AI4STEM: AI & Educational Robotics

• AI4STEM Curriculum focuses on the application of AI in educational robotics.
• It introduces the 5 Big Ideas in Artificial Intelligence using the Internet of Things (IoT) in STEM education.
• Published August 24, 2023 by SCHOLÉ, with project number 2022-1-FR01-KA220-SCH-000085611.

Revision History

• The document was drafted on 24.08.2023 by Sofia Sousa, Ana Barroca, and Joana Santos (SCHOLÉ).

Referenced Documents

• The document references the AI4STEM Proposal with ID 1, reference 2022-1-FR01-KA220-SCH-000085611.

Copyright

• The AI4STEM Consortium holds the copyright for this document, 2022-1-FR01-KA220-SCH-000085611.
• It is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Module Introduction

• The module explores AI and its applications in everyday life.
• It aims to help teachers understand AI and its intersection with the Internet of Things (IoT).
• Teachers can find information, examples, activities, and learning objectives for elementary and secondary school classrooms.
• Focus is given to approaching AI and IoT through the lens of the 5 Big Ideas (AI4K12 initiative), with emphasis on Societal Impact.

Glossary - Key Notions and Concepts

• Artificial Intelligence (AI) combines computer science and robust datasets to enable problem-solving.
• AI involves developing computer systems to perform tasks requiring human intelligence i.e. visual perception and speech recognition.
• A robot is an automatically operated machine that replaces human effort.
• Capacity-based AI is categorized into Artificial Narrow Intelligence, Artificial General Intelligence, and Artificial Superintelligence.
• Functionality-Based AI concerns how an AI applies its learning capabilities to process data, respond to stimuli, and interact with its environment; it can be sorted by four functionality types.
• The AI4K12 project focuses on teaching AI in K-12 in the US, aligning its vision for AI education with the CSTA standards for computer science.
• Internet of Things (IoT) uses computer tools to automate real-world processes and is defined as the ability of computing devices to transfer information over a network without direct human interaction.

Introduction to Artificial Intelligence (AI)

• AI is the ability of machines to simulate human intelligence processes using computer programs.
• AI focuses on developing algorithms/systems for tasks needing reasoning, self-learning, problem-solving, pattern recognition, decision-making, and language understanding.
• AI can learn through supervised, semi-supervised, and unsupervised methods.
• Supervised learning involves a person providing information to the AI.
• Semi-supervised learning involves the programmer providing the AI with a series of classified and unclassified images for it to classify.
• Unsupervised learning involves the AI identifying patterns and differentiating images autonomously.

Types of Artificial Intelligence

• Types of AI have applications and challenges; combined features allow for complex problem resolutions.
• The classification depends on the criteria used, capacity-based and intelligence and functionality-based:
• Artificial Narrow Intelligence (ANI) describes AI tools designed to execute very specific actions/commands. The use of image recognition software, self-driving cars, Siri and Alexa etc. use this common type.
• Artificial General Intelligence (AGI) describes AI that can learn, think, and perform tasks almost like a human. Such machines can multitask and act in everyday life as assistants for humans.
• Artificial Superintelligence (ASI) is the kind of machine often depicted in science fiction movies. It will keep learning and become stronger than any human. Movies like Ex Machina or I, Robot use this concept.
• AI functionality types sorted in four different types:
• Reactive Machines is the most fundamental type of AI. AI reacts in the immediate to a request but can't store any memory or learning from past experiences. A famous reactive machine from IBM was Deep Blue, able to beat Gary Kasparov back in 1997.
• Limited Memory refers to the AI that can store past data and make predictions using it. Most of the AI applications used nowadays are under this category, like chatbots or self-driving cars.
• Theory of Mind refers to a model of AI able to pick upon emotions and subtle environment changes and act on them.
• Self-Aware marks a type of AI that reaches its singularity, so the AI machines would be out of control because it'd be able to have a sense of self and think independently.

Task for the classroom - AI in the classroom

• To introduce the theme, the teacher can show students various images of AI use in everyday life.
• Theachers can choose images of robots, virtual assistants (like Alexa or Siri), facial recognition apps, self-driven cars, chatbot website (like chat GPT) and the recommendations part of Netflix.
• Teachers can introduces the different types of AI's and ask questions such as "Do all this AI's work in the same way?".

Application of AI in everyday life

• AI is present in everyday life through facial recognition, Netflix recommendations, and Amazon suggestions.
• AI is also used in healthcare, finance, manufacturing, transportation, entertainment, and education.
• AI algorithms are used in search engines (Google), social networks, virtual assistants (Siri, Alexa), autonomous transportation (TESLA, Efficiency Optimization, Intelligent Navigation), and medical diagnoses.
• Educational AI platforms like Khan Academy, Duolingo, and Squirrel AI adapt the contents to the needs in the progress of the student.
• AI is used in homes and smart homes, to learn user preferences, and better adjust to user expectations.

Task for the classroom – AI in everyday life

• Teachers can discusses with students how much their daily lives rely on AI:
  • "Do you think AI has made our life better or easier? Why?"
  • "Do you have any smart equipment at home? What does it do? How does it help you/your family?".

AI and the 5 Big Ideas

• The "Five Big Ideas" of Artificial Intelligence (AI) were proposed by philosopher and AI researcher, John Haugeland.
• The Five Big Ideas provide a foundation for understanding the principles/concepts of Artificial Intelligence and influence the development of systems that can perform intelligent tasks like humans.
• The Five Big Ideas of AI4K12 that have made the industry soar with AI are:
• Perception: Computers can perceive the world with sensors, enabling computers to 'see' and 'hear' for practical use.
• Representation & Reasoning: In processing data, agents maintain representations of the world and use them for reasoning.
• Learning: Computers can learn from inputs with machine learning, providing statistical interference for finding data patterns.
• Natural Interaction: AI agents require many kinds of knowledge to interact with humans naturally. Such as conversing in human languages, facial expressions, and emotions.
• Societal Impact: AI has both a positive and negative impact on society, thus there must be a balance between using technology and limiting it for enhanced human interactions.

Societal Impact in AI

• AI has a significant impact on society as it is used in various sectors to improve ares of peoples lives:
• Examples are for automating repetitive tasks in industries, Health and Medicine for improving the accuracy of treatments, transport for autonomous and intelligent driving, and education for personalization.
• There is always a chance of having biases, leading to different outcomes of the AI performance.

The Internet of Things and Artificial Intelligence

• IoT helps the AI to process and create an appropriate response to a real-world event signaled by the IoT; AI is essential to IoT.
• The question that IoT users and developers are asking is not if but how far AI can be taken which lies on the complexity of the real-world system the IoT supports.
• If a switch is pressed, and it's dark, AI would turn on light A, recognizing not only the event but also the state (the darkness).
• IoT is aimed at reducing the need for direct human work, but AI can help advance the capabilities of IoT tools.
• New AI applications to IoT will be greatly expanded in the future:
• Robots in Manufacturing - robots are equipped with AI algorithms so they can learn from the new data that sensors i.e. IoT provide, saving time and cost for manufacturing.
• Self-driving Cars - These cares predict the behaviour of pedestrians and other cars in different circumstances.
• Smart Thermostat Solution - Some thermostat solutions have AI-powered IoT, being able to check and manage the temperature from anywhere based on user schedule and temperature preferences.

Task for the classroom – Iot and Al

• After going through the theory about IoT and AI, the teacher can ask the students to think of some examples where they think the two are used and describe what is the role of the IoT and the role of the AI.
• If the subject is fully understood, teachers will understand the concept of IoT and of AI working together and how they can make lives easier.
• Smart examples include; Smart parking systems and healthy diet monotoring systems.

Ethics in Artificial Intelligence

• There are various ethical concerns about AI:
  • Fairness, data responsibility and privacy, climate degradation, threaten human rights, accountability, and technology misuse, among others.
• All AI-powered devices can collect a lot of data about children's behaviour, preferences and learning patterns, raising privacy concerns.
• Relying on AI can prevent students from developing certain social and personal skills like communication, perseverance, or critical thinking.

Task for the classroom – Ethics and AI

• Teachers can start by showing students some school equipment, asking them to describe its AI components, benefits, and drawbacks.
• Divide class into groups to focus on one piece of equipment, defining positive/negative aspects and safe usage rules,
• Teachers can create an “Ethics in the classroom” anchor chart with students suggestions.