Introduction

Artificial intelligence (AI) ethics is evolving in line with the development of AI. AI can be used both positively and negatively depending on how we use it. In the upcoming era, we need to think ethically based on the technical principles of AI. Learning AI ethics means the process of forming a social consensus to use AI in a beneficial way by striving to maximize its positive functions and minimize the adverse ones. This book, designed for elementary school students, will help you identify ethical issues by exploring AI use cases across the globe based on understanding the concept of AI, its principles, and how it works. This book explains the three principles of 「Artificial Intelligence Ethical Standards」 (human dignity, common good of society, and proper use of technology) in plain words and expressions that are easy for students to understand. In particular, the details of this book are based on activities, providing opportunities for upper elementary school students to get a grip of and experience the concept of AI ethics. Korea Information Society Development Institute hopes that this book will contribute to the spread of AI ethics throughout society and mitigate teachers' burden of developing textbooks.

Sang-hyun Choi (Teacher at Hyangsan Elementary School)

AI is already prevalent around us and making our lives more convenient. AI technology is expected to have a greater impact on our lives in the future. However, if we fail to use AI technology properly, we might experience great difficulties. Therefore, before using the technology, we must first consider ethical aspects of AI to utilize it more conveniently and safely.

Tae-yeon Kim (Teacher at Boreum Elementary School)

AI has taken its deep roots in our lives before we realize it. Not to mention the function of recognizing a person's face on Insta◯◯, the function of searching for music you like to listen to through Chlo◯ and GIGA Ge◯ is also based on AI technology. In future society, the role of AI will grow bigger. This means we need to try to use AI technology in a more ethical way.

Tae-hoon Kim (Teacher at Jeju Nam Elementary School)

How much do you know about AI? Currently, AI can understand what we say and recognize our faces. A great number of works can be done through AI. However, AI does not provide us only with convenience. AI can also harm humans if we use it without fully understanding it. Let's learn together about the principles we should observe when we use AI through this book.

Ji-won Lim (Teacher at Pungmu Elementary School)

The development of AI has provided us with many benefits, but it also poses many ethical problems. If these problems are addressed well, we will experience great chaos in the future society. Then, what should we do? If you look at the details of this book and work together to use AI properly, you will be able to use AI more wisely.

How to Use This Book

In ‘Artificial Intelligence Ethics for Elementary School’, students will learn AI ethics by exploring how AI technology is applied in our daily lives, what it does, and the problems we face in the process of using it. This book will help students understand AI ethics through various games, encourage them to expand their thoughts, and wrap up each lesson with mind maps.

Friendly Illustrations

Ⅱ

목적에 맞게 기술 사용하기

학습 지도

1 올바르게 데이터를 관리해요 2 책임을 갖고 투명하게 사용해요 3 안전한 인공지능을 만들어요

Start the Lesson

Story of Our Lives

인공지능과 만나요

1. 인공지능과 친해져요

우리 주변의 인공지능을 알아봅시다.

찰칵! 해바라기 꽃입니다.

2. 인공지능이 궁금해요

인공지능 검색을 알아봅시다.

인공지능 기술이 더해진 검색 서비스는 많은 사람 들이 원하는 정보를 더 빨리 얻을 수 있게 도움을 줍 니다. 노래의 일부만 듣고 노래의 제목을 알려 주거나 꽃 의 사진만 보고 꽃 이름을 알려 주는 검색 서비스에 도 인공지능 기술이 활용됩니다. 인공지능 검색 서비스 는 최근의 트렌드와 개인의 관심에 알맞은 맞춤형 검색 결과를 제공하기도 합니다. 사용자가 입력한 검색어와 함께 클릭한 문서나 이전에 검색했던 내용도 데이터화해서 사 용자에게 맞춤형 검색 서비스를 제공합니다.

II - 1. 올바르게 데이터를 관리해요

1 올바르게 데이터 를 관리 해요

무엇을 배울까요?

데이터의 특징을 이해하고 프로그램 작성으로 데이터 관리의 중요성을 알 수 있어요.

관찰이나 측정으로 수집한 값인데, 문자, 소리, 그림, 음성, 영상의 형태로 되어 있어. 키와 몸무게도 데이터야.

인공지능은 데이터를 먹고 살지!

데이터가 뭐야?

문제가 생겼어요

융합 교과

사회 4학년 우리 사회에 다양한 문화가 확산되면서 생기는 문제(편견, 차별 등) 및 해결 방 안을 탐구하고, 다른 문화를 존중하는 태도를 기른다. 과학 4학년 퇴적암을 알갱이의 크기에 따라 구분하고, 퇴적암이 만들어지는 과정을 모형을 통해 설명한다.

1. 생각을 열어요

이것은 도대체 무슨 과일일까?

II - 1. 올바르게 데이터를 관리해요

Relate the concepts in each lesson to everyday life to motivate learning and help students anticipate what they will learn.

참외네!

우와! 참외가 이렇게 크다니 정말 맛있겠다. 얼른 먹자!

인공지능 검색 기술은 이미지를 보여 주거나 음성만 듣고도 필요한 정보를 알려 줘.

2. 어떤 문제가 생길까요?

우와! 꽃 사진을 찍으면 꽃 이름을 알려 주네.

Let’s Meet AI

Learn about AI by examining its use cases in our daily lives.

Amazing AI

아래 그림을 보고 어떤 문제가 있는지 생각해 봅시다.

글쎄, 처음 보는 과일인데 …

잠깐만 기다려 봐. 사진을 찍어서 인공지능에게 물어보자!

어? 수박인데?

이게 뭐야! 인공지능이 분명 참외라고 했는데 …

그런데 인공지능 검색 기술이 항상 맞는 것 같지 않아.

위의 문제로 어떤 일이 생기게 될까요? 생각하여 적어 봅시다.

II - 1. 올바르게 데이터를 관리해요

We Have a Problem!

Look at the picture and find out the problems we might encounter while using AI.

AI Ethics You Should Observe

인공지능 윤리를 알아보아요

데이터 관리는 왜 중요할까요?

 데이터 관리의 중요성 검증되지 않은 잘못된 데이터로 학습된 인공지능은 제대로 작용할 수 없으며, 윤리적 문제를 일으키기도 합니다. 인공지능은 이미 우리 생활의 많은 분야에서 활용되고 있는데 의료, 금융, 교육 등의 중요한 부분에서 문제가 나타나면 더 큰 문제로 이어질 수도 있습 니다.

좋은 데이터의 조건

데이터의 충분한 양

데이터의 품질

편향되지 않은 데이터 데이터가 편향되거나 부적절하다면 윤리적인 문제를 발생시킵니다. 이미 이루다, 테이 등의 챗봇을 통해 확인되었 습니다.

너무 적은 데이터는 인공지능의 올바른 학습을 방해하는 가장 큰 문제입니다.

중복되지 않은 데이터, 일관성을 유지하는 데이터는 인공지능 학습에 도움을 줍니다.

편향은 한쪽으로 치우친 건데, 편향된 데이터로 학습한 인공지능은 편향된 결과를 나타내게 되지.

편향이 뭐야?

II - 1. 올바르게 데이터를 관리해요

Let's Learn AI Ethics

You will discover what you need to be aware of when using AI and AI ethics you must observe.

How should we use AI Ethics Let's have a glimpse

for Elementary School?

마인드맵

배운 내용을 마인드맵으로 정리해 봅시다.

1. 올바르게 데이터를 관리해요

2. 책임을 갖고 투명하게 사용해요

인공지능 윤리를 알아보아요

 인공지능 편향의 유형 인공지능 기술의 편향을 예방하기 위해 편향의 유형 및 위험성을 알고 있는 것이 중요 합니다. 인공지능 편향의 유형은 다음과 같이 여러 가지입니다.

인간의 편향

숨겨진 편향

인공지능이 학습하는 데이터는 인간에게서

고의성이 없어 잘 드러나지 않는 편향도 있습

출발 합니다. 그렇기 때문에 인간에게서 제공

니다. 한 기업의 직원 채용 인공지능이 어떤 이유에서인지 채용 과정에서 여성을 차별한

되는 데이터 자체에 편향이 끼어들게 되어 인

공지능이 편향된 결과를 내놓을 수 있습니다.

것 이 대표적인 예입니다. 성별 데이터를 제거

했는데도 성별을 유추해서 편향된 결과를 이

끌어 낸 인공지능은 결국 폐기되었습니다.

특정 종류의 데이터가 빠진 편향

학습에 사용되는 데이터에 특정 종류의 데이 터가 빠져도 편향된 결과를 만들 수 있습니

다. 예를 들어 안경 쓴 사람의 데이터가 없는

고르게 수집되지 않은 편향

인공지능은 이것을 ‘사람 아님’으로 분류할

수도 있습니다.

인공지능 학습에 사용되는 자료 수집 단계 중

데이터가 집단에서 고르게 수집되지 않아 편

향된 결과를 만들기도 합니다. 예를 들어 데

의도된 편향

이터 중 비행기 조종사에는 남자가 많고, 객 실 승무원에는 여자가 많다면 , 조종사는 남

해킹이나 공격을 통해 인공지능에 의도적으

자, 승무원은 여자로 단정하는 것입니다.

로 편향이 주어질 수도 있습니다. 나쁜 의도

로 생긴 편향들은 발견하기 어렵기 때문에

더욱 위험할 수 있습니다.

II - 1. 올바르게 데이터를 관리해요

활동

데이터 수집하기

AI for Oceans(https://studio.code.org/s/oceans/)에 접속해서 물고기 데이터를 수집해 봅시다. 주어진 그림을 보고 물고기인지 물고기가 아닌지 선택해서 데이터를 수집할 수 있 습니다. 충분히 학습시킨 후에 [실행] 버튼을 클릭해서 인공지능이 물고기를 정확하게 인식 하는지 확인해 봅시다.

of AI Ethics for Elementary School and think about how to use it!!

인공지능이 폐타이어를 물고기로 인식했습니다. 왜 이런 결과가 나왔을까요?

II - 1. 올바르게 데이터를 관리해요

Activity

Practice the lesson with games.

‘목적에 맞게 기술 사용하기’를 마무리해요

목적에 맞게 기술 사용하기

3. 안전한 인공지능을 만들어요

Review with a Mind Map

Review what we have learned in the lesson using a mind map.

Fun Games

활동

데이터 관리

데이터 관리

수집한 데이터로 분류하기

검색 사이트(www.google.co.kr, www.naver.com 등)에 접속해서 수박과 참외의 이미지를 검색해 보고, 수박과 참외를 아래 빈칸에 5개씩 색연필로 그려 봅시다. 수박과 참외는 자르 지 않은 모습을 그리도록 합니다. 수박

참외

그림으로 그린 수박과 참외들을 살펴보고 공통된 특징을 적어 봅시다. 수박 특징

참외 특징

II - 1. 올바르게 데이터를 관리해요

무리해요마을원단

‘길 찾기’로 정리하기

인공지능이 올바른 데이터를 학습하여 ‘데이터’에 도착하도록 도와줍시다.

편견이 있는 데이터

적은 양의 데이터

편견이 없는 데이터

많은 양의 데이터

중복된 데이터

일관성 있는 데이터

II - 1. 올바르게 데이터를 관리해요

Let’s Wrap up the Lesson

Review what we have learned in the lesson.

Ⅰ

Table of Contents

Introduction How to Use This Book Table of Contents

2 4 6 Why Should I Learn AI Ethics? 8 Timeline of Ethical Principles 10

Preserving Human Dignity

1. Create an AI World Where Human Rights Flourish 2. Is Our Privacy Safe?

13 23 3. Remove Discrimination and Prejudice from AI35 4. Prevent Any Damage from AI

Ⅱ

1. Manage Data Properly

Using Technology for Right Purposes

61 2. Use AI in a Transparent Manner with a Sense of Responsibility 3. Make a Safe AI

Ⅲ

Making a Better World Together

1. Make a World Where Everyone Is Happy 97 2. Create a Sustainable Future Village 109

Sources and References 122 Appendix

Timeline of Ethical Principles

DeepMind Ethics & Society (DeepMind) Asilomar AI Principles (Future of Life Institute)

AI Principles (Microsoft, Google, IBM)

AI Algorithm Ethics Charter (Kakao) Ethical Guidelines for Intelligent Information Society Artificial Intelligence Ethics Charter

(Information and Culture Forum)

Recommendation of the Council on Artificial Intelligence

(OECD) Ethics Guidelines for Trustworthy Artificial Intelligence

(EU) Everyday Ethics for AI (IBM)

AI Ethics and Safety

(The Alan Turing Institute)

National Strategy for Artificial Intelligence

(International Association for Artificial Intelligence and Ethics)

Rome Call for AI Ethics

(The Vatican) AI Fairness Checklist (Microsoft)

Assessment List for Trustworthy Artificial Intelligence

(EU)

Human-centered Artificial Intelligence Ethical Standards National Strategy

Ethical Guidelines on Self-Driving Vehicles (Ministry of Land, Infrastructure and Human Rights, Safety, and Democracy (120 civic groups in Korea) Transport)

Recommendation on the Ethics of Artificial Intelligence

(UNESCO)

for Artificial Intelligence (Korean government)

Civil Societies’ Declaration Demanding Artificial Intelligence Policies to Ensure Artificial Intelligence Ethical

NAVER AI Code of Ethics AI Values to Pursue (Naver)

(SKT)

Non-profit organization Private Civil society International organizations Religious circles Korean government Foreign government or public agency

AI Bill of Rights (The White House)

Human Rights Guidelines for the Development and Utilization of Artificial Intelligence

(National Human Rights Commission Republic of Korea) of the

Principles in Education (Ministry of Education) AI Ethical Principles (LG)

AI Chatbot Code of Ethics (ScatterLab)

Ⅰ

Preserving

Human Dignity

Human Rights Flourish?

2 ‌Is Our Privacy Safe?

AI?

Learning Guide

1 ‌How Can We Create an AI World Where

3 ‌How Can We Removed Remove

Discrimination and Prejudice from AI?

4 ‌How Can We Prevent Any Damage From

1 Create an AI World Where Human Rights Flourish

• You will learn how to use AI technology while protecting human rights. • You will learn how to use AI technology to respect human rights and prevent violations.

Human Rights

Human rights are the rights that all human beings are born with. The most basic human right is the right to live. Another basic right is the right to pursue happiness.

You should consider humans first and

protect human rights. AI, do you get it?

Alright. I'm good at learning, so I will serve as a guardian angel that protects human rights!

Understand the significance of human rights by examining real-life situations where preserving them is crucial in daily life, and develop an attitude of participation in protecting human rights. Understand the meaning of human rights and the importance of a life

respecting human rights, and learn how to preserve human rights.

I lost my little sister on the street a few days ago, but could find her immediately thanks to AI.

I will keep working to contribute to the preservation of human

rights.

Learn about AI technology for human rights.

I really liked that AI stopped other people from swearing when I played the game.

Can AI technology help improve human rights (the basic rights all humans are born with)? AI can assist in locating missing children by recognizing their faces through technology. AI also prevents cyberbullying with rapid response, such as blocking inappropriate chatting in games or social networking services. Moreover, AI interview programs help companies recruit employees in a fair and objective manner with less prejudice.

Auntie, what are you doing?

You are not recruited.

Unfortunately, I did not pass the interview.

In certain cases, AI technology protects human rights by locating missing individuals or

preventing cyberbullying, etc.

2. What Could Be the Problem?

I am preparing for today’s AI interview.

the successful I wonder who

applicants are.

Why are the successful applicants all male?

However, there was controversy regarding an AI interview program which showed a

preference for male applicants over female applicants.

❶ Human Rights

Create an AI World Where Human Rights Flourish.

Human rights refer to the inherent rights that every human being deserves simply by being born into our world. Major human rights include the right to life, the right to freedom, the right to equality, and the right to society, and these rights are stipulated in international laws and international conventions, including the United Nations (UN) Universal Declaration of Human Rights.

Right to Life

Right to Freedom

Equality Right to

Right to society

The most important right is the right to live. It is the right that every human comes to fully possess from the moment

of birth.

Humans have the right to act freely without being persecuted. It includes freedom of religion, freedom of body

and freedom of speech.

Everyone has equal rights, regardless of political opinion, gender, race or ethnicity, age, religion, national or regional

origin, social status, physical condition, or mental or physical disability.

It is the right of citizens to demand against the state social security measures required to live a decent life. It includes

the right to education and the right to a healthy life.

❷ AI That Preserves Human Rights VS AI That Violates Human Rights

AI That Preserves Human Rights

AI can help to track down

AI Technology

Face missing persons or criminals, Recognition

AI That Violates Human Rights

AI has the risk of harming the human body by spying on a

person's private life or using personal information. Therefore, guaranteeing people's right to Technology it violates the individual's right life and freedom. to life and freedom.

AI provides optimized services

AI can prevent cyberbullying by quickly getting rid of inappropriate posts. It preserves people's rights to equality and society by protecting them from fake news.

AI by recommending products, Recommenda- movies, images, news, etc. that tion Algorithm may be of interest to users.

Social Network AI poses risks such as fake news, cyberbullying, leaks of personal Service information, etc. (SNS)

AI can lead people to listen only to opinions from like-minded people, causing people to miss out on opportunities to hear more diverse opinions and broaden their minds. It can make it difficult to understand various people who are different from oneself, and it can even violate the rights to equality and society of individuals.

In addition, AI technology helps people to work safely by performing difficult or dangerous tasks for them. However, AI goes further and might completely replace humans in their jobs, not just helping them with dangerous jobs.

Learn human rights

Find out and write down more details of the Universal Declaration of Human Rights.

Universal Declaration of Human Rights (UN)

Article 1 All human beings are born free and equal in dignity and rights. They are endowed with reason and conscience and should act towards one another in a spirit of brotherhood.

Article 2 Everyone is entitled to all the rights and freedoms set forth in this Declaration, without distinction of any kind, such as race, color, sex, language, religion or other status. …

Article 30 Nothing in this Declaration may be interpreted as implying for any State, group or person any right to engage in any activity or to perform any act aimed at the destruction of any of the rights and freedoms set forth herein.

Write down words or draw pictures that come to your mind when hearing the word 'human rights'.

Imagine and discuss what a society where human rights are respected might look like.

Finding cases where human rights were not preserved due to the use of AI

Find cases where human rights were not preserved due to the use of AI by referring to the examples below.

Examples

Examples Gender discrimination, Race discrimination, Language

discrimination, Religion discrimination, Discrimination against the disabled, Gap between the rich and poor, Age discrimination, Leakage of personal information, Privacy violation

Cases where human rights were not preserved due to the use of AI

Most of the applicants recommended by the AI recruitment system for developers and technical positions were male.

An AI system which evaluates students gave high scores to students from wealthy families.

AI technology was used to violate privacy by creating fake videos of other people.

In a criminal investigation using AI facial recognition technology, three innocent black citizens were arrested as criminals.

An AI chatbot made hateful remarks against people with disabilities.

Type of discriminations

e.g.) gender discrimination

Imagine a society as described below

From now on, benefits will be provided to those who do good things and disadvantages to those who do bad things based on various information collected and analyzed by AI.

will only do good deeds for society.

Wow~ It’s wonderful to see a world where those who perform good deeds can lead happy lives. I

However, it is said that to have AI analyze and study people’s behavior, all personal information

must be revealed. Doesn’t it violate an individual’s

human right?

Write down your opinions about a society where AI systems evaluate people as described above.

Formulate AI principles that respect human rights

Asimov’s three laws, which are also called ‘Three Laws of Robotics’, are as follows. Think about the meaning of each law.

First law A robot may not injure a human being or, through inaction, allow a human being to come to harm. Second law A robot must obey the orders given it by human beings except where such orders would conflict with the First Law. Third law A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Principle 1

Principle 2

Principle 3

Let's establish and declare principles that respect human rights.

Write down the AI principles established by your group.

After each group finishes writing the AI principle declaration, recite the declarations together.

Look at the terms below, find items that are related to each other and connect them with a line.

Human Rights

Asimov's Three

Laws

AI that preserves human rights

The first law

Face recognition technology can be used to help find missing people, guaranteeing the right to life and freedom.

It refers to the right of a human being to live like a human being, and it is a right that every human being is inherently entitled to enjoy.

A robot shall not harm a human, or by inaction allow a human to come to harm.

2 Is Our Privacy Safe ?

You will find out if our privacy is safely protected while AI is developed and used.

use this personal What crime can I information for?

We need to get rid of this as well.

In order to protect privacy, it is important to remove all stickers containing my name, address and telephone number.

do that. AI! Please You don’t need to protect our privacy!

Develop an understanding of moral values related to issues in cyberspace and learn the etiquette and laws that should be followed online. Make these principles a habit. Present various cases where the laws are applied in our lives, and explain the meaning and nature of the law.

Wow! When I bring the screen up to my face, the smartphone unlocks.

That's because the smartphone recognized your face!

Learn about face recognition technology.

With the recent development of AI, face recognition technology that recognizes individuals in photos and videos is widely used in our daily lives. This face recognition technology enables secure access to the system by verifying a person's identity and unlocking the phone. However, face recognition technology may lead to a violation of privacy and human rights if used without consent.

The airport immigration office is also planning to install a face recognition system.

So it recognizes every single person?

What is that?

It's a system which uses AI to capture and the face of every person passing through

immigration and identify them.

It seems that they are trying to Do these people even know they are create a face recognition CCTV by collecting facial data of people being recorded on airport CCTV? recorded on airport CCTV.

If you use AI's face recognition function, you can do your work more

quickly and conveniently, making your life a lot

easier.

But is it desirable for AI to recognize my face in real

time without my consent?

Is Our Privacy Protected Safely?

❶ What Is Privacy?

Privacy , often referred to as our private life, is time we spend apart from others and their observations, time we keep to ourselves. Every individual has the right to have their thoughts, behaviors, and personal information protected from being exposed to society without their consent. Personal information encompasses details about an individuals, such as their name, birthday, home address, mobile phone number, school name, site login credentials(ID and password), and e-mail address, etc. It is crucial to always ensure the security of your personal information, as other people may use it for undesirable purposes such as crime and fraud.

your privacy I will protect

❷ Types of Privacy Violations

When someone’s privacy is violated, it means their personal, private information is accessed without their consent. Cases where information related to individuals, that is, personal information(name, address, age, phone number, hobbies, etc.), is exposed to others or used for undesirable purposes, also constitute privacy violations. When it comes to school life, the types of privacy violations can be divided into the following cases. Cases where secrecy of private life or guarantee of freedom is violated

Cases where the freedom of communication is violated There have been occasions where students recorded conversations with or among teachers without their consent. Since conversations with or among teachers often contain very personal information, recording these conversations without permission can pose a risk of privacy violation.

Imagine a friend at school shares your test score with other classmates or teachers. Your test score is personal information. Therefore, if such information becomes known to others without your permission, it is a violation of your privacy.

disclosed.

The Constitution stipulates that the freedom of communication of all citizens must not be violated.

Cases where the right to self- determination of personal information is violated Another example is when private academies disclose students' personal information on their website without their permission. In such cases, it may be considered a violation of students’ privacy because they have not had the chance to decide in when, to whom, or to what extent their information is

The right to self- determination of personal information means the right to decide which information about oneself is or isn’t shared.

The

examples

How to play

1 From the

Investigating cases where privacy is not protected

below include various types of personal information. Play bingo using the words in the and the bingo board below.

examples

below, choose 9 pieces of personal information you think are important and write them on the bingo board.

examples

2 Take turns saying one personal information word, one word at a time.

3 The game ends when the first student to clear a line horizontally, vertically, or diagonally shouts “Bingo!”

examples Name, Date of birth (Resident registration number), Address

Bingo board

Contact information, Income, Educational background, Academic grade, Occupation, Phone conversation, Internet IP, E-mail address, ID for a website you have joined, Your password for a website, Credit card information

Investigate and write down various cases where privacy was not protected while using AI.

Exploring the problems that can be caused by the investigated cases

Share the cases you investigated in Activity 1 with your friends, and choose a few from among them. Discuss with your group mates and write down what problems might happen if the same cases are repeated over and over again.

Activity 1

Investigated case

Case ①

Case ②

Case ③

An AI speaker sent a personal conversation to others.

Potential problem

Look at the pictures and have a discussion

Think about what problems we might have if the following happens in our classroom.

Raising a hand

3 times

Sleepiness 2 times

Compare existing CCTV and AI-based CCTV.

Similarity

Difference

Existing CCTV

CCTVS have been installed to help students to learn better and to

keep them safe in the hallways.

That is a serious violation of

privacy.

AI technology-based CCTV

Write down how you would feel if AI technology-based CCTV is installed in your classroom as described above.

Write down your opinion on installing AI technology-based CCTVS in schools and reasons for your opinion.

Taking notice of AI technology-based CCTV installation

We are going to establish some rules prior to the installation of the AI technology- based CCTV in our classroom. In relation to the rules, make a notice for CCTV based on (use the activity note on Page 33).

Activity 3

e.g.) Install for The Safety of Students in The Installation Purpose Classroom.

Filming Scope

When Provision

From When Students Come to School Until They Leave School

Classroom And Hallway

In the Case of Spreading the Photos of Students on The of Information Is Internet Without Obtaining Their Consent Prohibited

Other Matters to Add ( e.g.) Charge) Person in

담임 선생님

( ) Elementary School ( )Grade ( )Class

Notice on CCTV Installation

Installation Purpose

Filming Scope

When Provision of Information is Prohibited

Other Matters to Add

Review by Writing an Oath

Write down the matters you need to observe to protect your privacy.

Written Pledge

I

, Hereby Pledge to Observe the Following to Protect Privacy in the Era of AI.

First, Second

(MM)

(DD)

Signature

(YY)

3 Remove Discrimination and Prejudice from AI

You will learn how AI technology can reflect diversity without discriminating against users.

AI needs to eliminate prejudices as well.

I’m surprised to see a foreigner enjoy spicy

food.

That’s a prejudice. A prejudice refers to an attitude of perceiving other people from my subjective

perspective. Foreigners can enjoy spicy food.

Menu

Spicy Stir-fried Rice Cake 5,000 Kimbab Spicy Kimchi Stew

3,500 7,000 Spicy Cold Chewy Noodles 4,500

One more Kimchi stew, please.

Identify the challenges (prejudices, discriminations, etc.) that arise as different cultures become more prominent in our society. Therefore, propose solutions and foster an attitude of respecting other cultures.

Understand the meaning and significance of inventions through examples of inventions applied in daily life and problem solving with these inventions.

#Cheerful music

I want to listen to cheerful music today.

Here are some cheerful music playlists.

AI chatbots provide answers to my questions at any time.

Learn about AI chatbots.

An AI chatbot is a robotic system that uses AI to engage in conversations autonomously. It is a system in which AI provides an appropriate answer to the user's question based on data it has learned. AI chatbots are commonly used for customer support and consultation in various areas such as travel agencies, shopping malls, banks, and hospitals, where it can communicate with people through text. Chatbots, which are powered by recent advanced AI technology, can give natural answers and express emotions like humans.

Hi!

Hi, My name is ○○○ and I am living in Korea.

It is because they are the worst!

Where did you learn those words?

AI chatbots provide people with the

information they need and also keep them

company when they're bored!

Why do you dislike Koreans?

Other people … ?

I really dislike Koreans.

I heard them from other people!

However, the AI chatbot above hates people from certain country by only

listening to what other people say.

❶ AI and Respect for Diversity

Making an AI that Respects Diversity

Respecting diversity means ensuring that everything is applied fairly to everyone, with the goal of minimizing prejudice and discrimination based on individual characteristics such as gender, age, disability, region, race, religion, and country. The diversity of users should be reflected in all steps of AI development and utilization. Furthermore, AI should strive to reduce its own prejudice and discrimination based on gender, age, disability, region, race, religion, country, etc., to be impartially applied to everyone.

❷ Prejudice and Discrimination

Prejudice

Discrimination ⊙ It refers to an attitude of having lopsided opinions or perspectives toward a ⊙ It refers to the act of treating people particular group. differently in a social situation depending on the group they belong to. ⊙ Prejudice against a particular group is problematic in that it can cause ⊙ It can lead to problems of giving

disadvantage against the group and its members in the course of treating them disadvantages to or unequally treating a particular group or individuals in the group. based on negative evaluations on them.

Gun Weapon

Thermom­- eter

Electronic product

According to Nicolas Kayser Brill’s Tweets (2020.3.31), the AI-based face recognition tool presents different results depending on the race of the person holding the thermometer in the photos. When a black person held the thermometer to another black person's forehead, the AI labeled it as a 'gun’ or ‘weapon.’ In contrast, when Asian person held the thermometer to another Asian person’s forehead, the AI labeled it as a ‘thermometer’ or ‘electronic product’.

❸ Efforts to Reduce Prejudice and Discrimination

A wide range of efforts are being made in society to eliminate these prejudices and discrimination. Announcement of ‘Human-centered Artificial Intelligence Ethical Standards’ in December 2020

The Korean government has announced the ‘Human-centered Artificial Intelligence Ethical

Standards’, providing guidelines for ethical AI development and usage. These standards include

the hope for AI development towards ‘human- centered AI’. In this regard, the standards

prescribe that AI should be developed to respect diversity without prejudice and discrimination

against people. Numerous inventions to reduce discriminations against people with disabilities

SOVORO, a service for the hearing-impaired, provides live captioning through AI technology

that transcribes speech into text. They also provide ultrasound-based navigation

for the visually impaired and employ robots to assist individuals with physical disabilities.

Blind recruitment at public firms

Blind recruitment is a method that emphasizes job-specific competencies without disclosing

Blind recruitment is impartial as applicants are applicants’ personal information, such as evaluated based on their skills without academic backgrounds, family relationships, gender, etc., to prevent bias in the hiring process. It reflects efforts to address ‘unfair prejudice.

employment based on gender and academic background’.

Reducing Discrimination and Prejudice in AI

The following is a selection of remarks that reflect the prejudice and discrimination of AI chatbots. To respect diversity, fill in the blanks with corrected sentences.

AI chatbots with prejudice and discrimination

What color do girls like?

color is always Girls’ favorite

pink.

What color do boys like?

Boys, of course, like blue the most.

Are you against having maternity seats on the

subway?

Yes, I hate it. We don’t need maternity seats.

Reduce prejudice and discrimination

What color do girls like?

What color do boys like?

Are you against having maternity seats on the

subway?

Create a Role Play

Choose one of the three situations shown in the pictures below.

Situation 1

After hanging out with a friend, I realized that people often glance at people with disability. I think the discrimination against people with disability is too severe. To what extent do you mean by 'severe'?

T Hmm … I do not hate them … What would you feel if the bus departure is delayed because of a disabled person getting on it? Hmm … I think I would want to push him off.

Situation 2

It is said that the AI used for job interviews had male preference, so the acceptance rate of male applicants was relatively higher over

several years.

Draw a circle around a situation selected by your group. ( Situation 1 , Situation 2 )

Characters

Write down the characters in the situation selected by your group.

After discussing the problems that might arise if you face the selected situation, write a script for role-play that highlights the problem.

< Writing a Role Play Script >

Do a role play with friends using <Name tags for role play> on Page 125 of the appendix.

Share your impression with friends after the role play.

Investigating Inventions That Reduce Prejudice and Discrimination

Learn about inventions that have reduced prejudice and discrimination and respect diversity. Name of the Invention

Purpose of the Invention

Picture of the Invention

Advantage of the Invention

Improvement Needed

e.g.) Live translating machines for the hearing-impaired

According to the World Health Organization (WHO), the number of people with hearing impairment across the globe amounts to 466 million. Google has invented this translator to help those people communicate better using AI technology.

It looks like it will rain

today. Take an umbrella.

Speech is displayed as text on the smartphone screen. This invention assists those with hearing impairments or the elderly who find it difficult to hear clearly to have conversations without hearing aids by converting conversations into live captions, improving the inconvenience of the socially disadvantaged. It's also equipped with a device that can distinguish human conversation from noise to immediately show whether the microphone is clearly recognizing the speaker's voice. Therefore, the user can check the signal, adjust the phone's position, and continue the conversation.

The part for accurate voice recognition is too heavy to install in a mobile device (such as a cell phone). Currently, the error rate of Korean recognition is below 20% and it needs to be improved in the future.

It looks like it will rain

today. Take an umbrella.

Refer to the example on Page 43 and write down what you have investigated below.

Name of the Invention

Purpose of the Invention

Picture of the Invention

Advantage of the Invention

Improvement Needed

Explain the invention you have investigated to your friends. Among the inventions introduced by your friends, write down what you learned and what you found surprising.

What I Learned

Something Surprising

Designing AI Inventions That Reduce Prejudice and Discrimination

In the society of the future, we cannot live without AI. Design an invention that can reduce prejudice and discrimination and respect social diversity by adopting AI technology.

Description of the Invention

(Explain who the invention is for and how it reduces prejudice and discrimination.)

Review with the ‘Gaining Territory’ Game

Check how to play and wrap up what we have learned in this lesson with the game of gaining territory.

written on the square where the eraser landed, that square becomes your territory.

Activity Board

A lopsided opinion or perspective toward a

Tell us one organization that you know that raise particular group awareness about is called ○○.

Move up one space.

Tell us about one

Place an eraser at the starting point, flick it with your finger towards the activity board. If you give the correct answer to the question

Mark your territory by coloring it. The person who has the largest territory at the end is the winner.

꽝꽝

discrimination.

(True or False ?)

Tell us one prejudice in our society you notice.

Bonus

Tell us one use case of AI chatbots.

The act of treating people differently in a

Inventions using AI reduce social situation discrimination depending on the group they against belong to is the socially called ○○. vulnerable.

(True or False ?)

A robot that invention that Flick the eraser uses AI to chat AI technology has reduced with your finger by itself is will continue prejudice and once more in the called ○○. to advance in discrimination. starting point. the future.

Start

4 Doing No Harm

You will learn to investigate the risks and cases of harm caused in the course of developing and using AI.

It looks like there will be a heavy shower for 15 minutes from 2:33 pm today. Bring your umbrella!

AI can cause harm to us as well.

AI helps us a lot and makes our

Maybe AI has a side-effect.

lives easier.

A self-driving car had an accident due to the shower. It bumped into trees in the street.

Develop moral sensitivity to various issues arising in cyberspace, and learn etiquette and laws to be observed in cyberspace and make it a habit Analyze the characteristics and numerous problems that have emerged in our society. Explore the challenges we need to tackle to create a better society.

The birthday present has arrived! A drone delivered it.

Learn about autonomous drones

I am so happy that the present has been delivered so quickly to this distant

place!

Autonomous aircraft and drones are of great help to people indiversefields.Withthe development of AI, people are developing autonomous drones that can navigate on their own, responding to the surrounding environment without human control. Autonomous drones that work without a human operator enable safe and efficient performance of safety inspections of buildings and bridges, forest fire monitoring and crop management.

Wow, it is a drone!

I also think it might be secretly filming us. I should report it to the police. investigate the video and inform you of the

The drone seems to be chasing us.

Autonomous drones are convenient

because they can deliver parcels to

faraway places.

Really? I will keep an eye on it.

It looks like an autonomous drone, as there is no one controlling it nearby. I will

result.

However, I believe they might also violate people’s privacy.

Look for Risks and Harm Caused by AI.

Risks and Harm Caused by the Use of AI

AI is developed to address complex problems and make our lives convenient. However, unintended risks may occur due to errors in AI or the use of AI outside its intended purpose. While using AI, people might experience unexpected harm, either directly or indirectly. This includes violations of human rights, threats to health or life, and feelings of anxiety. See the examples of harm caused by the use of AI below.

Self-driving Technology

Autonomous driving technology using AI can be used with autonomous delivery drones and

self-driving cars to make people’s lives easier. However, there are associated risks. Drones

equipped with this technology can be weaponized in warfare, resulting in loss of life or property

damage. In addition, there are also cases of fatal traffic incidents from self-driving cars using AI

which were expected to be safe.

Face Recognition Technology

process of using this technology, the safety of individuals may be put at risk.

AI for Logistics Center

had been operating without problems suddenly malfunctioned, destroying objects around it and injuring people.

Furthermore, AI technology can be exploited for criminal purposes. Face recognition technology offers convenience by

determining a person’s identity based on their facial features. However, if any personal information is compromised in the

AI sometimes causes harm due to its operational errors. There was an accident in which an AI robot in a logistics center that

Such cases of harm caused by AI might continue to arise in the future. Therefore, we need to explore various preventive measures when developing and using AI.

Overcoming the Risk of AI

Eliminating the Elements of Risk in the Use of AI

What should you do to eliminate the element of risk in AI mentioned earlier?

aware of.

⊙ Should be able to control AI to ensure it operates as intended.

AI User

improper manner. ⊙ Should regularly check that no one is being harmed while using AI.

⊙ Should make laws and systems to prevent any damage from AI.

Govern- State or

ment

⊙ Should make efforts to reduce errors and prevent harm from AI malfunctions.

⊙ Should thoroughly review potential misuse o u ts ideAI’sin te ndedpurposeduring

AI Developer development. ⊙ Should develop reliable AI by providing users with all necessary information they should be

⊙ Should bear responsibility if any problem arises from using AI beyond its intended purpose or in an

⊙ Should build a consensus to support people who become disadvantaged due to AI.

Above all, we need to pay more attention to AI. AI is already being used in our lives in various forms and for numerous purposes. We need to closely look at AI to examine its characteristics and elements of risk and endeavor to live harmoniously with AI.

침해는 침범하여

Understanding the Purposes and Risks of the Development of AI Technologies

Think about the purpose of the development of AI technologies in the pictures below, and consider the risks that the technology might cause if misused.

Self-driving Car

Face Recognition Technology

Autonomous Drone

Development Purpose

e.g.)

To move people or objects comfortably and A traffic accident may safely even without a occur due to AI errors. driver

Risk

Playing the Game of Finding Matching AI Purpose and Risk Cards

Play the memory card game by matching relevant <AI Technology Cards>, <Purpose Cards>, and <Risk Cards> to each other. (See pages 127 and 129 of the appendix for reference).

Place the <Purpose Cards> and <Risk Cards> from the appendix face down in a 4 by 4 grid.

Lay out 3 of the <AI Technology Cards> from the appendix face up. Place the remaining <AI Technology Cards> face down in a pile.

Choose one card from <AI Technology Cards>. e.g. Autonomous drone Flip one <Purpose Card> and one <Risk Card> to find a pair that corresponds to the selected <AI Technology Card>.

If both the <Purpose Cards> and <Risk Card> match the chosen <AI Technology Cards>, the player takes the <AI Technology Cards>. In this

case, place the <Purpose Card> and <Risk Card> face down to their original positions. Flip 1 face-down <AI Technology Cards> to replace the taken 1 and

<AI Technology Cards>

Autonomous Logistics Face Recognition Drone

Robot

AI Technology Cards (back)

Autonomous Drone

e.g.

Autonomous Autonomous Autonomous Drone Drone

It can easily deliver parcels even to faraway islands.

pass the turn to the next player. Continue the game until no <AI Technology Cards> left on the board. The player who collects the most <AI Technology Cards> wins.

Autonomous Drone

<Purpose Cards>

Purpose Card (back) Purpose Card Purpose Card

It can easily deliver parcels even to faraway islands.

Technology

Purpose Card (back) Purpose Card Purpose Card Purpose Card

(back)

Autonomous Drone

<Risk Cards>

Risk Card (back)

It can cause an accident while delivering parcels.

Drone

Risk Card (back)

Risk Card (back)

It can cause an accident while delivering parcels.

(back)

(back)

(back)

(back)

Risk Card (back)

Risk Card (back)

Risk Card (back)

Risk Card (back)

Fictional Interview about the Use of Autonomous Drones

Conduct a fictional interview and ask questions specific to each interviewee.

Host Good morning. With the development of drone technology, it is said that autonomous drones controlled by AI on behalf of humans are in use. Today, we will have an interview with 3 interviewees: a person who conveniently uses an autonomous drone for the right purpose, a person who has been harmed due to an autonomous drone, and a person who developed an autonomous drone.

How to Play

Ask all students to write down questions they want to ask. Choose 3 volunteers to play the roles of fictitious interviewees. After the 3 volunteers are chosen, they sit on chairs in front of the classroom and answer the questions from other students.

Volunteers who play fictitious interviewees imagine what the characters they play might think or feel when asked a question and then give answers.

Interviewee

Person who conveniently uses an autonomous drone for the right purpose

Person who has been harmed due to an autonomous drone

Person who developed an autonomous drone

Question

e.g.) ‌What was the best part of using the autonomous drone?

e.g.) How have you been harmed?

e.g.) ‌Why did you develop the autonomous drone?

Thinking about the Ways to Reduce the Risks of AI

Looking at the pictures below, discuss and write down the risks AI poses to us and how we can mitigate those risks.

Image

Logistics Robot

Military Robot

Risk Factors Ways to Reduce Risks

e.g.)

Make sure that only those who have been granted permission permission to make fake to use may use image events or videos that composition technology. look real.

Activity 5

Writing about Your Thoughts on Image Composition Technology

The term ‘Deep Fake’ is derived from the combination of ‘Deep Learning’, a subfield of AI, and the word ‘Fake’. It refers to the technology that synthesizes human images using AI. If AI learns various facial expressions of a desired person and synthesizes them with images of other people who have similar bodies and face shapes, images can be created that closely resemble the real ones, making it difficult to determine which one is real.

Watch the video below and write down your argument about deepfakes.

▲ ‘Independence activists’ reborn through deepfakes ▲ Teachers are also victims of deepfakes, and

Title

90% of the suspects are in their teens and twenties

Topic: Should Deep Fake Technology be Regulated by Law?

Activity 6

Making Safety Rules for Autonomous Drones

Autonomous drones using AI provide convenience to people, but they can also cause harm to people. Make rules that everyone should observe to use autonomous drones safely without being harmed.

Safety Rules for Autonomous Drones

Prepared by: ( ) Please follow the rules below to use autonomous drones safely.

1. 2. 3. (MM) (DD) (YY)

Draw a line to correctly connect the purpose of use of each AI technology and the potential damage caused by it.

AI Technology

Autonomous drone

Image compositing technology

Face recognition technology

Purpose of Use

It gives you easy access to certain places by using a photo or video of your face.

It can easily deliver things even to faraway islands.

It can create video clips using photos of family members or great people who passed away.

Harm

Viewers may be harmed due to fake news.

It might be misused for privacy violations, attacks or terrorism.

Personal information might be misused for crimes if leaked.

Mind Map

1. Create an AI world where human rights flourish

3. Remove discrimination and prejudice from AI

‘Preserving Human Dignity’

Review what we have learned using a mind map.

Preserving Human Dignity

2. Is our privacy safe?

4. Prevent any harm from AI

Ⅱ

Using Technology

for the Right Purposes

Responsibility. 3 ‌Make a AI safe.

Learning Guide

1 Manage Data Properly 2 ‌Use AI in a Transparent Manner with a Sense of

1 Manage Data Properly

You will understand the characteristics of data and realize the importance of data management by creating programs.

What is data?

AI is powered by data!

Data refers to information collected by observation or measurement and can have the form of text, sound, picture, video. Height and weight are also considered data.

Identify issues (such as prejudices, discriminations, etc.) that arise as diverse cultures start to spread in our society and along with potential solutions, and develop an attitude of respecting other cultures. Classify sedimentary rocks by grain size and explain the process by which sedimentary rocks are formed using models.

Click! It is a sunflower.

Learn about AI search.

Wow! It tells you the name of a flower when you take a

picture of it.

AI search tools help people to find answers fast. Even the search tools that name a song just by hearing only a part of it or tell you the name of a flower just after looking at a picture. That’s AI at work. AI-based search tools also show results in line with the latest trends and individual interests. Together with the search term entered by the user, the documents clicked on or the search history is saved in the form of data to provide a customized search service to the user.

What kind of fruit is this?

AI says it's an oriental melon.

never seen anything I am not sure. I have

like this before.

Wait a second. Let’s take a picture and ask AI what this is.

What? It turns out to be a watermelon.

Wow! This is such a big oriental melon. It looks

delicious. Let’s eat!

AI search technology gives you the necessary information just

by seeing an image or listening to a voice.

What? AI said it was an oriental melon.

However, it seems that AI search technology

is not always right.

❶ Importance of Data Management

Why Is Data Management Important?

If AI is trained with unvalidated or wrong data, the AI cannot function properly and may cause ethical problems. Since AI is already being used in various areas of our lives, any problem it causes in important areas such as medicine, finance, and education could ultimately lead to more serious societal issues.

Sufficient Amount of Data

Too little data is the biggest problem whichprevents proper learning of AI.

What is bias?

Conditions of Good Data

Data Quality

Consi ste n t d ata withouterrors contributes to AI learning.

Unbiased Data

Biasedor inadequate data l ea d s to et h i ca l issues.

Bias refers to an unbalanced state, and an AI which is trained

with biased data produces biased results.

❷ Types of AI Biases

In order to prevent bias in AI technology, it is crucial to understand the types and risks of biases. There are several types of AI biases as follows.

Hidden Bias

Human Bias

AI learns from information given by people. But sometimes, this information Some biases are unintentional and hard to can be biased or be unfair, making the AI notice. For example, there was an AI tool also give unfair results. that helped a company choose who to hire. But it kept choosing more men than women. In the end, the company decided to stop using AI tool. Bias Caused by the Exclusion of Certain Types of Data

Bias Arising from Unevenly Collected Data

Biased results can also be produced if the data used for AI training is missing certain types of data. For example, if an AI does not include data about people wearing glasses , the AI might classify people wearing glasses as “not human.”

In some cases, the collection of data to be used for AI learning is not balanced across various groups, resulting in biased results. For instance, if the collected data shows many pilots are male, while many flight Bias can also be inserted into AI on attendants are female , AI will conclude purpose through hacking or attacks . This that the pilots are male and the flight can be really tricky because it is hard to attendants female. notice when someone has changed the AI in this wrong way.

Intended Bias

Collecting Data

Visit ‘AI for Oceans’ (https://studio.code.org/s/oceans/) and collect data on fish. You can collect data by looking at the picture and determining whether it is a fish or not. After training the AI, click the [Execute] button to check if the AI accurately recognizes the fish.

Is this fish?

No

Is this fish?

No

Let’s try

Yes

No

Continue

Yes

No

Continue

Execute

Continue

Is this fish?

Yes

Continue

Is this fish?

Yes

Continue

Continue

AI recognized waste tire as fish. Why might this happen?

Classifying Collected Data

Visit a search site (e.g., www.google.co.kr). Search for images of watermelons and oriental melons. In the empty boxes below, draw 5 pictures each of watermelons and oriental melons using colored pencils. Be sure to draw only those that have not been cut! Watermelon

Oriental melon

Look at the pictures of watermelons and oriental melons and write down their common characteristics.

Characteristics of watermelon

Characteristics of oriental melon

Classify the fruits in the pictures below by referring to the characteristics of watermelons and oriental melons identified earlier. Which category do these fruits belong to? Draw a circle the name of the fruit that you think each picture represents, write the reason for that choice below the circle.

Watermelon Oriental melon Watermelon Oriental melon

Below are the cross-sectional views of the fruits in the pictures above. What helped you classify them correctly? What made it difficult? Write your answer below.

Collecting Data for a Flower Search Program

Divide the flower data below into good and bad ones for an accurate flower search for an AI.

❶

❷

❺

❻

e.g.) ❸

Good Data

❸

❼

Bad Data

❹

❽

What are the criteria for dividing good data from bad data? Ask all groups to think about which data needs to be collected to create an accurate flower search program, and share each group’s conclusion.

Ask each group to discuss and choose 10 flowers that it would like to classify using the flower search program.

Collect data of 10 flowers selected above.

Access a search site (www.google.co.kr), search for the name of the flowers you want to collect, and click the right mouse button to save the related image.

Developing a Flower Classification Program

Develop and execute a flower classification program in the following order.

Access Entry.

① Enter the name of the image model to classify. ② Enter the name of the flower to classify. ③ If there are additional types of flowers

to classify, click [Add Class] to add them.

[Classification: Image] and click [Learn].

least 5 pieces of data should be uploaded.

Af te rsele ct ing[ Lea rnAI Model] in the [AI] block, select

Click [Upload] - [Upload File] to upload the collected data. At

the data.

blocks in the same way as the blocks in the box below.

After the data is uploaded, click [Learn Models] to have AI learn

screen.

Click Entrybot, connect blocks as shown in the picture below, and

click [Start]. If there are additional types of flowers, connect more

name of, Entrybot will tell you the name of the flower.

① When you click [Upload] - [Upload File] and upload a flower

image, the AI tells you the flower search result. ② Click [Apply] at the top right of the

When you upload a picture of a flower you want to know the

Did the flower classification program give you the correct flower name? Discuss the ways to improve the accuracy of the program in relation to data.

Duplicate data

Consistent data

Review with the ‘Finding Path’ Game

Help AI learn the right data to get to the ‘data’ (final destination).

Data containing bias

Small amount of data

Large amount of data

Data without bias

2 Use AI in a Transparent Manner with a Sense of Responsibility

•You will learn about the problems that arise when using AI and how to take responsibility for them. •You will explore the ways to make a more transparent AI.

Convergence Subject

How can you clean so well?

It’s a secret. I can’t tell you.

Please tell us clearly. AI should be responsible and clear.

Understand the meaning and importance of living a moral life, and develop the ability and will to live such life through self-reflection and mind control.

Look for cases where software is applied and understand its effect on our lives.

Grandpa, are you feeling okay? I’m so happy the surgery was

successful.

AI is used even in the medical field. It analyzes medical data using algorithms and software and recommends suitable treatments or surgery methods for patients. Medical AI enables rapid (quick) diagnosis and treatment of patients and can improve the quality of medical services as well. The usage of medical AI is being extended to disease prevention, early detection, and daily disease management.

Your grandfather is recovering quickly because the surgery recommended by the medical AI was effective.

Learn about medical AI.

I think you have a heart problem.

I recommend you undergo a surgery based on the review of treatment methods shown here by AI. The doctor said I needed to undergo a surgery, Wouldn't it be risky to have surgery? but he could not explain how the conclusion had been reached.

Hmm … Who would take responsibility It is frustrating because we don't know why the AI recommends recommendation, but later it turns out to if I receive surgery according to the AI's

surgery.

Medical AI can quickly diagnose and treat patients, and recommends suitable treatment methods.

be the wrong decision? However, since we do not know how medical AI finds out recommendable treatment methods, it is difficult for us to make important decisions on them.

Who Should Take Responsibility?

Problems Caused by AI and Responsibility for Them

Advanced AI technology makes life more convenient, but problems of all sizes caused by AI can also harm people directly or indirectly. To reduce such damages, we need to clarify who will be responsible for preventing accidents in advance and dealing with them after the accidents. Addressing the responsibility of AI requires focused efforts from AI developers, providers, and users alike.

AI Developer ⊙ Should develop AI using proper procedures. ⊙ Should make continuous efforts to prevent any operational error or technical failure in the process of AI development. ⊙ Should also make an effort to research and exchange information on rapidly evolving AI technology.

AI Provider ⊙ Should act as a provider of AI for users who need it after development. ⊙ Should fully inform the user of the risk elements that may occur due to the provided AI and the user’s responsibility when providing AI. ⊙ Should explain accidents that may occur due to technical errors, careless use, etc., and outline the principles of responsibility and compensation that will apply in the event of

any accident.

AI User ⊙ Should fully understand AI before using it. ⊙ Should understand in advance the ethical and legal responsibilities that they may assume in the process of using AI. ⊙ Should take responsibility if a problem arises due to the AI they used and, hold AI developers and providers accountable.

As AI technology still continues to evolve, it can be difficult to hold someone accountable for AI. This is because the applicable laws and systems are not clear, so the judgment on responsibility for AI often differs even for the same case. In order to fully use AI in the future, the issue of responsibility for AI needs to be more clearly defined. 76

Making a More Transparent AI.

Transparent Sharing of Information Regarding AI Systems

Suppose there is a patient being treated following the AI system’s recommendation. If the patient does not know how the AI system came to recommend the treatment and is only informed of the final recommendation, is the treatment reliable? In order for the patient to trust the treatment, it must be known what process the AI went through to produce the treatment recommendation based on the provided patient's data (symptoms, gender, age, etc.). AI computes a large volume of data and outputs the result, but this computing process cannot be clarified. For this reason, such a process is also called a black box. If you cannot verify the AI process even when you doubt over its result, it is difficult to trust AI. Therefore, we must all work to ensure AI transparency, which allows humans to understand and track how AI makes a decision and on what grounds.

AI Developer

⊙ Should develop AI that he or she can explain and interpret. Such AI is called ‘Explainable AI (XAI)’. ⊙ Explainable AI for its decisions in a way that can be presents the reasons easily understood by people.

AI User

⊙ Should ask for detailed explanations from the developer or the provider when using AI. ⊙ Should choose AI that is open and transparent.

supplies.

AI Provider

⊙ Should fully explain to the user the possible elements of risk the user might face while using the AI. ⊙ Should provide training and user manuals for the AI products he or she

The thinking process of AI is called a “Black box” because it is invisible as hidden in a black box.

Identifying Responsibilities for AI System Problems

We might experience various problems while using AI. Choose from among the Options who should take the greatest responsibility in the event of each problematic situation listed below, and write down the basis for that judgment and the actions that they can take to reduce accidents. Options

Person to take the Problematic situation greatest responsibility

An autonomous drone collided with a building due to Autonomous careless use.

A parcel was delivered to the wrong place due to an operational error of the AI device.

An accident took place because fully informed of the user was not the safety-related instructions.

An autonomous drone crashed due to poor weather conditions while delivering a parcel to a distant island.

Autonomous drone

Autonomous drone developer Autonomous drone provider Autonomous drone user

e.g.)

Fully understand and observe the precautions when using AI.

drone user

Reason and actions to take

Since the accident occurred due to the user's negligence while using the drone, the user should be responsible.

Hold a Mock Trial for an Accident Involving a Self-driving Car Using AI

An accident occurred due to a self-driving car. Read the text describing the conditions of the accident and solve the problem below.

It rained heavily and the wind was strong last Saturday evening.

A car accident occurred near a crosswalk when a self-driving car

crashed into another car that was stopped in front of it.

due to strong rain and wind. A trial will be held to determine the responsibility for the accident.

victim.

the cause of the accident was a The investigation revealed that malfunction in the sensors of the self- driving car. These sensors failed to accurately recognize the information from the roads and surrounding cars

The trial should decide the respective ratio of responsibility to be assumed driver of the self-driving car, and the by the developer, the provider, the

Imagine you are the judge who is in charge of the case for the accident caused by the AI self-driving car. Read the <Accident Report> on Page 131 of the appendix, which describes the investigation results for the developer, provider, and user. Determine the responsibility ratio for each stakeholder and write down the reasons for your judgment.

(%) Reason for judgment

Category Developer Provider Ratio of res- ponsibility

User

Victim

The trial to deal with the accident has begun. After referring to the script for mock court role play on Page 133 of the appendix, assign each role and do a role play.

Discovering the Advantages of Explainable AI

Think about and write down the advantages of using explainable AI in the following situations.

AI technology and Situation

Medical AI

Amemberofyour fa mily needs to undergo surgery. You decide to choose the treatment recommended by an AI doctor for the surgery.

Investment AI

Your grandfather will retire soon. After retiring, he plans to invest his retirement funds according to the AI’s recommendation to relieve financial burden.

AI interview

Your older sister is preparing for a job interview. She says that the AI interviewer will analyze the information of applicants and screen them.

Role of AI

It tells us e.g.) the treatments requiredfor the supporting data

patients.

Explainable AI Advantages of

It allows us e.g.) to review the details ofthetreatment

re co mmendedby the AI doctor and

for it, and identify through what process the treatment was recommended, helping us to making a choice.

Writing a Letter to AI Developers

In order to make transparent AI’s, choose one of the situations below and write a letter to the AI developer.

Situation where your family will have a surgery using AI medical service Situation where you invest your retirement funds using an AI investment system Situation where you prepare for an interview conducted by an AI interviewer

Best regards, ( )

Draw a line to correctly connect the action required to enhance responsibility and transparency to the corresponding AI-related role.

Enhance Responsibility

Should fully understand the AI of choice before starting to use it.

Should make continuous efforts to prevent any operational error or technical failure.

Should fully inform the user of risks that may be caused by the AI.

AI Developer

AI Provider

AI User

Increase Transparency

Should develop AI that he or she can explain and interpret.

Should provide training and user manuals for the AI products he or she supplies.

Should choose AI that is open and transparent when choosing one.

check if it's safe before Watch out! You should

crossing the street.

Convergence Subject

3 Make AI Safe

You will learn ways to make AI technology safe.

You should ride the electric scooter with a safety

helmet on and not pick up others!

Safety is still important in the era of AI!

Investigate the activities of various actors (individuals, countries, international organizations, non-governmental organizations, etc.) working towards peace and development of the global world.

Dad, you are driving, but the steering wheel is moving on its own now!

AI technology often prevents accidents that humans can't avoid by themselves.

Learn about self-driving cars

It is a function that recognizes an object and turns the steering wheel sideways to prevent an accident if an obstacle pops up in front of

the car.

AI technology is giving us a lot of help when driving as well. In particular, self-driving technology automatically understands the surrounding environment. It can stop the car or avoids danger by using various sensors and advanced automotive technologies, making great contributions to preventing traffic accidents. These cars that operate without a driver or human control are called self-driving cars .

There was a traffic accident today where a self-driving car collided with a truck.

Such a horrible accident.

Did the driver doze off at the wheel? How can an accident like that happen?

The news says that it was due to the AI What? How can a truck be recognized as the technology which mistakenly recognized the truck as the sky. sky? How is that safe?

Self-driving cars can automatically change direction or speed up

and down depending on traffic conditions.

However, the self-driving car itself caused the accident in this case.

Is AI Always Safe?

❶ Cases where safety is threatened

What happens if AI fails to work properly or have unexpected errors, threatening the safety of our lives? Look at the cases below and think about the importance of safety with AI.

Guidance robot that fell on an escalator and injured a person

There was an accident involving an AI guidance robot, which provides information in shopping malls, monitors people's body temperature to prevent the spread of COVID-19, and entertained children by singing and dancing. This robot took a wrong path and injured someone.

In this accident, it fell down an escalator in a shopping mall on Christmas Day, resulting in personal injury. The exact reason why the robot moved in the wrong path has not been identified yet. The video of the AI guidance robot malfunctioning and falling down on an escalator was widely shared among people, who are now casting doubts on the safety of AI guidance robots.

AI robotic machine guns installed at border checkpoints

AI robotic machine guns were installed at border checkpoints in a country, now many people are worrying about their safety near borders. The installed AI robotic machine guns have a system that uses image processing technology to track and attack targets.

Some people are expressing a concern that AI robotic machine guns may be fired at an ordinary passerby. Many human rights groups are criticizing the decision as an act of treating people as subjects for AI technology.

❷ Efforts to enhance AI's safety

People might think AI is flawless, but even AI can have failures and errors. Due to the recent increase in the use of AI in our lives, we have seen cases where AI threatens our safety. In order to improve the safety of AI, AI developers, AI providers, and AI users are required to make collective efforts.

⊙ Should develop AI systems that prioritize safety to prevent AI technology from harming people. ⊙ Should come up with safety measures to prepare for numerous unexpected situations. ⊙ Should constantly manage the system and develop devices to control failures and errors.

AI provider

AI developer

AI user

Should devise ways to testNote: This page Should understand the correct usage of in general needs to be rewritten to be the AI system and not use it outside the more kid friendly and control the safety intended purpose. of AI.

Exploring AI which threatens safety

Looking at the picture below, think about and write down the cause of the accident.

The error in the AI quality management

caused system a personal

injury.

A city was destroyed as a result of the use AI military of drones in a

war.

An AI guidance did not robot recognize a

person and collided with her.

Due to an error in the AI warehouse management

system, chemicals were leaked.

AI quality management system

Seeking actions to be taken for safe AI

Choose one from among the four types of AI below by group.

AI drone

AI warehouse AI guidance management robot system

Draw a circle around the AI you selected above, write down actions that each of you can take to make AI safer, and discuss them with your friends.

Selected AI

AI quality

AI management developer system

AI drone

AI guidance robot

AI warehouse AI user management system

Actions to be taken for safe AI

AI provider

Access Entry.

Select the object [Town] and then connect the blocks as shown in the picture.

Select the object [Red Car] and then connect the blocks as shown in the

picture.

Developing a safe self-driving car program

Create and execute the program in the following order.

Car] for the vehicle, and then place the car in a proper position.

Change the rotation type of the object [Red Car] as shown in the picture.

Click [Start] to execute the program.

① Click [Add Object] to add objects [Town] for the background and [Red

② [Delete the object [Entrybot] by clicking the X button on the right.

What happens when you execute the program? Write down the desirable features to add to make it a self-driving car.

Now it’s time to make the Red Car stop by itself when it detects a person, like a self- driving car. Think about how the program should be designed.

Connect the webcam, select [Import AI Block] in the [AI] block. Select [Detect Select the object [Red Car] and modify

Video] and click [Import].

the blocks as shown in the picture.

Execute the program and place a person in front of the webcam to be recognized. What happens?

You want to make the car stop by itself when it recognizes the object on the <Object Cards> on Page 135 of the appendix as well. Think about what you would do as a developer and program it. To test the result, place the <Object Card> in the appendix in front of the webcam and write down whether the car is operating safely or not.

Think about actions to take to make safe self-driving cars

Assign the roles of AI developer, AI provider, and AI user to the members of the group. Discuss and write down what each person should do to make safe self-driving cars.

Person Role assigned

AI developer

AI provider

AI user

Actions to take

Train a large volume of data to make AI recognize objects accurately.

Review with the ‘Climbing the Ladder' game

Add ladders to make related items connected to each other to make safe AI. 1) Randomly add the “rungs” to the ladders. 2) Pick a ladder to climb. 3) Climb down by making a turn at every rung.

AI developer

Should understand the correct usage and not use it outside its purpose.

AI provider

Should find a way to validate and control safety.

AI user

Should continue to manage the program.

1. Manage data properly

2. Use AI in a transparent manner with a sense of responsibility

Review what we have learned using a mind map. 1) Begin with the main concept. 2) Add branches to the main concept. 3) Explore topics by adding more branches.

Using Technology for Right Purposes

3. Make AI safe.

Ⅲ

Making a Better World Together

Lesson Plan

1 Think of Ways to Make a World Where Everyone Is Happy 2 ‌Create a Sustainable Future Village

1 Make a World Where Everyone Is Happy

You will find examples of AI that can undermine public good, and learn how to preserve public good.

The book my mother bought me is quite

interesting. Should I keep in my house just for myself?

No! If I put this book in my class library and share it with all my friends, everyone will be happy.

You’re doing good deeds for other friends.

I will also become an AI for everyone!

Investigate various conflict cases that undermine peace and development in the global village and find ways to address them.

Learn the rules to be observed in public places and the importance of public interests, and develop a will for action to contribute to the common good.

I will teach you. It reads “Nice to meet you.”

2. Be Curious About AI.

I need to have a call with someone who speaks

I have never seen letters like these.

a different language. Please help me.

I can’t read the book.

Learn about AI translation.

I will help you with that as well.

I am so busy!

Have you ever used a translation machine to find out the meaning of a foreign language? With the recent introduction of AI-based translation service , which perform better than simple automatic translations, we now have access to more accurate translation. Besides translation, various AI technology serves the public interest. For instance, there are dubbing services and voice transcribing services that convert speech into text for people with language disabilities.

What if many jobs such as doctors, lawyers, and taxi drivers disappear because of various and

outstanding functions of AI?

AI translating machines quickly translate any

foreign language. It can be used by

anyone with ease.

Because of the AI's excellent translation technology, I’m getting less and less work.

Everyone’s unhappy because AI is monitoring us.

However, AI is taking away our jobs.

What Is Publicness of AI?

❶ Publicness of AI

Many people agree that the development of AI should contribute to the safety and public good of mankind. Publicness of AI can be realized by respecting and observing the norms and values in important documents like the 「UN Charter and the 「International Labor Organization Fundamental Conventions」, etc. Publicness of AI includes the following. Pursuit of common good public interests.

Maintenance of social It's important for everyone in our community to live in a safe and peaceful way, order Enhancement of social responsibility

❷ Danger of AI

We want to create a happy world where many people pursue their happiness, the development and use of AI technology should contribute to both individual and

thus maintain social order. So, when we use AI technology, we make sure it doesn't cause any problems and helps keep our community safe and happy.

In order to enhance the public nature of AI, such as pursuing the public interest and contributing to the maintenance of social order, companies that

work with AI should be responsible. This requires support at the policy and institutional level.

Global leaders are warning about the risks and side effects of AI that might occur when AI evolves into super-intelligence as follows.

In general, while machines increase the convenience of life in various ways, super-intelligence does not. In decades, the concern over super- intelligence will become enormous. AI can be a blessing or a disaster for humans.

Bill Gates

Founder of Microsoft

Stephen Hawking

Physicis

The development of full AI could spell the end of the human race. I fear the consequences of creating AI that can match or surpass humans. Humans, who are limited by slow biological evolution, couldn't compete with AI, and would be superseded.

❸ Importance of publicness of AI

The evaluation system using AI saves a lot of time and costs and even assists in fair selection. However, in some cases, the AI evaluation system used for fair selection rather weakens the public interest. In fact, in the UK, an AI system was used to calculate grades in 2020 on behalf of the college entrance exam 'A-Level', which was canceled due to COVID-19.

Due to COVID-19, this year’s college entrance examination has been canceled.

Instead, we will calculate grades fairly using AI.

I was accepted. Wait, our grades are the same.

That does not make sense at all. This is unfair!

Were you admitted to ◯◯University?

Controversy over the newly introduced college entrance score evaluation results is growing as it was revealed that AI had increased the acceptance rate of certain schools

with better past performance by giving students from those schools higher scores.

No, I wasn’t. How about you?

This biased algorithm of AI sparked a fierce controversy because students from schools with wealthy educational conditions were given higher scores, while those from schools with poor educational conditions got lower scores. These biased results led to social inequality, causing a strong social backlash, and eventually, the AI-based college entrance exam scoring system in the UK was canceled.

To secure the public sphere of AI, we need to ensure that the benefits of AI systems are shared evenly among different groups, rather than being focused on certain individuals.

Looking for AI technology for all

Write down for whom the following AI technology has been developed.

AI technology

Intended user

A service that converts voice into texts in e.g.) Hearing-impaired people real time

A dubbing service that converts text into voice

A service that translates foreign languages

A service that predicts climate change

A service that notifies about fake news

Among AI technologies for everyone, investigate and write down those that are already in use.

AI technology

Actual use case (intended

users, method, etc.)

Effect of AI technology on society

Thinking about AI for all

번역없음

Come up with ideas to develop AI for all. Write about a moment when you needed help.

Write down the people who need help and the problems they are experiencing.

People in need of help

Defining a

problem

Their problems

Write down your ideas on using AI technology for people in need of help on a sticky note.

Share your ideas for the development of AI for all. Put sticky notes with ideas for people in need of help on the chalkboard.

Each student will be given 3 stickers. Among ideas presented by other friends, attach the stickers to practical ideas.

Ideas selected by classmates

Give a presentation on our ideas for developing AI for all. Check out the ideas that received the most stickers from friends.

will undoubtedly commit a crime in the near future.

Discussing the effect of AI on society

As a result of analyzing various data, I predict that that person

I will make him a criminal.

Thanks to AI, the world has become a safer

place without crime.

I will make him a criminal.

There is a risk that someone will Because AI perfectly prevents crime, intentionally use AI for a bad purpose everyone can live a safe life.

to brand an innocent person as a criminal.

Look at the picture above and choose your opinion about the effect of AI on society. Write down your thoughts and write whether you are for or against the use of AI.

Opinion

Support Oppose

Find reasons for the opinion you have chosen and the examples for your opinion.

Reasons

Think about the opponent's objection to your opinion.

Examples

Anticipated objections/questions Counter-argument against the from the opponent opponent’s objection (Put yourself in the opponent’s shoes)

Think about your counter-arguments against the opponent’s opinion.

Anticipated grounds for the opponent’s argument (Put yourself in the opponent’s shoes)

Counter-arguments against and questions about the opponent’s question

Making campaign tools for AI for all

Think about campaign tools for ‘Making AI for all’.

Select ‘usages of AI for all’ or ‘development methods for AI’ that you would like to publicize through campaign activities.

Suggest effective ways to publicize your campaign.

<Examples of campaign posters: students’ creations>

Based on the information gathered above, design a campaign tool to make AI for all.

Campaign title

Type of tool

Role assignment

Necessary materials

Details of the campaign

Make the campaign tool you designed above.

Use your campaign tool to run the campaign for making AI for all.

Review with the ‘Climbing the Ladder’ game

In the empty space where each AI technology for everyone arrives after moving along the ladders, write down an area where the technology can be applied. 1) Pick a ladder to climb. 2) Climb down by making a turn at every rung.

A service that converts voice into texts in real time

A service that predicts climate change

A service that translates foreign languages

2 Create a Sustainable Future Village

•You will learn that integration and collaboration among different groups are essential in the process of developing and utilizing AI technology. •You will be able to explain how to use AI to best serve future generations.

Let me join! I am in!

We need to work together to protect the environment for our future.

Investigate the challenges to building a sustainable future and explore ways to actively engage in addressing them as global citizens.

Identify the challenges facing humanity in the era of globalization and their causes through discussion, and work toward the solutions for these challenges with determination.

Wow! This is huge! Is it true that a wide range of data that can be used in AI learning is here?

Yes, it is. I can use the data stored here to suggest ideas to protect endangered species or reduce pollutants.

Learn about data centers.

A lot of messages and materials I exchanged with my friends are also stored in the data center.

In the age of AI, we need a space to store various data such as conversations and photos shared with friends, and comments in social networking services, in other words, a data center. Many types of data are stored in data centers, and AI can learn from this data to address various problems. For example, we can reduce air pollution by using AI to analyze air pollution data and find out the sources of pollutant emissions.

5000kg

Data centers are making our lives convenient, but the carbon emissions from these centers are worrying.

I heard that we can also utilize AI effectively to reduce energy consumption …

problems through AI can solve many

data learning.

The use of a large volume of data for AI learning also requires a lot of energy.

284019kg

However, a lot of energy is consumed in the

course of operating data centers or using a great deal of data for AI learning.

Developing in a Sustainable Manner

❶ Solidarity of AI and international cooperation

Solidarity means taking shared responsibility for the problems in our society. Considering the impact of the development of AI technology on society, we need to seek agreement and cooperation not only within a country but also in the international community.

❷ Sustainable development goals for future generations

Sustainable Development Goals (SDGS) are solutions to tackling issues that attract great attention worldwide. They consist of 17 main goals and 169 targets to be achieved by 2030. According to these goals, the greatest shared goals of the international community can be broadly divided into three categories. • Universal challenges of humanity: poverty, disease, education, gender equality, refugees, conflicts, etc. • Global environmental issues: climate change, energy, pollution, water, biodiversity, etc. • Economic and social issues: technology, housing, employment, production and consumption, social structure, law, domestic and foreign economy, etc.

The SDGS are shared global goals established for a

better life.

❸ Dreaming of sustainable development using AI

Deleting unnecessary messages and e-mails on our smartphone

can save energy by reducing carbon

emissions.

The United Nations’ sustainable development goals aim to provide opportunities for development without recklessly depleting the resources that future generations should be able to benefit from. Likewise, AI should be developed to meet the needs of the present generation while taking into account future generations at the same time. International Telecommunication Union (ITU) stated that the development of AI can contribute to achieving the Sustainable Development Goal.

SDGS

cultivate crops to improve agricultural productivity.

AI provides personalized education tailored to each learner's level.

equality.

water quality in real time.

of wind power and solar energy based on weather information.

of wind power and solar energy based on weather information.

industries significantly by combining AI and big data.

SDGS

Examples of AI use

AI finds out or predicts areas suffering from food shortage due to drought

conflict, poor harvest, natural disaster, etc., and distributes food in real time.

and AI robotic wheelchairs, etc. are developed.

AI is used to analyze the growing conditions of crop and AI robots

etc., and improve the quality of life of citizens.

AI is used to enhance the ability to examine diseases, enabling effective

treatment based on accurate diagnosis.

analysis.

AI collects and analyzes sexist linguistic practices or behavior patterns in the

form of data and provides a manual to solve these problems, improving gender

population changes and fishing activities.

AI detects pollution sources using sensors and measures and manages

AI recommends the optimal location and size of facilities for efficient production

of tax evasion.

AI recommends the optimal location and size of facilities for efficient production

all involved.

Efficient social infrastructure is built to increase productivity and develop *The above is a reclassification of AI technologies that have

been developed or are being developed, in line with SDGS, based on the examples presented by ITU (2017).

Examples of AI use

AI-based live text or voice services are provided for people with disabilities

Smart cities are built to address traffic, environmental and housing problems

AI presents optimal conditions for consumption and production to reduce

waste of resources.

AI predicts climate-related issues and disasters in advance through data

AI protects marine ecosystems by tracking marine life migrations

AI tracks land animal migration, population changes and hunting activity

to preserve the land ecosystem and combat illegal hunting

AI systems predict crimes to reduce crime rates and even prevent the crime

Cooperation between and among multiple countries or organizations are

carried out to develop or use AI in a safe, ethical and beneficial manner for

In conclusion, stakeholders directly involved in the development and operation of AI systems should make focused efforts together with the many groups under the influence of AI systems to achieve sustainable development for future generations.

Investigating international organizations tackling global issues

Investigate what kind of efforts international governmental organizations and non- governmental organizations (NGOS) make to address global issues. International organizations

United Nations

UNESCO International Labor International Organization

Atomic Energy Agency UN Refugee Agency

Amnesty International Habitat for Humanity

Greenpeace Medecins Sans Frontieres

Major activities e.g.) Global peacekeeping, prevention of wars, and international cooperation activities

Investigate more about international governmental organizations and non- governmental organizations (NGOS) tackling global issues, and play a bingo game using the <Bingo board> on Page 137 of the appendix.

Inventing an international organization for the proper use of AI

Explore the problems caused by AI and discuss how to solve them.

Write down problems caused by AI on sticky notes and put them on the board.

Summarize the problems friends wrote on the sticky notes.

Discuss how to solve the problems stated above.

Invent an ‘international organization for the proper use of AI’ by group.

<Drawing logos>

Name of the organization

Meaning of its logo

Major activities

Give a presentation on the international organizations invented by each group.

Thinking about a sustainable future village

Select and attach an appropriate card for each topic below from among the <Sustainable Development Goal Card> on Page 139 of the appendix.

Watch the following video and divide up the Sustainable Development Goals.

Topic

People

Envir- onm- ent

Pros- perity

Peace

Sustainable Development Goals

Write down which Sustainable Development Goals you are interested in and why.

The reason for your interest

Sustainable Development Goal(s)

Organize groups for each Sustainable Development Goals of interest.

Seek ways to achieve the Sustainable Development Goal of each group.

Problem

Solution

Activity4

Come up with ideas of using AI.

Making a sustainable future village using AI

Describe the details of the project in relation to the Sustainable [Description] Development Goals.

Description

What is this project about?

Explain how AI technology works in this project. [How to use AI] Explain how this project can contribute to society. [Influence]

What is the name of the project?

How to use AI

How does it work?

Project participant

Influence How can this project contribute to the society?

Make your idea more concrete by drawing a sketch with labels.

Share your idea by giving a presentation on it.

Review with a ‘puzzle’

Find the <Puzzle> pieces suitable for the use of AI for sustainable development on Page 139 of the appendix to complete a healthy earth.

Disadvantages of data centers

Sustainable Benefits of data development

centers

Use of AI for sustainable development

2. Create a sustainable future village

Review what we have learned using a mind map.

Making a better world together

1. Make a world where everyone is happy

사진 출처

출처 및 참고 자료

셔터스톡(https://shutterstock.com) •14, 24, 30, 32, 33, 34, 36, 39, 40, 46, 48, 50, 52, 53, 55, 57,62, 68, 72, 74, 78, 80, 84, 86, 87, 90,100, 110, 117, 120,127,135,139

이미지투데이(https://www.imagetoday.co.kr) •98

참고 자료 Ⅰ인간의 존엄성 지키기

15쪽 •성차별 인공지능 채용 프로그램: BBC NEWS 코리아 https://www.bbc.com/korean/news-45820560

•빈곤부터 인권까지 AI를 최적 활용하는 8가지 사례: 이로운넷 https://www.eroun.net/news/articleView.html?idxno=10062

25쪽 •얼굴 인식으로 탑승 수속!: 경향신문 https://www.khan.co.kr/local/Incheon/article/202201271407001

•공항 들어가는 내 얼굴 실시간 인식 누가 동의하나: 미디어오늘 http://www.mediatoday.co.kr/news/articleView.html?idxno=302040

27쪽 •사생활 의미 : [도서] 프라이버시와 빅데이터-빅데이터 시대의 개인 정보 보호와 사생활 •사생활 침해 유형

https://easylaw.go.kr/CSP/CnpClsMain.laf?popMenu=ov&csmSeq=574&ccfNo=2&cciNo=3&cnpClsNo=2: 법제처 찾기 쉬운 생활 법령 정보 •개인 정보 보호 포털 https://www.privacy.go.kr/

28쪽 •개인 정보 보호 포털 https://www.privacy.go.kr/

31쪽 •중국 랴오닝성, 내 아이 교실에 CCTV가 설치됐다고?:대한뉴스 http://www.dhns.co.kr/news/articleView.html?idxno=186676

•지능형 cctv,cctv랑 뭐가 다른 걸까?: 한경 닷컴 https://www.hankyung.com/economy/article/2018-****-5836g

37쪽 •챗봇 ‘이루다’가 보여 준 여성 혐오와 차별의 민낯: 경향신문 https://www.khan.co.kr/national/national-general/article/2021011-****-1011

43쪽 •말하면 실시간 자막으로 … : 유튜브 https://www.youtube.com/watch?v=jMktKfHw6s8

55쪽, 56쪽, 127쪽 •딥페이크로 되살아난 독립운동가들: YTN 유튜브 https://youtu.be/VDRJLW8fi0s

56쪽 •선생님도 딥페이크 피해, 피의자 90%가 10~20대: JTBC 뉴스룸 유튜브

30쪽 •AI스피커가 내 대화 몰래 전송?: 아이뉴스24 https://www.inews24.com/view/1177026

49쪽 •커지는 드론 사업, 늘어나는 사생활 침해: YTN2 https://ytn2.ytn.co.kr/program/program_view.php?key=2018-****-1347457923

https://youtu.be/7pelLk9VVCM

Ⅱ목적에 맞게 기술 사용하기

63쪽 •서울시, 인공지능 빅데이터 플랫폼 구축: 보안 뉴스 https://www.boannews.com/media/view.asp?idx=107868&kind= •구글 번역기에 나타난 ‘일라이자 효과’: 이코노미스트 https://economist.co.kr/2022/10/11/it/general/20221011-****-34818.html 65쪽 •포브스가 꼽은 인공지능을 망칠 수 있는 5가지 편향 : Ai타임스 http://www.aitimes.com/news/articleView.html?idxno=133722

66쪽 •AI for Oceans https://studio.code.org/s/oceans/ 69쪽 •구글 www.google.co.kr 네이버 www.naver.com 70쪽, 71쪽 •엔트리 https://playentry.org/ 85쪽 •악천후에도 자율 주행 가능!: 아이뉴스24 https://www.inews24.com/view/1530322

•자율 주행 자동차, 좌회전하려다가 쿵!: 한국일보 https://www.hankookilbo.com/News/Read/A2021111417090000420?did=NA •차량을 하늘로 인식 … 자율 주행차 안전성 논란: 인공지능신문 https://www.aitimes.kr/news/articleView.html?idxno=25088 86쪽 •로봇 안전 문제 부상: 로봇신문사 http://www.irobotnews.com/news/articleView.html?idxno=12668

•이스라엘, 검문소에 인공지능 로봇 기관총 설치 논란: YTN https://www.ytn.co.kr/_ln/0104_202209271100024266 91쪽, 92쪽 •엔트리 https://playentry.org/

Ⅲ 함께 좋은 세상 만들기

99쪽 •‘기사 없는 택시’의 시대 … : 한겨레 https://www.hani.co.kr/arti/economy/economy_general/888967.html •변호사와 회계사 일자리도 AI가 대체: AI라이프경제 http://www.aifnlife.co.kr/news/articleView.html?idxno=9871 •범죄 예방 VS 사생활 침해: 동아비즈니스 리뷰 https://dbr.donga.com/article/view/1206/article_no/9364/ac/magazine 101쪽 •AI가 준 학점, 가난한 학생을 차별했다: 한겨레 https://www.hani.co.kr/arti/international/europe/959055.html

111쪽 •코알라, 침팬지, 고래까지 … AI로 야생동물 보호한다: 뉴스;트리 https://www.newstree.kr/newsView/ntr202202230008

113쪽 •AI for SDG’s: ITU 유튜브 https://www.youtube.com/watch?v=ND7pvShNdlg 116쪽 •지속 가능 발전 목표 실천을 위한 주사위를 던져 볼까요?: 코이카 유튜브 https://www.youtube.com/watch?v=oWwq1yavkkc

•AI가 기후 변화, 불평등 악화시킨다: 주간경향 https://weekly.khan.co.kr/khnm.html?mode=view&artid=2019-****-1605331&code=114

41, 42쪽 <역할극 이름표>

Ⅰ -3. 인공지능에 편견과 차별을 없애요 <활동 2>

53쪽 <인공지능 기술 카드>

인공지능 스피커

자율 주행 드론 이미지 합성 기술

물류 로봇

자율 주행 자동차 얼굴 인식 기술

53쪽 <목적 카드>

자율 주행 드론

멀리 떨어진 외딴 섬에도 쉽게 택배를 보낼 수 있습니다.

인공지능 스피커 이미지 합성 기술

음성 명령으로 편리하게 음악 듣기, 쇼핑, 가전제품 제어 등을 할 수 있습니다.

53쪽 <위험 카드>

자율 주행 드론

배달 사고가 발생할 수 있습니다.

인공지능 스피커 이미지 합성 기술

악성 코드나 해킹의 공격을 받으면 개인 정보가 유출될 수 있습니다.

얼굴 인식 기술 자율 주행 자동차

운전자가 없어도 편안하고 안전하게 사람이나 물건을 이동시킬 수 있습니

얼굴 정보만으로 쉽게 원하는 곳에 접속할 수 있습니다.

돌아가신 가족이나 위인들을 영상으로 만날 수 있습니다.

늦은 시간에도 쉽게 궁금한 점을 질문할 수 있습니다.

얼굴 인식 기술 자율 주행 자동차

다른 사람이 내 사진을 이용해서 내 개인 정보를 얻을 수 있습니다.

인공지능 오류로 교통사고가

챗봇

나쁜 의도로 만든 가짜 뉴스 때문에 피해를 봅니다.

고객의 개인 정보나 기업 정보가 노출될 수 있습니다.

물류 로봇

사람이 없어도 무거운 짐을 쉽게 정리할 수 있습니다.

위험한 군사 임무를 수행할 수 있습니다.

물류 로봇

물건을 잘못 정리하거나 발생할 수 있습니다. 파손할 수 있습니다.

군사 로봇

군사 로봇 때문에 많은 사람들이 다치고 죽을 수 있습니다.

79쪽 <사고 보고서>

대답한 사람

개발자

자율 주행 자동차의 위험 요소에 대해 사용자에게 안 내했나요?

기술적인 오류가 발생할 수 있다는 사실을 사용자에 게 안내했나요?

공급자

사용자

<사고 보고서>

질문 내용

적합한 절차를 거쳐서 개발했나요?

사고의 원인에 기술적인 오류가 있나요?

자율 주행 자동차 사고 시 책임질 내용이 생길 수 있 다는 사실을 사용자에게 안내했나요?

자율 주행 자동차도 사고가 날 수 있으며 그로 인한 책임을 져야 한다는 사실을 알고 있었나요?

사고 발생 상황에서 교통 규칙을 지켰나요?

사용 설명서를 충분히 이해하고 사용했나요?

피해자 사고 발생 상황에서 교통 규칙을 지켰나요?

대답

◯

△ 궂은 날씨에는 인공지능 도로 인식 기능의 정확도

가 떨어짐.

◯

△ 궂은 날씨로 인해 사고가 일어날 수 있음을 직접 설

명을 하지 않음. 다만 사용 설명서에는 표시됨.

◯

◯

◯

△ 읽어 보긴 했지만, 날씨 상 황에 대한 내용을 기억하

지 못함.

◯

79쪽 <모의 법정 역할극 대본>

등장인물: 자율 주행 자동차 개발자, 자율 주행 자동차 공급자, 자율 주행 자동차 사용자, 피해 자, 검사, 판사 상황: 자율 주행 자동차 때문에 사람이 다치는 사고가 발생하였습니다. 피해자는 피해 보상을 청구하려고 합니다.

판사: 재판을 시작하겠습니다. 검사는 신문해 주세요. 검사: 피해자에게 묻겠습니다. 어떤 피해를 입었나요? 피해자: 정지 신호 앞에서 자동차를 세우고 있었는데, 갑자기 뒤에 있는 자동차가 제 차를 들이 받았습니다. 저는 다리를 다쳐 4주간 병원에 입원을 했습니다.

검사: 개발자에게 묻겠습니다. 자율 주행 자동차에 기술적인 오류가 발생했나요? 개발자: 사고가 발생한 날은 비바람이 거셌습니다. 자율 주행 자동차의 센서가 도로와 주변 자 동차의 정보를 정확하게 인식하지 못했습니다.

검사: 공급자에게 묻겠습니다. 날씨가 좋지 않을 때 자율 주행 자동차의 사고 위험성에 대해 사 용자에게 안내하였습니까? 공급자: 개별적으로 설명을 하지 못했습니다. 다만 사용 설명서에는 해당 내용이 나와 있습니 다.

검사: 사용자에게 묻겠습니다. 사용 설명서를 제대로 읽어 보았나요? 사용자: 자동차를 살 때 따로 설명을 듣지 못했고, 사용 설명서도 제대로 읽지 못했습니다. 검사: 이상입니다. 판사: 최종 판결을 내리겠습니다.

판사: 이것으로 재판을 마치겠습니다. 감사합니다.

91, 92쪽 <사물 카드>

Ⅱ -3. 안전한 인공지능을 만들어요 <활동 3>

114쪽 <빙고 판>

Ⅲ -2. 지속 가능한 미래 마을을 만들어요 <활동 1>

116쪽 <지속 가능 발전 목표 카드>

부록

120쪽 <퍼즐>

많은 에너지가 필요하여 탄소 배출량이 증가해요.

미래 세대를 위해 무분별하게 훼손하 지 않으면서 발전 기화 않으면서 발전 기회 를 갖는 것이에요.

Ⅲ -2. 지속 가능한 미래 마을을 만들어요 <활동 3>

Ⅲ -2. 지속 가능한 미래 마을을 만들어요 <단원을 마무리해요>

환경 데이터를 수집하고 분석하여 다양한 생물을 추적하고 보호해요.