Artificial Intelligence Foundations

Student AI Notebook
Introduction
Welcome to the Artificial Intelligence Foundations course! Throughout the five online modules,
you will have the opportunity to learn about and engage with AI applications.

This document is your AI Notebook and it will be your accompanying guide to complete short
tasks, answer reflection questions, and ultimately work on your AI Design Challenge. Be sure to
have access to it whenever you are working on the course.

You have a few options for how to use your AI Notebook:

1) Make a copy of the Google Doc, save it to your Google Drive, and work through it
digitally.
2) Download as a Word document, save on your device, and work through it digitally.
3) Print this document and work through it on paper.

Throughout the course, you’ll see references to your AI Notebook. Whenever you see this icon,
there is something for you to do in your AI Notebook:

1
Table of Contents
Question Generator 4
Module 1: What is AI? 6
Lesson 1.1: This is AI 6
Lesson 1.2: History of AI 9
Lesson 1.3: AI and HI 13
Module 2: AI and You 14
Lesson 2.1: AI and Humanity 14
Lesson 2.2: Risks of AI 17
Lesson 2.3: Ethics in AI 18
Lesson 2.4: The Design Thinking Method 20
Module 3: Machine Learning in AI 25
Lesson 3.1: Machine Learning and AI Challenge 25
Lesson 3.2: Machine Learning and AI Challenge 29
Lesson 3.3: Machine Learning with Algorithms 32
Lesson 3.4: Machine Learning with Neural Networks 34
Module 4 : Machine Learning in AI 36
Lesson 4.1: Natural Interaction 36
Lesson 4.2: Computer Vision 39
Lesson 4.3: Recommender Systems 41
Lesson 4.4: Autonomous Systems 44
Module 5: Bringing It All Together 46
Lesson 5.1: Project Build Day 1 46
Lesson 5.2: Project Build Day 2 51
Lesson 5.3: Project Build Day 3 57
AI Design Challenge Rubric 59

2
3
Question Generator

Type of Question Definition and Example

Knowledge Questions based on identifying and recalling information.

Asking topic questions about the:

● Who
● What
● Where
● When
● How

Questions that describe what something is.

Example: How is AI used?

Comprehension Questions based on organizing and selecting facts and ideas.

● What is the main idea?

● What would a brief outline of the topic look like?
● What differences exist between ____?

Example: What differences exist between AI and robotics?

Application Questions based on using the topic’s facts, rules or principles to

then apply to something more.

● How is the topic an example of_____?

● How is the topic related to____?
● Why is the topic significant?
● Do you know of another instance where_____?
● Could this have happened in______?

Example: How is AI an example of computer science?

Analysis Questions based on separating the larger portion of the topic into

4
 smaller components.

● What are the parts or features of ______?

● How would you outline/diagram/map this topic?
● How does _____ compare/contrast with_____?
● What evidence can you present for_____?

Example: How does machine learning compare with deep learning.

Synthesis Questions based on combining ideas to form a new whole idea.

● What would you predict/infer from _____?

● What ideas can you add to_____?
● How would you create/design a new_____?
● What solutions would you suggest from_____?
● What might happen if you combined _____ with _____?

Example: What might happen if you combined the power of AI with

online tutors?

Evaluation Questions based on developing opinions, judgments, or decisions.

● Do you agree that ________? Explain.

● What do you think about __________?
● What is most important?
● How would you prioritize______?
● What criteria would you use to assess________?

Example: Do you all agree that someday everyone will know how to
program AI technologies?

5
Module 1: What is AI?

Lesson 1.1: This is AI

What Do You Know About AI?

K-W-L Chart
What I Know What I Want to Know What I Learned
What do I already know about What do I want to know about What did I learn about AI?
AI? AI?

What is AI?

How would you describe AI?

6
What can AI do?
What can AI currently do?

What can AI currently not do?

Concerns About AI
Skill Builder: What questions do you have about AI?

Type of Question Questions About AI

Analysis Example: What are the parts or features of AI?

Knowledge

Comprehension

Application

Synthesis

Evaluation

Reflect

7
Revisit the K-W-L Chart at the start of this lesson. What have you learned today that can be
added to the “WHAT I HAVE LEARNED” section of the chart?

8
Lesson 1.2: History of AI

Evolution of AI
Which events would you consider the greatest in AI advancements for each era?

Era Significant Historical Events

1900s-1940s

1950s

1960s

1970s-1980s

1990s-2000s

2010s

2020s and
Beyond

9
Top Three AI Advancements

List the THREE AI advancements that you consider the most important.

1.

2.

3.

Why did you select these three? What have you learned so far that supports your choices?

What about the future? Think about what technology currently does not use AI. How would
adding an AI feature advance that device?

Create a news headline around one of the major advancements or about a future potential
advancement. Your headline should capture the significance of the advancement.

Event News Headline

10
Imagining the Future of AI
What if you could interview Futurist Mike Walsh? What questions would you ask him about his
career? What else would you ask about the future of AI?

Generate three specific interview questions around the future of AI.

1.

2.

3.

Do you think all futurists would answer your questions the same? Why or why not?

Reflect

In the first activity, how could you tell whether the conversation was between two humans or a
machine and a computer?

In the second activity, which types of questions helped the interrogator determine whether the
answers were from a computer?

11
Think about the purpose of virtual assistants and chatbots. Why do people interact with them?
Does it matter if they can pass the Turing Test? In what situations might it matter? In what
situations might it not matter?

What was Alan Turing trying to prove with the Turing Test? How might the Turing test be useful
for measuring machine intelligence today? In what ways is it not useful?

12
Lesson 1.3: AI and HI

Strengths and Limitations

What are some strengths of AI?

Automatic tasks that have clearly defined rules and can be represented
by data.
Rapid data processing
Complex computations

What are some limitations of AI?

Tasks that require common sense and abstract decisions

Reflect

How can creating art, music, stories or other forms of creative expression alongside AI expand
or limit your creativity? Does the work you produce feel authentic? Are you proud of it in the
same way you would be if you created it without the help of AI?

As long as the creative process is within our control (direct or indirect) AI is an enabler of it.

It enhances my creativity

The work is still authentic in this case

I would probably feel proud the same way

13
Module 2: AI and You

Lesson 2.1: AI and Humanity

AI for Good

Use the table to capture the ways that AI is being used to help people and improve their lives in
these different industries.

Industry Benefits of AI

Health Provide enhanced mobility for people with disabilities, can receive
operation and care from everywhere by remote, with some doctors that
know how to move in there

Safety Protect people from cyber crimes, know by maps which are the place
more infamous

Environment Increase efficiency in agricolture, can grow the plants with tecnology,
with a simple click

Education Give timely, personalized coaching to increase learning, do remote

school form anywhere (for exemple can do MIT even thought you are
form india)

Transportation Reduce the number of collision, had only self-guided car, with all the
informations to work and to remove all the accident (like the tesla one)

14
Cities Have the maps to move everywhere in the city, and the opportunity to
follow the right path

15
Benefits to Humanity

Think Outside of the Box: What are other ways AI can be used to help humanity?

List two or three “what if” questions below. Don’t limit your ideas.

1. can receive operation and care from everywhere by remote, with some doctors that
know how to move in there

2. had only self-guided car, with all the informations to work and to remove all the accident
(like the tesla one)

3. do remote school form anywhere (for exemple can do MIT even thought you are form
india)

Human-Centered Design

In your own words, describe what human-centered design is.

Human-centered design can help human to focus on their perspective, to encourage and
empathize human to make their own idea, to believe and make them

Reflect

As you use explore of the listed tools, either as a user or as a creator, answer these questions:
1. What problems does this tool solve or what task does it support?

It easier for anyone to start exploring machine learning, set out to explore wheter
machine learning that could be used to writers to in unblock and enrich their process

2. Who is it for?

For writers, for technicians, there is also a machine learning for kids

3. What features best meet the user’s needs?

16
Knowing the operation of the machine learning, knowing how to program something,

17
4. What features should be included to better meet the user’s needs?

The users needs including a better preparation on what they have to do

18
Lesson 2.2: Risks of AI

Negative Impacts of AI

How can some uses of AI be harmful to humans?

For safety AI can cause some problem when that work without human surgeon ,In data AI can
change what people prefer, AI algorithms could lead to biased reccomendation to decision-
makers, deep fakes could made fake some speech about a person, making him say
something that he don’t believe, in the environment the energy can be converted to carbon
dioxide emission, AI in jobs can help and accelerate some process, but also can replace
something that we need, AI can let us have more prejudices

Types of Bias
Reflect on your personal experience with bias. What are some ways that you have shown bias or
experienced bias? How can practicing empathy, which is assuming the perspective of someone
else, help prevent bias?

Personal bias: a judment that we had for a thing or for a people that we don’t really know, but
based on stereotypes

Cultural Bias: making judment on a culture, making comparison with other nations

Emphty gap bias: The tendency to underestimate the influence or strength of feelings,
in either oneself or other

Stereotypiting: Expecting a member of a group to have certain characteristics without having

actual information about that individual.

Anchoring bias: To rely too much on one trait or piece of information when making decisions
(usually the first piece of information that we acquire on that subject).

Bias blind spot: The tendency to see oneself as less biased than others, or to be able to
identify more cognitive biases in others than in oneself.

Confirmation Bias: The tendency to search for, interpret, or focus on information in a way that

19
 confirms one’s preconceptions.

Bandwagon Bias: The tendency to do or believe things because many other people do.
(Groupthink)

I think that bias are a bad things, but also I think that is impossible haven’t bias,
because human judge all the things, in my life I notice that I have some time
Bandwagon bias, also some stereotypiting, but in the last year I limited my prejudice,
‘cause I notice that I had a lot of that, and open the eyes to see how many bias we
had is the only thing that we can do to don’t let us influence by that

Bias in AI-Decision-Making
How can biased decision-making happen in machines?

Machines gave us a lot information that obviously influence us, and the information that
machines gave us are influenced by the searches that we had done, so if we reaserch only
things that interest a man, probably if we search for exemple “tennis player” in the picture,
we’ll foud only male tennis player.

What is one way to avoid algorithmic bias?

open the eyes to see how many bias we had is the only thing that we can do to don’t
let us influence by that, so recognize when the algorithm do that can be usefull

Lesson 2.3: Ethics in AI

Ethical or Unethical?

Reflect the Ethical or Unethical survey. Were you surprised by the results? If so, what surprised
you about the results?

20
 I’m surprised to see that the 55% of people find etichal that a robot simulate some fake
emotion and that we can fall in lve with them or had some relationship, ‘cause I think that this
type of feelings, if them are fake can hurt a lot human

Ethics Questions

What concerns do you have about AI and its impact on individuals and society? Share 1-2
ethical questions related to a particular use of AI. What are some ideas you have for addressing
these issues?

Ethical Question Ideas for Addressing

I am concerned about the

possibility that someone can For this problem can help increasing the privacy of the
had a lot of my data and can do information that are exposed on the platform

what he want

A problem on AI could be that

internet should be the platform Focus less on the searches what the people had search
where the people can explore yet, permitting to explore everything

and go out from their cave, their

habits; but in reality we had
always the same information,
and that dies nothing else that
close us in our cave

21
Engage: Evaluate an AI System

Which AI technology did you explore?

Deep angel, affectiva, talk to transformer, gan paint

1. Does the technology promote well-being?

2. Does the technology protect user’s data and privacy?

3. Does the technology give insight into how it works?

22
Lesson 2.4: The Design Thinking Method

Design Thinking and You

How can the design thinking method help to build human-centered AI applications?

Design thinking method help human, ‘cause with a lot of information about empathize, define,
ideate, prototype a group of people can elaborate some incredible ideas that can help to make
situation more appreciable

Stages of Design Thinking

What are the five stages of the design-thinking process?

1. Empathize

2. Define

3. Ideate

4. Prototype

5. Test

Engage: Create an Empathy Map

Practice creating an empathy map by using one of the scenarios.

1. Read each scenario and the related personas.

2. Select a scenario and person to focus on for your empathy map.
3. Think about the perspective of the user and complete the empathy map.
a. Write the name of your persona in the middle of the map.
b. Use the information from the scenario, add observations to the appropriate
section of the map.

23
24
Scenarios and Personas

Scenario 1: Tutor Me
You’ve been asked to build a mobile app that will help connect students and tutors.

● Persona 1: Jack is a high school student and is focused on maintaining a high GPA to
increase his chances of getting into his first-choice college. He is struggling with his
Spanish class and wants to find a tutor. He is looking for someone in his neighborhood
who he can meet with after school, possibly Saturday mornings.
● Persona 2: Samantha is a Spanish-speaking high school student who would like to make
a little extra money by helping other students. She hopes to be a teacher one day and
thinks being a tutor would help her gain experience and build her resume. She would like
to offer her services to students looking for a Spanish tutor.
● Persona 3: Mr. Jasminder Smith is a high school teacher and has several students
struggling with their Spanish assignments. He would like to be able to direct his students
to available tutors to help them improve their grades and catch up with the rest of the
class. He also wants to be able to check the progress of his students to ensure they are
taking appropriate steps to improve.

Scenario 2: Rapid Rescue

You’ve been asked to build a mobile app that will help people find lost pets.

● Persona 1: Lisa’s small dog got scared by a car during a walk, broke the leash, and ran
away. Lisa desperately wants to find her precious dog as quickly as possible. She wants
to search a list of found pets or post information about her dog so others can help her
find him.
● Persona 2: Didi found a lost dog and wants to find the owner. The pet has no tag, so Didi
cannot call the owner directly. They don’t see any flyers in the neighborhood.
● Persona 3: Petra works at a local pet shelter and truly cares for the cats and dogs that
end up at the shelter. She wants to keep informed of any lost pets in the area in case any
of them arrive at the shelter. She wants to be able to identify lost pets and reunite them
with their owners before they get added to the adoption website.

Scenario 3: Style Me
You’ve been asked to build a mobile app that kids can use to create new outfits from the clothes
and accessories they already own and get advice on what to buy to enhance their wardrobe.

● Persona 1: Sam is a middle schooler who is tired of wearing the same outfits every
week. He wants some advice or ideas for mixing and matching the clothes he already
owns to create new looks.

25
 ● Persona 2: MaryAnn would like to learn more about style and accessorizing her outfits.
She owns a lot of clothes and accessories, but isn’t always sure what looks good
together. She would like help matching tops, pants/skirts, and jewelry.
● Persona 3: Nicki and Jo would like to enhance their wardrobes. They want to purchase
items wisely to match what they already have. They want some advice on pieces they
could buy that would work with multiple items they already have.

Scenario 4: BullyMeNot
You’ve been asked to build a mobile app that kids can use to find help when they are being
bullied or provide resources to others who want to help prevent bullying.

● Persona 1: Claire is a middle schooler who gets teased daily for one reason or another.
She doesn’t know how to talk to her best friend, her parents, or her big brother about the
stressful situation. She wants to get help to learn how to cope.
● Persona 2: Nigel has a best friend who gets picked on frequently at school. He would
like to help his friend and learn more about preventing bullying behavior with his peers.
● Persona 3: Tracy suspects her daughter is being bullied at school, but Sasha denies it
when Tracy asks her about her concerns. Tracy would like to find resources for her
daughter to help with the stressful situation.

Empathy Map Example

26
Empathy Map

Says Thinks

Did you see my dog, he was ran away few How will I do? I need someone to console me
seconds ago, ‘cause his leash was broken,
please help me to find him I had to pay more attention to the leash, I’m
stupid

Where I can print some sheet with the photo I must to find him, I can’t stay with this knot,
of my dog, I must to find him and without my dog

Persona

Lisa

Does Feels

Ran toward the direction where the dog went Sad

Ask to everybody nearby if they have seen the Angry with herself
dog
nervous

27
Reflect

How might an activity like this empathy map activity help you to understand and get to know
your users? If this were a real-life scenario, what questions might you ask your potential users?
What might your next steps be?

28
Module 3: Machine Learning in AI

Lesson 3.1: Machine Learning and AI Challenge

The Machine Learning Process

Provide a basic overview of how machine learning works.

Machine learns from data and

training without being explicitly
programmed. Essentially, machine learning takes in data, finds a
pattern in the data, and makes a decision.

What are the key components required for a machine to learn?

Algorithms, data, features

Get Inspired
Log information from the tools and applications you explored.

Machine Learning Tool

Research Notes

Tool Explored: Teachable MAchine

URL: https://teachablemachine.withgoogle.com/train/audio

29
 Interesting Solution/Application 1: Language recognition Italian english

URL:

How it Works:

Able to detect when I speak English or Italian

Interesting Solution/Application 2:

URL:

How it Works:

Getting Started: AI Design Challenge

Step 1: Empathize – Choose Your User

Who do you want to design your solution for? Who would you like to help increase their
knowledge or improve their skills?

Choose someone:
❏ that you have regular access to
❏ who is willing to be interviewed
❏ who is willing to provide feedback on your prototype

You may choose a small group of people if you like.

Name of your user:

30
Step 2: Empathize – Interview, Observe, Research

Write down 8-10 questions below for your interview that will help you understand your user
better. You may not end up using them all. You can use the example questions to start you off.

Interview tips:
● Avoid yes or no questions. Or at least follow up with why and how?
● Use questions that begin with “tell me about the last time you _______?"
● Ask open ended questions that lead to stories, as stories often provide insights to
designers.
● Ask questions about how the user feels about certain situations.

Engage: Conduct Your Interview

Step 3: Empathize - Conduct an Interview

Now conduct your interview! Record notes on your user’s responses. With the person’s
permission, you may also want to audio record the interview.

# Interview Questions Users Responses

What is a topic you’d like to learn

1 more about or a skill that you
would like to improve upon?

Why are you interested in that

2
topic or skill?

What is your current level of

3 knowledge about this topic or skill
level?

31
 How do you currently practice this
4 skill or gain knowledge on this
topic?

10

32
Lesson 3.2: Machine Learning and AI Challenge

Deep Learning

Briefly explain the relationship between machine learning and artificial intelligence.

AI is the ability to do tasks that requiring human intelligence, machine learning do the same
but without human involvement, they acquire skills with experience

How Do You Ensure Quality Data?

Complete the checklist for how to evaluate the quality of a data that can be used to teach a
machine. What criteria should exist?

Data should be:

❏ Recent (data relevant to what is being investigated)

❏ Complete (No missing information)
❏ appropriate
❏ accurately labeled
❏ large
❏ representative sample
❏ checked for bias

33
Engage: Train a Machine
Log your information from the activity you selected.

Machine Training Tool

Research Notes

Option Selected:

URL:

What project(s) did you work on?

What data sets did you use to train a machine?

Other notes:

34
Reflect

What did you find most interesting about machine learning from your explorations today?
Answer in 2-3 sentences.

I think that is incredible how the machines can be influenced by the bias, and that alllt he
things that they do is based on that

35
Lesson 3.3: Machine Learning with Algorithms

Algorithms

Explain how algorithms are used to teach machines in 2-3 sentences.

Algorithms are used to teach machines, because when they see some new data they
understand that and remember that to exploit it in another moment. Also with a tree model
make decision, if I choose something the next chose Is based on that

Create a Decision Tree

Choose one of the below topics or select your own topic and create a decision tree. You may
use an online tool to create your decision tree or you can create your diagram using paper and
pencil.

Suggested Topics:
● What will I do today?
● What animal is this?

Algorithmic Bias

What are two challenges and solutions related to algorithmic bias in AI?

For me, machines to be better have to not influenced a lot by bias, but can learn also from
another things, and not even with singular data, be influenced of the whole data can be usefull

Also can be usefull that the machine can learn something about the person with some test or
some question, to be ready in different way to organize their work

36
Reflect

What did you learn about machine learning from your explorations today? Answer in 2-3
sentences.

Now I know more the step that machine use to work well, and to make successful decisions

37
Lesson 3.4: Machine Learning with Neural
Networks

What Are Neural Networks?

Explain the basics of how artificial neural networks process information.

Artificial neural works with some essential information that can recognize when see
something, for example the machine know that a cat had 4 legs, had whiskers, had fur, and
that is small, so with this information the machine can recognize when there is a cat and
when no, but machine can be wrong, ‘cause also dog had all these features

Machine Learning and Deep Learning

What is the relationship between machine learning and deep learning?

Machine learning know and learn with the experience, and with that have some skills, instead
deep learning is based on algorithms inspired by the human brain, machine learning isn’t
influenced by human, we can say that the experience that machine learning do can be usefull
to the algorithms of deep learning

38
Engage: AI Design Challenge - Empathy Map

Use the information you gathered from your interview, observations, and research to complete
an empathy map for your user.

Says Thinks

Did you see my dog, he was ran away few How will I do? I need someone to console me
seconds ago, ‘cause his leash was broken,
please help me to find him I had to pay more attention to the leash, I’m
stupid

Where I can print some sheet with the photo I must to find him, I can’t stay with this knot,
of my dog, I must to find him and without my dog

Persona

lisa

Does Feels

Ran toward the direction where the dog went Sad

Ask to everybody nearby if they have seen the Angry with herself
dog
nervous

39
Module 4 : Machine Learning in AI

Lesson 4.1: Natural Interaction

What is Natural Interaction?

During this lesson, look for information to fill in this table.

Share three common areas of natural interactions, supporting examples of these areas in
everyday AI applications, and the limitations of these natural interactions.

Areas of Natural Examples in Everyday AI Applications Limitations

Interactions

SPEECH Alexa, siri, all the voice assistant They do really specific
tasks, so in some cases
they don’t recognize some
words, the accent, the child
voice

AFFECT IMDR Is difficult go in deep of

emotions and understand
how someone feel

GESTURES
Kinect for xbox Sometimes they are unable
to detect motion for some
factors like distance from
the sensor, and also we can
say that this machine have
some limitation because
there aren’t a really rule for
gesture, different people use

40
different gesture

41
Natural Interactions and Human Impact

Can you think of other ways that applications that use natural interactions through speech,
gesture, and affective computing impact humans? Jot down one idea here.

That applications impact humans in everyday life, because someone that can feel alone with
some affective Ai can feel better, cause they respond based on. The emotion that the person
express

Reflect
Answer the following questions as you reflect on the experience.

1. Which application did you interact with?

With siri and alexa

2. What area(s) of natural interaction does the program address (speech, gestures, affect)?

speech

3. Was the response to your input expected and accurate? Why do you think that is?

Yes, but not always, because sometimes they didn’t understand what I mean

4. How might an application like this be used to help people learn or improve their skills?

This machine are usefull to save time, and ask something that we don’t know in some
seconds

42
These ideas may help you as you plan your AI Design Challenge.

43
Lesson 4.2: Computer Vision

What is Computer Vision?

Use the table to share the basics of how computer vision works.

Forms of Computer How It Works

Vision

Image Sorting images into different categories

classification

Recognizing and classifying a specific object

Object detection

Object tracking Using object detection, following images as they move from frame to
frame

Facial recognition Recognizing a human face and identify the actual person

Image Separating an image into multiple areas and analyzing them separately
segmentation

Edge detection Identifying the edge of object

Pattern detection Finding patterns related to color, shape, objects

44
45
Applications of Computer Vision

Share ways computer vision is used in everyday applications.

We can use a lot of applications that can hep us in everyday life, for example google leans is
useful for translate some word from one language to one other, also to ease car accident
there are so many self-driving car, for grow well the plants data-driven-farming is good

Can you think of other uses of computer vision that would benefit people? Share one or two
ideas.

Robots that can help people to doing some important decision, focusing on what that person
really want, or motivational coach robot to doing sport in correct way

Computer Vision and Human Impact

How have applications with computer vision impacted humans?

Have application change our life, because they let make some thing that help some people
that can’t do something, and facilitate other people to do specific task, for example there is an
app for Czechs that say what he can’t see, what he had in front of, and this is really usefull

46
Lesson 4.3: Recommender Systems

What are Recommender Systems?

During the course of this lesson, look for information to populate this table.

Identify the two types of recommender systems and explain how each type works. Provide
examples of the two types of recommenders.

Recommender How It Works Example

System

Content- based Recommend information that can If I search some crime book,
filtering interest to a person, because he/she probably I will find some
searched something that is similar suggest about a crime best
seller

Collaborative Make recommendation based on If me and other people order

filtering similarities between two users protein bar , probably I will
find suggest about some
new economic bar

Recommender System Scavenger Hunt

Pick a different sport from what you previously selected during the Engage activity. (Select one
that preferably you are not as interested in.) See if you notice any differences from your initial
experience.

Complete the table on the next page as you explore the platform.

You will uncover the ways in which YouTube or Google Play uses AI. For each distinct feature,
describe:
1. the data that is being used to influence the recommendation algorithm

47
 2. what the algorithm is trying to learn from that data,
3. what is the algorithm “optimizing” for. In other words, what does the platform want you
to do; what are the goals? For example, online shopping platforms may be “optimizing”
for items clicked because they want you to buy products.

Feature Data What is it trying to What is it optimizing

“learn”? for?

Want that I persist to

It is trying to learn search what they think,
Suggested search My previous what I prefer to because they know
research search, and what I well what I just
(autofill) want in this time searched compared to
what I didn’t search

It want to recommend
It is trying to learn
what it think that I want
What I search in the what are my interest,
based on my recent
Recommendations past and if are the same of
research, and similar
the past
research

What I comment in It want to suggest

Is trying to learn which
Comments section the past, in other good word for my
are my ideas
video though

It is trying to learn
It want to give me the
What I search in the what are the
Search results information that I want
past information that
per first
impact on me

Ads
What I search in the It is trying to learn It want to give me the
past what ads I open ads that I wat to

48
 immediately receive

It want to give me
Auto play Which video I have It is trying to learn
some video that can
(YouTube only) seen what video I like
interest me

Now scroll to the bottom of the page and select and read the Privacy section of the platform
(bottom middle on Google Play, bottom left-hand sidebar YouTube). How is your data being
used? How do you feel about that?
I found always the same video, or something that is similar to it, and that for some way is
positive because save time to search this type of video, but in other way is negative because
limit my possibility to know something else, and have new interest that I didn’t know to have

49
Lesson 4.4: Autonomous Systems

What are Autonomous Systems?

Describe the basics of how autonomous systems work.

Autonomous system work is based on adapt to reach a goal and to sense their environment

Autonomous Transportation

Identify the examples of autonomous systems in everyday AI applications.

Autonomous transportation is a system that let reduce incredibly car problem, ensuring safety
and increased mobility

Reflect
Why did you make some of the decisions you made during the MIT Moral Machine activity?
What do you see as some limitations of autonomous vehicles? What are some ethical questions
regarding autonomous vehicles?

Definitely autonomous transportation are more safe, but life is unexpected and machine
sometimes can’t react well as a human when there is an unexpected event

50
AI Design Challenge
Over the past few lessons, you have been exploring different machine learning tools and
applications. You have also been spending time with your target user(s) to gain insights about
their needs and preferences. Now it is time for you to begin focusing on the specific problem
that you are solving for your user. Review any notes from observations, interview responses,
and your empathy map. Use insights from those sources to craft your problem statement.

You may use this sentence starter as a guide:

_______[user] needs a way to ______ [do what] because _____[why].

Problem Statement:

A tennis player needs a way to calm down when the game goes differently than he had
expected because he is one of the most talented player of the world, but when her anger go
out he lost a lot of match

51
Module 5: Bringing It All Together

Lesson 5.1: Project Build Day 1

AI Design Challenge Rubric

As you continue to work on your AI Design Challenge, use the rubric (found at the end of your
notebook) to self-assess your project.

Ideate
Braindump instructions:
1. Reread your problem statement from the previous lesson (previous page).

2. As you generate ideas keep your problem statement in mind and respond to the
challenge question:
How might we use the power of machine learning to help people increase their knowledge
or improve their skills?

3. Set a timer for five minutes and write at least five ideas below.

Idea #1

Idea #2

Idea #3

52
 Idea #4

Idea #5

Ideate with Constraints:

Due to time, toolsets, and expertise, you now should simplify your idea based on your comfort
with the tools. At a minimum, the solution must use machine learning to recognize and classify
an image, sound, or pose.

Take another five minutes to refine the ideas for your solution. Write up to three ideas for your
solution that can be designed within a limited timeframe (about one hour total) and using one of
the tools that you are comfortable with.

Idea #1

Idea #2

Idea #3

Choose an Idea:
Choose one idea to focus on as you prototype a solution.

Briefly summarize the idea for your solution in a few sentences and be sure to identify the tool
that you will use.

Example
My AI solution will help someone learn yoga poses. I will train a ML model using Teachable
Machine to recognize different yoga poses. The tool’s confidence level at identifying the pose will
help the person know if they have done the pose correctly and how they can adjust their position

53
to increase the machine’s confidence level. A high confidence level means that the person is
doing the pose correctly.

Your AI Design Challenge Idea:

Plan for Prototyping

Answer the following questions to plan for designing your prototype:

1. What is the name of your AI solution?

2. What tool will you use to develop your design?

3. What is the goal of your solution?

4. How will it help the user to learn or practice a skill?

54
5. What ethical considerations are included in your solution? Is there potential for the
solution to harm anyone or introduce bias?

6. What data will you use?

7. How many different classes or categories are needed?

8. How will your solution gather or use data to make decisions?

9. What decisions or outputs will your tool generate? Will it take any further action after a
decision is made?

55
10. Map the user’s experience with the tool. What key steps will the user complete while
interacting with your tool? Describe what the user does first, what happens next, and so
on. It is okay if you need to supplement the tool with instructions for the user. For
example, you may tell your user to complete a task and the tool will indicate whether
they are correct by classifying an image, sound, or pose with a high level of confidence.

1.

2.

3.

4.

Example
1. The user does a yoga pose in front of the computer webcam.
2. The program indicates the name of the pose and the confidence level that the
name of the pose is correct.
3. The user keeps practicing the pose until the confidence level is high and they are
satisfied that they have done the pose correctly.
4. The user repeats this process with another pose.

56
Lesson 5.2: Project Build Day 2

Prototype
This lesson is dedicated to working on your prototype. Your prototype should be complete by
the end of today’s lesson.

Instructions for working prototype:

1. Gather the training data you need to train your model. If you are generating data in real-
time, for example, posing in front of a webcam, list the actions or items you will show.
2. Follow the instructions for the tool you are using and train your model.
3. Test your model using new data to see how well it recognizes or responds to the
information.
4. Add more training data as needed to increase the tool's accuracy.

Instructions for paper prototype*:

1. Use the storyboard template starting on the next page to sketch the details of each
screen the user will encounter as they interact with the tool.
2. Think about how the user will interact with your solution. Show how users will transition
from one step to the next and include features of the tool (screens, buttons, overall
layout, etc.).
3. Provide a brief explanation that explains each step in the user’s experience. Include
information about how the solution uses data, makes decisions, and the final output or
action.

*Although it is called a “paper prototype” you may also choose to create a digital prototype
using a drawing tool on your device or a presentation tool like PowerPoint or Google Slides.

57
Paper Prototype – Storyboard Template

58
59
60
61
62
Lesson 5.3: Project Build Day 3

Test
Tips for testing with your user(s):
● Take detailed notes in your notebook as you observe and ask questions during testing.
● When you test your prototype, allow your user to experience the solution without
explaining it. Give only basic information to get them started, but let them explore how it
functions. If you are testing a paper prototype, allow your user to examine the visual
representations and read the explanations for each step.
● Ask them about their experience or impressions as they are exploring the prototype. You
might ask, “What are you thinking as you explore this tool?”
● During testing of a working prototype, allow them to make mistakes. Don’t correct them
right away if they do something wrong. This is valuable information that you can use to
determine if something is unclear about your solution or how users might interpret it in a
different way.
● Take note of their questions. These questions provide insight into areas that are not
clear in your design and can also provide inspiration for new features.
● Once the user has completed the testing, ask them a few follow-up questions. You may
come up with your own questions, but here are a few examples:
○ How did this solution make you feel? Why?
○ What confused you?
○ What surprised you?
○ What do you wish the tool would do? Why?

Testing Notes

63
Reflect
Answer the following questions based on your testing:

1. What did you discover about your prototype?

2. What areas might you improve?

3. What did you learn about your user?

64
65
AI Design Challenge Rubric
Element 3 points 2 points 1 point 0 points Points
Given
I completed all the steps to
empathize with a user:
● I chose a target I partially I completed one of I did not
user. completed the the steps required complete any
● I prepared
Empathize steps to to adequately of the steps to
questions.
empathize with empathize with my empathize with
● I conducted an
my user. user. my user.
interview.
● I completed an
empathy map.
I crafted a
problem I crafted a problem
I crafted a problem
statement that statement, but I I did not
statement using insights
used some of the did not rely on complete a
Define from my observations,
insights I insights learned problem
interview responses, and
discovered from the statement.
my empathy map.
during the empathize stage.
empathize stage.
I completed all the ideate
steps:
● I spent time
generating and
refining my idea.
I partially
● I considered the
constraints I would completed the I completed one I did not
Ideate need to work with ideate stages to step during the complete the
to create my select my idea ideate stage. ideate stage.
solution. for a solution.
● I chose one idea to
focus on to
prototype a
solution.

Prototype I completed the prototype I completed a I only completed, I did not build a
steps: portion of the or partially prototype of
● I answered the steps toward completed, one of my solution to
questions around prototyping my the prototype my user needs
my plan for solution. steps.

66
 designing my
prototype.
● I selected the type
of prototype I
would build
(working or paper)
● I worked on the
prototype of my
solution

I tested my
solution and
I only partially
followed up in
tested my solution
some ways to
and had sought
receive a portion
I tested my solution and limited feedback
of feedback from I did not
received feedback. I also from my users.
Test my users. I complete my
reflected and planned how
completed some test.
to iterate on my design. I did not reflect or
reflection and
plan any further
considered the
iteration on my
iteration still
design.
needed on my
design.
I planned a
prototype design
that was mostly I planned a
based on a clear prototype design
I planned a prototype AI goal. I that was limited in
design with a clear AI goal. I somewhat an AI goal. I
I did not plan a
identified appropriate data identified trained and tested
clear goal or
Knowledge of AI to train and test the AI appropriate data the AI system with
test the system
system. I ensured that the to train and test inappropriate
with data.
data was organized into the AI system. I data. I did not
appropriate categories. mostly ensured ensure the data
that the data was was organized
organized into appropriately.
appropriate
categories.
I identified how AI could be I considered a I only partially I did not use
Critical and a creative solution to a solution to a worked at a critical or
Creative problem. I considered problem that solution that could creative
Thinking constraints during the used AI. I use AI. I did not thinking efforts
design process. Challenges partially recognize in my design

67
 recognized
constraints
during the
constraints or
process. I
addressed
addressed
challenges that
with the solution were challenges as
came up in the
anticipated and addressed they came up in
process. I mostly
during the design process. I the design
forgot about the
mapped and considered the process. I
user during the
user’s experience from remembered at project.
design process. I
beginning to end. I reflected times the user
only partially
on the insights gained and experience
answered, or did
areas of growth from my during my
not complete, the
design experience. design. I
reflection
answered the
questions at the
reflection
end of the process.
questions at the
end of the
process.

Total Points

68