Class VIII

Artificial
Intelligence
By:
Computer Department
DPS-Bopal
2021-22

Page 1 of 25 Computer Department, DPS-Bopal

 Unit 1:: Excite

¨ What is Artificial Intelligence

Artificial Intelligence is a field of computer science aimed at developing machines that
are intelligent enough to do certain tasks which are otherwise considered fit to be
performed by human beings only.

In artificial intelligence, computer systems try to partially mimic the functionality of

human brain. i.e., using inexact and incomplete knowledge to make prediction. On of the
greatest advantage of AI systems is that they have the capacity to enhance and improve
their learning themselves, over a regular period of time.

¨ Types of Artificial Intelligence

Mainly there are two types of Artificial Intelligence.
Artificial
Intelligence

Artifical General
Artifical Narrow
Intelligence
Intelligent (ANI)
(AGI)

¨ Artificial Narrow Intelligence (ANI)

An existing type of AI that is capable of performing simple tasks requires basic
intelligence. The intelligence of these system is narrow in the sense that they are only
capable of performing only one task, but in many cases, they can perform this one task
better than human beings.

For example, Google Home and Alexa can recognise and speak multiple languages. The
number of languages spoken and understood by these systems are more than what human
can speak and understand how ever these system are not capable of doing everything that
human does for example Google Home or Alexa cannot drive a car.

So what exactly an ANI can do?

à Processing video for understanding them. For example, recognising when an intruder
enters a house or when small children are in danger in the house
à Performing routine of everyday tasks, which might require some limited decision
making. For example, ordering milk, switching off power in empty rooms, Switching on
ACs when needed.
àPerforming simple tasks done by secretaries. For example taking dictation, scheduling
appointments, taking minutes of meeting, reading emails etc.
àPerforming tasks that require processing of data. For example, assisting doctors in
diagnosing diseases based on symptoms.
Page 2 of 25 Computer Department, DPS-Bopal
 ¨ Artificial General Intelligence (AGI)
Artificial General Intelligence refers to the science fiction idea of artificial intelligence.
i.e., the computer systems, which are at least intelligent as human beings, if not more.
These systems will be capable of learning and developing themselves. They will not be
limited by the tasks that they can perform. These systems are in the research stage, and
much breakthrough has not been made regarding them

Weak AI VS Strong AI
à Weak AI is the intelligence that mimics intelligence but does not process it or in other
words, acts intelligently without having any intelligence. All of the AI systems in
existence today fall under this category. For example, a self-driving car might recognise
that it is raining, but will not understand what rain is. A smart home security can
recognise intruder without understanding what an intruder is.
à Strong AI systems are theoretical systems that actually process intelligence. These
systems currently are in the research stage.

¨ Is there another way to classify AI?

There are other parameters using which we can classify AI. For example time.

In above image we can observe that how four waves in different timelines categorize AI

We can also classify Ai based on functionality

Page 3 of 25 Computer Department, DPS-Bopal

 ¨ What contributes to Artificial Intelligence
Artificial intelligence is a science and technology based on disciplines such as Computer
Science, Biology, Psychology, Linguistics, Mathematics, and Engineering. A major
thrust of AI is in the development of computer functions associated with human
intelligence, such as reasoning, learning, and problem solving.
Out of the following areas, one or multiple areas can contribute to build an intelligent
system.

¨ What is Artificial Intelligence technique

In the real world, the knowledge has some unwelcomed properties −

• Its volume is huge, next to unimaginable.

• It is not well-organized or well-formatted.
• It keeps changing constantly.
AI Technique is a manner to organize and use the knowledge efficiently in such a way
that −

¨ It should be perceivable by the people who provide it.

¨ It should be easily modifiable to correct errors.
¨ It should be useful in many situations though it is incomplete or inaccurate.

Page 4 of 25 Computer Department, DPS-Bopal

 ¨ Programming without Artificial Intelligence or with Artificial
intelligence
The programming without and with AI is different in following ways −

Programming Without AI Programming With AI

A computer program without AI can A computer program with AI can answer

answer the specific questions it is meant the generic questions it is meant to solve.
to solve.

AI programs can absorb new modifications by

putting highly independent pieces of
Modification in the program leads to
information together. Hence you can modify
change in its structure.
even a minute piece of information of
program without affecting its structure.

Modification is not quick and easy. It

may lead to affecting the program Quick and Easy program modification.
adversely.

¨ Application of Artificial Intelligence

AI has been dominant in various fields such as −
¨ Gaming − AI plays crucial role in strategic games such as chess, poker, tic-tac-
toe, etc., where machine can think of large number of possible positions based on
heuristic knowledge.
¨ Natural Language Processing − It is possible to interact with the computer that
understands natural language spoken by humans.
¨ Expert Systems − There are some applications which integrate machine,
software, and special information to impart reasoning and advising. They provide
explanation and advice to the users.
¨ Vision Systems − These systems understand, interpret, and comprehend visual
input on the computer. For example,
o A spying aeroplane takes photographs, which are used to figure out spatial
information or map of the areas.
o Doctors use clinical expert system to diagnose the patient.

Page 5 of 25 Computer Department, DPS-Bopal

 o Police use computer software that can recognize the face of criminal with
the stored portrait made by forensic artist.
¨ Speech Recognition − Some intelligent systems are capable of hearing and
comprehending the language in terms of sentences and their meanings while a
human talks to it. It can handle different accents, slang words, noise in the
background, change in human’s noise due to cold, etc.
¨ Handwriting Recognition − The handwriting recognition software reads the text
written on paper by a pen or on screen by a stylus. It can recognize the shapes of
the letters and convert it into editable text.
¨ Intelligent Robots − Robots are able to perform the tasks given by a human.
They have sensors to detect physical data from the real world such as light, heat,
temperature, movement, sound, bump, and pressure. They have efficient
processors, multiple sensors and huge memory, to exhibit intelligence. In
addition, they are capable of learning from their mistakes and they can adapt to
the new environment.
¨ Limitation of Artificial Intelligence
Currently AI systems are not perfect, and they suffer from a number of limitation. Open
ended world many times require unconstraint thinking and solution. AI systems are
constrained by a number of variables. The following table will better help understanding
the limitation of the current AI systems.
AI systems are capable of Ai system are not capable of
Recognising face in images. Understanding natural language.
Partially self-driving vehicles. Surfing the internet.
Simple machine translations. Interpreting visuals(images and videos)
Partially understanding human speech. Learning natural language
Partially predicting human behaviour

¨ Three domains of Artificial Intelligence

(1) Data
Data in artificial intelligent systems refers to the data that is right for the AI systems in
question. For example, a system developed for recognising the images of students will
not find the data of animal faces useful.
Data is at core of artificial intelligence. Current AI systems cannot developed or made
functional without adequate data. For example, an image recognition AI will need to be
trained on data containing image and the items in the images.

Types of Data:
Artificial intelligence uses different type of data for different purposes. For example
à Sound data for voice recognition systems,
à Text data for automatic classifying systems, and
à Image and video data for computer vision systems.

Page 6 of 25 Computer Department, DPS-Bopal

For Ai developers, finding the suitable data in the right format is challenging. Collecting
and structuring data is time consuming and difficult.

(2) Computer Vision (CV)

Computer vision is technology that is used for making machines see and perceive the
human world as human beings do. This requires image or visual data. The development
of computer vision is progressing at a fast pace. There is a great demand for application
that are capable for detecting and identifying objects from images or videos. Such
applications have multiple uses.

Uses for computer vision technology includes:

à Face recognition: It is used for applying filter and recognising images. Google photos,
Snapchat, Facebook, etc. are using this technology
à Image retrieval: It is used for identifying the images of objects. Search engines like
Google, Bing, etc. use this technology to provide result of image searches. There are
even algorithms that can provide result based on image used for searching.
à Gaming and control: Microsoft Kinect is one of the best example of successful
commercial product based on AI. Other example include games like Emoji Scavenger
Hunt.
à Surveillance: It includes identifying criminals, or illegal or suspicious behaviour, etc.
from video feed.
à Smart cars: Self-driving vehicles will need this technology for detecting traffic signs,
lights, other vehicles etc.

(3) Natural Language Processing (NLP)

The third domain of AI is Natural Language Processing. It works on enabling the
computers to understand naturally written or spoken language, like English or Hindi.
This will make computers capable of natural communication.

Components of NLP:
Natural Language processing contains two main components:
(1) Natural Language Understanding (NLU): This component deals with
understanding spoken or written language. It involves:
à establishing linkage with natural language input and what they represent.
à analysing different aspects of the language.
(2) Natural Language Generation (NLG): This component involves the process of
producing meaningful phrases and sentences in the form of natural language from some
internal from some internal representation. It involves:
à Text planning: Retrieving relevant text from the data stores.
à Sentence planning: Deciding on the correct words, linking them into meaningful
phrases, etc.
àText realisation: Combining phases and words for forming sentences
For AI systems NLU is difficult then NLG, human being find NLU easier than NLG

Page 7 of 25 Computer Department, DPS-Bopal

 Unit 2:: Relate

¨ Relevance Of AI in our daily life.

Technological advancement and breakthrough have altered the shapes of the
world in which we live. We no longer do things manually, as our forefather
used to do. We are blessed to be a part of a new generation in which a
significant part of the routine manual work is being taken over by computers
and machines. This has become possible only for a considerable advancement
in the field of artificial intelligence. It is slowly developing to a point where it
is able to perform various tasks, which needed human interventions in the
past.

Today, AI has become a part of our life. Both big and small companies are
using this technology to reduce work pressure and increase productivity. This
translates into better customer experience and comforts. AI has revolutionised
every industry and field.

Below listed are some of the industries where AI is being is used:

à Automobile industry
à Transportation and logistics corporations
à E-commerce organisations
à Financial and insurance companies
à Digital assistants
à Social networks
¨ How is Artificial Intelligence Relevant in Our Lives?

Page 8 of 25 Computer Department, DPS-Bopal

AI automates repetitive learning and discovering through data: This automation
distinct from the earlier hardware-based robotic automation that was limited. That AI
based automation is about taking care of work that is complex for us but simple for AI.
This automation learns and becomes better with time. For example, a simple task of
clicking photos- the AI technology in Google Pixel smart phones learns and improves it.

AI augments intelligence of existing product: Having separate AI products for final

consumers rarely happens. Normally, AI technology is used to improve the existing
products. For example, self-driving vehicles are not new products. Vehicles were
invented a long time ago. AI technology is only expanding their capability by making
them capable of self-driving

AI adapts through progressive learning algorithm to let the data do programming:

AI technologies are all about creating systems that can learn from the data fed to them. In
other words, these systems use data to program themselves. For example, take the case of
Amazon recommendation algorithm. This AI system studies customer’s buying habits
and recommends products that he/she will need and buy.

AI analyses more and deeper data using neural networks that have many hidden
layers:
These hidden layers provides better analysis and understanding. Take a simple example
of fraud detection by financial corporation. This multiple layers allows for a deeper
analysis and better predictions. In simple words, it means the more the data AI system
receives, the better it becomes.

AI achieves incredible accuracy: Take an simple examples Bing or Google search. AI

technology allows these search engine to provide better and more accurate results.

AI acquires the most out of data: For AI systems, data becomes prominent. In many
cases, it can also become intellectual property. The solutions are already there in the data.
Inferiorly designed AI system with better data can many times performs better than better
designed systems with poor data.

¨ Smart Living: Smart Homes to Smart cities

¨ Smart Homes to Smart Cities
As we have read in previous chapters, AI technology is changing the way we live our
life. One of the ways in which it is doing so, is by changing our work and home
environment. Many homes, airports, hospitals, universities, office, etc., are becoming
smarter, i.e., are using AI technology to make themselves better.

Some of the world cities are also turning smart. These cities are using AI technologies
to improve the lives of their citizens. These technologies will have an important role
to play in future as urbanisation increase and resource become more and more scarce.
For sustainable development, smart cities and smart homes provide an answer.
Page 9 of 25 Computer Department, DPS-Bopal
 A Smart Home has smart devices along with AI-powered technology to provide new
possibilities.
AI allows us to transform our homes into state-of-the-art machine. These machine can
be controlled from anywhere on Earth!
Smart homes can monitor children and take care of their safety, and inform parents if
something abnormal is detected. These homes can sense that the home owners are
leaving from office or schools and can switch on the AC, monitor milk and food
usage and order groceries, etc. the functionality and capability of smart homes are
only limited by our imagination!

Page 10 of 25 Computer Department, DPS-Bopal

 Smart cities: are the cities that use AI technologies along with the other digital and
information-communication technologies to make life of its inhabitants and more
comfortable.

¨ How can Smart Homes connect to Smart Cities?

Smart building do not exist in isolation. These building connect with each other and interact with smart
cities to provide a number of benefits.
Smart Buildings Smart Cities
Benefits for individuals Benefits for the community
Smart building provides benefits to the users of the Smart cities are evolving around the world.
buildings. The number of smart building is
increasing in the world.
The feature and capabilities of smart building are Initially smart city concept did not use AI
also increasing. technology and only provided simple community
services.
Smart building depend on smart appliances that In the second phase, smart cities evolved by
interact with each other and with the internet. combining sensors and analytical systems. For
example, traffic light sensors that change the
traffic lights depending on the traffic. This was the
phase that started using AI.
The benefits of smart buildings are limited to the The latest phase of the smart cities involves citizen
individual using the building. participation in decision-making.
Smart homes are evolving and most of the so-
called smart house are only partly smart, i.e., they
provide only limited specific functions.

Page 11 of 25 Computer Department, DPS-Bopal

 ¨ Data for Smart Homes
Smart homes and building are made up of a number of interconnected device. These
device gather a lot of data when they are running. This data helps smart homes to become
smarter. For example Google’s nest thermostat uses data from the internet about the
weather, time when you enter home, and uses it to set optimum temperatures for the
smart building heating and cooling device.

How will smart buildings data connect to smart cities?

In increasing number of smart building clubbed with smart city infrastructure will start
making cities increasingly smarter. Some of the different areas in which the smart cities
work are as follows:

Electricity: Smart building know their energy requirement during different times,
seasons, phases, etc. Smart building also see how they can save electricity and reduce
bills. With the growth in the number of smart buildings this data can be used by cities to
see how to reduce bills. With the growth in the number of smart building this data can be
used by cities to see how to reduce electricity consumption. This data can help power
companies to regulate power generation and supply.

Maintenance: This data from smart buildings can help manufactures know when a
product needs servicing and replacement. For companies, it can help in reducing the
warranty and after sale costs. For consumers, it leads to reduction in product failures.

Health: Hospitals can gather patient-related data from house without difficulty and use it
for better serving the elderly, disabled, and sick. For example, smart building can
continuously monitor vital signs of the patient and inform hospitals if any emergency
takes place. Doctor can also provide 24 x 7 monitoring to patient in their homes and
reduce the burden on health infrastructure.

Smart Citizen: Connecting individuals with community

Page 12 of 25 Computer Department, DPS-Bopal

 Unit 3:: Purpose

¨ Sustainable Development Goals (SDGs)

The term ‘Sustainable Development’ can mean different things to different people in
different situations. This makes the precise definition situations. This makes the precise
definition of the term difficult. One of the most frequently used definition of sustainable
development is:
‘Sustainable development is the development that meets the needs of the present without
compromising the ability of future generation to meet their own needs.’

This definition can be broken into the following components:

1. The development that satisfies the present needs.
2. The development that will allow future generation to meet their needs.

Sustainable Development Goals (SDGs) are set goals that the nations of the world
adopted at the United Nations Conference on Sustainable Development in Rio de Janeiro
in 2012. The SDGs were adopted to meet the urgent challenge of the world covering
following three dimensions:
1. Environmental
2. Social
3. Economical

Page 13 of 25 Computer Department, DPS-Bopal

 ¨ Appreciating the complexity of social issues
Social Issues may be interlinked and integrated across SDGs
Social issues are not simple. They are complex and interlinked. Isolating them or trying
to deal with one issue without understanding the other interlinked issue does not help.
For example, take the issue of ending poverty in all its form. Nations of the world have
accepted this as the greatest global challenge restricting sustainable development and as
such made it a Goal 1. But almost all the other SDGs are linked to it. For example
without addressing poverty we cannot address the issue hunger (goal 2) or good health
and well being (goal 3). Absence of Quality Education (goal 4) increase poverty and
poverty makes it difficult to attain quality education.

The inter-linked and integrated nature of the SDGs are of crucial importance. The
sustainable Development Agenda cannot be achieved without dealing with this inter-
linked
Page 14 of 25 Computer Department, DPS-Bopal
 ¨ Using AI to help achieve SDGs
One of the best example of AI, which is used to make Sustainable Development a reality,
is of Hurricane Harvey. In 2017, Houston was devasted by this hurricane. It led to a
situation where many roads in the city became flooded. Artificial intelligence used
satellite imagery data along with the object-detection techniques. This allowed disaster
management teams to better plan and coordinate the rescue attempts. Another example is
of a free Microsoft App for smart phones. A visually-challenged individual can use this
App to recognise friends and various objects.

Contribution to SDGs through evolution of technology

Development of artificial intelligence did not take place spontaneously, rather it evolved
over a period of time, something that can be best termed as waves of technological
advancements.
For example
à Sustainable Development Goal 6: ‘Smart water management’ systems are already
working to help in the attainment of this goal. In some areas, they have helped in
reducing water consumption by as much as 15 precent. These intelligent system use one
or more of the following component: smart pipes, soil sensors, smart irrigations
management, smart rain water harvesting, smart water meters, smart water flow control
software, etc.
à Sustainable Development Goal 7: ‘Smart energy’ solution are expected to results in
energy saving in excess of 1.3 billion MWh by 2030. These system use smart grids,
smart grids, smart appliances, smart energy storage, smart meters, sensors, smart demand
regulations etc.
à Sustainable Development Goal 11: Smart phones based disaster information sharing
systems and using disaster simulators to create hazard maps are helping governments
across the globe in both disaster prevention and disaster management.
à Sustainable Development Goal 12: AI technology like smart farm management,
smart integration precision agriculture etc. are helping in boosting farm productivity and
reducing wastage.
à Sustainable Development Goal 13: Different AI based solutions are helping to
reduce Green House gas emissions, and encourage adoption and use of non-pollution
renewable energy source.
Artificial Intelligence (AI) and robotics driven technologies are helping companies and
households across the world become more efficient and reduce consumptions. Adoption
of these technologies by all the different sections of global society will help in the
optimised use of resources that will in turn help steer the Earth toward sustainable
development.

Page 15 of 25 Computer Department, DPS-Bopal

AI technologies can help in achieving significant leaps towards sustainable development.
However the adoption of these technologies across the globe is not easy and the
following two factors substantially restrict their adoption:
(i) Problems with data availability and accessibility: Data is one of the three primary
domains of AI technologies, and this data is not freely available and accessible. Further,
the collection and retention of data raise privacy and ethical concern (the next chapter
will look at these concerns in details)
(ii) Chronic shortage of talent is able to improve the AI capabilities, improves
models, and implement solution: AI technologies are the latest development. This
means that there is shortage of manpower to develop these systems.

Page 16 of 25 Computer Department, DPS-Bopal

 Unit 4:: Possibilities

¨ Emerging of Artificial Intelligence (AI)

The development and adoption of Artificial Intelligence has changed our world. In the
previous chapter, we learnt about how AI is helping us in the future. However, before
that, let us first have a look at how AI has evolved.

Artificial Intelligence History

The term Artificial Intelligence or AI was first used in 1956, but it did not gain
popularity until much later. With time, the use of computer increased, the internet was
born and mobile internet picked pace. All of this led to an increase in data volumes, the
improvement in the computer’s computing power, increased data storage capabilities and
as a result the AI gained popularity.
The research in AI during the 1950s was restricted to problem solving and symbolic
methods. During 1960s, the pioneering work in the AI field was done by the US
Department of Defence that was working on creating computers capable of acting like
human beings. In 1970s, the US Defence Advanced Research Projects Agency (DAPRA)
completed the street mapping project. Incidentally, the first intelligent personal assistant
was produced in 2003, long before Siri, Alexa, or Cortona.
Human likes robots and artificial intelligence of science fiction are far away from reality.
The development of self-aware Also has not shown much progress and it will continue to
exist in the field of science fiction for some more time to come.

¨ Possibilities of AI in various fields

Everyday influences of AI globally
How AI has revolutionised and changed our lives. A person from 50 to 60 years ago will
be unable to imagine the world in which we are living today.

Artificial Intelligence Influences Business:

Commercial established and big business, along with the government, are the drivers for
artificial intelligence development. Adopting AI-based solution allows businesses to cut
both cost and time and provide better service and products to the customers. Leading
companies in every field are using AI solutions to stay ahead of the competition.

Life-Saving AI:
Healthcare industry is using AI to improve the quality of life and save it. The
development in this field include personalised drug treatment based on 24X7 monitoring
using various sensor. Robot assisted surgeries improve diagnosis. In the field of disaster
management, these technique are helping in both reducing the impact of disaster and
managing the post disaster rescue.

Page 17 of 25 Computer Department, DPS-Bopal

Entertaining AI:
Entertainment industry is experimenting with AI to produce original music, books,
recopies, etc. Companies like Netflix and Spotify use AI to predict what the users would
like to watch or listen.
Illustration- Smart Living visit to an AI-based Hospital in a Smart
City
The development is the field of AI will also change the manner in which we interact with
our environment. Modern Ai system use various sensor, cameras, speakers, microphones,
etc. for better interactions. These systems know what we are doing, what our preferences
are, and help make our life better. Let us look at how this happens by imagining a visit to
a hospital.

Setting 1: You walk into the reception area of the hospital. The hospital AI controller
recognises you based on your earlier visits to the hospital. It already knows the patient
you have come to see. It checks and sees if your name is still on the list of visitor who are
allowed to see the patient. You approach the reception and the AI notices that you are
breathing hard and you are empty handed. The reception robot asks you if you would like
to go directly to meet the patient, get something to drink at the canteen, or visit the gift
shops in the hospital. You tell the robot that you would like to visit the canteen, or visit
the gift shops in the hospital. You tell the robot that you would like to visit the canteen
and then go to the gift shop before visiting the patient.

Setting 2: The hospital AI updates your smart device with the directions to the canteen
and gift shop. You visit the canteen and pick up a soda from the beverages sections. The
payment for the soda is automatically deducted from your account. The cafeteria AI
system records different things like the weather, time your reason for visiting, etc. This
data will be used to make your future visits even more comfortable.

Setting 3: The gift shop AI talks to the Hospital AI and finds out that the patient you are
visiting is a five years old girl and that you arrived empty handed to the Hospital. The AI
also checks with the Hospital AI to find out the items that are appropriate for the patient
(for example due to allergies, medication, etc.) The robot at the gift shop shows you a
different range of items suited for a five years old girl.

Setting 4: You use the smart device to check the route to the patient’s room and head
towards it. On reaching the patient room you are informed that the visiting hours would
end at 6:00 PM. You are unable to access the latest charts of the patient because you are
not a close relative.

¨ Skills sets needed for jobs in AI field

The growth and development of Artificial Intelligent (AI) has increase the demand for
specific jobs in the industry. It has created a situation in which there is critical shortage
of trained individual in certain areas. Most of the artificial intelligence related jobs
require knowledge of computer and math. The entry level jobs require graduation.
Page 18 of 25 Computer Department, DPS-Bopal
Advanced jobs require post-graduation or doctorates. The field of study depends on the
category of jobs.

Type of AI career
AI careers can be of two types. The first relates to developing and deploying Ai systems.
These are highly specialised jobs that require advance training and education. The second
are related to operating AI system. For these kinds of jobs; the entry level requirements
are comparatively lower.
The career related to the core development and deployment of AI systems include:
à Software analysts
à Software developers
à Computer scientist
à Computer engineer
à Algorithm specialists
In addition, the development and deployment of AI systems need specialists from
different areas and fields. These careers help in the development of AI systems related to
their fields.
à Mechanical, manufacturing, electrical engineers
à Medical professionals
à Defence specialists
à Graphic design
à Architects
à Musicians and other entertainment specialists

Which companies are hiring for Artificial Intelligence Roles?

Companies that hire top AI talent from start-ups like Argo AI to tech giants like IBM.
Following are the leading employers who have hired top AI talent over the past years.

Page 19 of 25 Computer Department, DPS-Bopal

Skills Set in demand for Jobs in AI
Here are a few of the specific AI career along with the skills set require for them:

1. Machine Learning Engineer

Machine learning engineer is one of the most demanded AI related career. These
engineers develop and manage machine learning projects. The skills required for these
jobs include:
(a) understanding of different programming languages
(b) understanding of applied research and data science
(c) strong mathematical skills

2. Data Scientist
Data scientists are scientist who work with data. They collect, analyse, and draw
interpretation form huge datasets. This is a perfect job for individuals who like to play
with data. The skill requires for these jobs include:
(a) knowledge of big data platforms like Hive, Hadoop, MapReduce, etc.
(b) knowledge of statical computing and programming language like Perl and Python,
and
(c) strong analytical and communication skills
Data analysts normally have post-graduate or doctorate degree in field like engineering,
mathematics, or computer science.

3. Business Intelligence Developer

Business intelligence developers can be regarded as a subset of data scientist jobs. They
analyse data to identify business and market trends. This involves designing, analysing,
interpretation, maintenance, etc., of data sets.
The skills require for these jobs include:
(a) strong technical and analytical skills
(b) background of engineering, computer science or related fields and
(c) experience of business and markets

4. Research Scientist
Research scientist is related to research in the field of artificial intelligence. They
research and advance our knowledge of Ais. The skills required for these jobs include:
(a) master’s or doctored degree in computer science or related fields.
(b) strong background in applied mathematics

Page 20 of 25 Computer Department, DPS-Bopal

 Unit 5:: AI Ethics

¨ Ethical Concerns of AI related to Data (Bias and Inclusions

Concerns)
Artificial Intelligence is already being used to make everything decisions that have
serious implications on the lives of people. Companies are using AI systems to make
recruitment, promotion, and termination decision and evaluate loan applications. In
judiciary. AI systems are being used to help judge, and jury make better and more
informed judgements by predicting the chances of re-offenses by criminals. In campuses.
AI systems are being used to evaluate the potential of the incoming students for deciding
on their future careers. These uses of AI systems are raising serious ethical concerns
around the world. We already know that Data is one of the three core domains of AI. AI
systems have potential of analysing large amount of data and learning from this data;
something which human beings are incapable of doing.

Bias in real World Data

The problem here is that AI system learns from the real-world data fed into it. This
means that AI systems can reinforce the biases found in AI systems. For example, a
computer system trained on the data for last 200 years might find that more females were
involved in certain jobs or that more percentage of successful business were established
by men and conclude that specific gender are better equipped for handling certain jobs
(gender bias)

Understanding or even detecting such biases is not easy because many AI systems act as
black boxes. The reason behind their decision-making is not easy or in some cases, even
possible to understand. Many times, programmers of AI systems themselves cannot
explain the logic behind decision taken by the AI systems.

Two famous Cases of AI Bias

Amazon Recruitment ProPublica Judiciary services
In 2014, scientist at Amazon developed a In 2016, the scientist at ProPublica
recruitment AI to select the software developed a system to predict the chances
engineer that the company might consider of repeated offence by criminals. The idea
for hiring. was to help judges make better judgement
Amazon quickly found out that this system with this additional information.
had picked gender bias from the data that The system picked up racial bias existed in
was used for making the selection and as a America for hundreds of years and became
result was discriminating against women. biased towards the black community.
Amazon was forced to abandon the AI
recruitment system in 2017.

Page 21 of 25 Computer Department, DPS-Bopal

Since the beginning of computer programming, the programmers understood the classic
computer maxim; Garbage in Garbage out, i.e., the output provided by computers
depends on the input given to it . Feed useless or irrelevant inputs and the outputs will
also be useless and irrelevant.

This also holds true for AI systems. But in case of Ais, this problem is further
complicated because AI systems learn from the data fed into them. This real world data is
rarely free form bias! AI systems trained on biased data also learn the bias from the data.

Problem in Inclusion
AI system trained on biased real world data create problem of inclusion, i.e., the problem
that some people are left out of AI decision-making system. Consider the example of AI
system used by Amazon for recruitment. This created a situation in which many eligible
females were left out of consideration. This is known as the problem of inclusion.

Problems of facts and their interpretation:

The root cores of the bias in AI systems come from the problem between facts and
interpretation of facts. AI systems can scan data and draw learning from them but AI
systems are not equipped to understand the reason behind particular conclusion or
learning.
Microsoft developed and launched an experimental twitter chatbot as ‘Tay’. Tay was
designed to experiment with conversation understanding. While launching Tay.
Microsoft claimed that, the more you chat with Tay, the smarter it will get, learning to
engage people through causal and playful conversations. In less than 23 hours of launch,
Tay learned from thousands of tweets and developed a racial misogynistic personality.

Page 22 of 25 Computer Department, DPS-Bopal

There was no programming error or hacking for the consequences. The problem was that
Tay did exactly what it was decided to do. It started mimicking the users who were
talking to it. It learned from what was being fed to it without interpreting the inputs.
¨ Ethical Concerns related to AI Access
The second sets of AI ethical problems is related to the implication of the AI systems.
These AI ethical concerns can be divided into two parts:

Privacy Concerns of AI systems

One of the most serious implication of data collection is related to privacy concerns.
Take a simple example of your Android Smart Phone. The following is a list of some
information that Google possesses about you (if you have given permission).
1. Contact List: All of the numbers on your phone along with the name. google knows
who are your favourite (based on the amount of time you spend talking to them), the best
time to call you, the people who you do not like to talk to etc.

2 Location: Google knows when you leave for school, the route which you take, the time
you spend in school, the route you take back, the shops that you like to visit, the park
where you go for walk, the restaurant where you eat. Etc.

3. Email: The complete list of all the mails sent and received by you (including the ones
you deleted) Every personal and official mail.

4. Hangout: All of your chat records.

5. Photos: All the images of you and your fiends along with their names. Linked with
contact list google knows their phone numbers, email-IDs, addresses etc. Linked with
calendar it also knows about important events like their birthdays, marriage
anniversaries, birth of their children, etc.
Most of time we accept lengthy user agreement without even realising the implication of
these agreement on our privacy rights. Data related to us is stored somewhere in some
large database. This data can be used for a number of purposes.
The information is collected in lieu of providing us with helpful services, but the
potential risk here is very high. For example, take facial recognition system. The AI
system can use this technology along with the information provided by Apps like Contact
and Photos to create a wide surveillance network that can monitor you 24 x 7

Components of a Good AI policy

A good Ai policy must have five components:
à The AI system must be transparent.
à An AI must have the right to the information it is collecting.
à Consumer must have the right to the information it is collecting.
à Consumers must have a choice of leaving the system.
à The data collected and the purpose of the AI must be limited by design.
à Data must be deleted upon consumer request.
Page 23 of 25 Computer Department, DPS-Bopal
Ethical concerns related to adoption of AI systems
One of the major concerns related to the adoption of AI systems is:

Job Loss: For a long time now people have been arguing about the number of the jobs
that will be lost due to adoption of AI systems and advanced robots. There have been
research published by different organization that measure this impact differently.
In previous units we saw the potential benefits of AI technology. The ethical question
before us is evaluate these benefits against human cost and see if is worth it or not.
Another way of looking at it is to find how to adapt to the changing future and look for
new methods of employment generation. We must not forget that most of the
technological advancements have led to job losses.

Increasing Inequalities: The current economic system provides economic rewards based
on contribution to the economy. The AI era will reduce the number of people required for
doing jobs. This will also reduce economical remuneration given to them. In this case,
most economic benefits will go to companies who own AI systems. This will further
widen the already wide income gaps.

Negative Adoptions: Ai technologies, like all the other technologies, can be used for both
good and evil. This technology, in the hands of terrorists can be used to spread terror.
They can use AI based technologies to make attacks that can do most damage in terms of
life and infrastructure losses. Cyber criminals can use AI technology to hack into more
systems and for doing more damage. Countries can use this technology to increase war
damage inflicted by them.

Black Box Problem: Black Box problem refers to inherent problems of AI systems.
Sometimes what happens in the AI systems is normally not easily understood by a
common individual and not even by programmer. This problem stems out of the fact that
AI systems do not provide reasoning of decision made by them.

Cost benefit Analysis of the AI system

In Unit 2 and 3, we learnt about the different benefits of the AI technology and the
difference that they are making in the world. In Unit 4, we have looked at the
possibilities of the AI systems and in the current Unit we looked at the ethical concerns
or dilemmas related to the AI systems.
Every technology development can be used for the benefits of the mankind as well as for
destruction. We can use dynamics to make mining easy and we can also use them as
weapons. Nuclear power can be used to create energy and it can also be used to destroy
cities and countries. Advancement in medicines can be used to heal and improve the
quality of life human beings as well as create biological weapons.
AI technology is no different. It can be used for the good of the mankind or for its
destruction.

Page 24 of 25 Computer Department, DPS-Bopal

Overall, the society will have to come up with ethical guidelines on data and AI systems
as it has done for many other technologies. The violation of these guidelines, rules or
regulations will have to be punished appropriately

Reference Taken from:

https://www.tutorialspoint.com/artificial_intelligence/

Page 25 of 25 Computer Department, DPS-Bopal