reach the treasure. As a reward, they play an online video game together.
50 - 55 minutes 9 - 12 years old Difficulty: lowImage recognition
Now that we understand how a machine learning AI works, let's try some!
In these activities we will see how an AI captures the image to be identified, extracts its characteristics, compares this new information with the one is on its database or is found on the Internet, and then it makes decisions about it, just as we have done with our pupils in previous activities!
Learning about our environment using AI
Learn how to use a tool that uses AI to recognise plants and animals.
55 - 60 minutes 9 - 12 years old Difficulty: intermediateWhat is what?
In this activity, pupils play a board game in which they try to identify objects using different techniques, in order to understand how artificial intelligence tools work in order to recognise objects from images.
50 - 60 minutes 9 - 12 years old Difficulty: lowArtificial intelligence and language
Another type of AI is Natural Language Processing (NLP), a system that allows computers to understand and use human language.
Some examples of these AIs are translators: they understand the meaning of words in one language and are able to translate them into another language autonomously. Let's see an example with this activity!
Mysterious words
Use of the image translator to discover the meaning of some words that we will have to draw in Scratch to finish a story.
55 - 60 minutes 6 - 8 years old Difficulty: intermediateArtificial intelligence as a source of information: chatbots
The most obvious use of Natural Language Processing AI is chatbots. We interact with them every day, but how do they work? By recognising language, autonomously searching the Internet for information and automatically learning from the conversations they have, they are able to generate answers to the questions we ask them.
Have we experienced some chatbots?
For the youngest pupils, we propose to discover how this type of AI works in a controlled environment, in which the chatbot only generates answers from a predetermined base of questions and answers, without access to the Internet.
Looking for the lost piece of art
Discovery of Frida Khalo interacting with a chatbot. The activity ends with a virtual visit to the artist's museum.
55 - 60 minutes 6 - 8 years old Difficulty: intermediate
With the older pupils, we will work with one of the best known chatbot in the world: Chat GPT. We will experiment with it, reflecting on how it works and the risks and opportunities it poses.
Chatting with AI: discover ChatGPT
The pupils learn about ChatGPT by discussing its use and limitations and interacting with this powerful chatbot.
65 - 70 minutes 9 - 12 years old Difficulty: lowCan artificial intelligence be creative?
So far, AI has been used to work on realities that already exist. However, AI can also be creative! Thanks to generative AI, we can use machine learning models to learn patterns in a series of texts, images, data, etc. and generate entirely new content.
Let's learn about this side of AI and discover how technology also impacts creativity and art.
Family virtual gallery
We will create works of art as a family using the Google Arts & Culture app. Furthermore, we will also create a virtual gallery with our drawings in peopleArt factory.
50 - 60 minutes 9 - 12 years old Difficulty: intermediateDo we use artificial intelligence to learn about emotions?
As we have seen, AI can recognise images, such as facial expressions. Therefore, it is possible to use this technology to help our pupils explore emotions and know how to identify them in their environment.
This is our proposal to deepen social-emotional learning thanks to AI, while they work in teams and develop their creativity.
Happy artificial intelligence
The activity involves team work on the expression of emotions and their recognition through mimics. An artificial intelligence capable of recognising drawings that represent these emotions is then introduced.
50 - 55 minutes 6 - 8 years old Difficulty: intermediateArtificial intelligence (AI) is a very useful tool that can help us in various situations in our lives. However, as with any other technology, it has some limitations, and it is important to be aware of them to be able to use it safely and critically.
In this section we will discuss the ethical dilemmas and biases posed by AI and we will work through the safety considerations associated with its use. Finally, we will see how to train our own AI!
Remember that you can find resources for working on the principles and basic uses of AI in the section AI 1: basic principles and applications.
What biases are there in AI and why are they so important?
In the context of AI, biases can be identified as the existence of systematic prejudices in the results that this technology provides. They can appear as discrimination in race, gender, age, or in many other areas…
To understand and mitigate biases in AI it is essential to ensure that automatic decisions are fair and equal for everyone.
Biases can occur in different stages of the development of an AI, for example, during its training: if the data used contains errors or does not adequately represent all groups, the AI may learn to perpetuate them, leading to unfair or discriminatory situations.
In this activity we will see an example: are our pupils capable of training an AI without bias?
The people of Raticulin
Pupils discover the training bias of artificial intelligences through a short video game and by searching for images that can represent human beings.
50 - 60 minutes 9 - 12 years old Difficulty: low
The biases may have serious consequences for people’s lives. For example, if an AI system used to contract staff were trained with data that reflected gender prejudice, it could favour one gender over another, perpetuating inequality.
In this activity we will analyse some of these situations in the classroom and we will see the importance of good training.
Uncovering bias in AI
In this activity we learn that AI, although often viewed as neutral or emotionless, can have biases.
45 - 60 minutes +12 years old Difficulty: IntermediateGenerative artificial intelligence: let’s be aware
One of the most cutting-edge types of AI is generative AI, which is capable of creating new content (texts, images, music, etc.,) from patterns that it has already learnt.
It is a very powerful tool but, should we still have blind-faith in it?
In the next activity we will try an AI image generator, and we will discuss the biases that this type of technology may present.
Bias Detectives
In this activity we explore how diverse AI really is. You will use a text-to-image generator to create pictures from different prompts and analyze the results to understand how AI can sometimes play favorites. Let's learn how to make AI fairer together!
45 - 60 minutes +12 years old Difficulty: IntermediateDeepfakes: appearances mislead
And speaking of generative AI…do your pupils know what a deepfake is? It is audiovisual content manipulated through this type of AI to make whoever sees the content believe that someone is doing or saying something that never really happened.
To explore more about this in the classroom, you can use this activity on deepfakes and their relationship with cyberbullying.
The story of a deepfake
Pupils make a timeline with the different stages of a cyberbullying case while reflecting on the use of deepfakes.
50 - 60 minutes 9 - 12 years old Difficulty: lowEthical dilemmas in the use of artificial intelligence
If a self-driving car, that is directed by AI, commits an offence: who is guilty? The AI? The person that programmed it? The owner of the vehicle? Is it ethical to use AI to do homework?
The use of artificial intelligence may be controversial, and it is important to develop our pupils’ critical thinking regarding it. In the following activities we will provide answers to these questions and to many more!
Interview with an AI
Pupils create a podcast on an ethical dilemma related to the use of AIs, interacting with an AI to ask its opinion.
50 - 60 minutes 9 - 12 years old Difficulty: intermediateExperiencing AI ethical dilemma
In this activity, students explore ethical dilemmas surrounding AI.
45 - 60 minutes +12 years old Difficulty: intermediateExploring ethical dilemmas in AI City
In this activity, you’ll explore a virtual 3D world which shows a city with different AI features and its challenges. You'll learn about various real-world scenarios where AI is used and consider the pros and cons that come with it.
45 - 60 minutes +12 years Difficulty: intermediatePersonalised content: why does the artificial intelligence know so much about me?
The AI learns from the models we show it. In some cases these models are our own data, our likes and preferences.
Is this positive or negative? Perhaps there is no answer to this question, but we can reflect on it in this activity.
The attractive algorithm
Pupils discover and discuss the ability of artificial intelligences to recommend personalised content to us while knowing what we like and by being able to access our data.
50 - 60 minutes 9 - 12 years old Difficulty: lowTraining artificial intelligences.
Now that we know the limits of AI and we know how to use it safely, we are going to train some AI!
We will generate our models; we will look at where they have succeeded and where they failed and then perfect them.
We suggest that you start with an activity to train an image recognition AI that can distinguish between cats and apples. Will it be able to do this?
In this activity, we use the tool Teachable Machine. If you want to learn more about it access our course.
Training an artificial intelligence
Pupils learn about artificial intelligences that recognise images and train one to recognise pictures of cats and apples found on the Internet.
50 - 55 minutes 9 - 12 years old Difficulty: high
The AI isn’t only able to train with images, but also with texts. In the next activity we will combine training safely with the digital world so that our AI tells us if we can or can’t publish specific personal information on the internet.
The Monitoring chat
Pupils train, test and implement an artificial intelligence on internet safety, capable of detecting if they expose data about who they are or where they live.
50 - 55 minutes 9 - 12 years old Difficulty: highArtificial intelligence used for the common good
From applications that help to predict natural disasters to those that improve medical attention, AI has become an incredibly powerful tool in helping humanity deal with global challenges and promote sustainable practices.
Below are two suggested activities for training waste recycling systems with AI, one of them is totally off line!
AI for Good: Understanding waste sorting through neural networks
In this activity, students will first gather some basic information about recycling and then work in small groups to sort waste manually.
45 - 60 minutes 9 - 12 years old Difficulty: intermediate
Eco-Tech: Waste separation with AI
In this activity, you’ll use Machine Learning (ML) to help sort waste.
45 - 60 minutes 9 - 12 years old Difficulty: intermediate
