CSE FY 2023-24 SEM II

Dr. Vasantraodada Patil Shetkari Shiksha mandal’s

Padmabhooshan Vasantraodada Patil Institute of Technology
Budhagaon

Department of Computer Science Engineering

SEMINAR REPORT ON :-

“Artificial Intelligence”

Under the guidance of

Mrs. Smita .P. Mandale

Submitted by:-

Roll no Student name

1231 Akash Honrao

Academic year:-2023-24

1
CSE FY 2023-24 SEM II

Padmabhooshan Vasantraodada Patil Institute of

Technology Budhgaon

Department of Computer Science Engineering

CERTIFICATE

This is to certify that have satisfactorily submitted a report on

Artificial Intelligence

Submitted by :-

Name Akash Honrao

Roll no 1231
PRN no 23062691242104

Incharge H.O.D

Mrs.S.P.Mandale Prof.Dr.A.A.Patil

2
CSE FY 2023-24 SEM II

ACKNOWLEDGEMENT
We express our sincere gratitude to Mrs.S.P.Mandale. , Assistant Professor,
Department of First year. Engineering, PVP Institute of Technology Budhagaon , as the
mentor for my seminar. It is our great fortune that we have got opportunity to carry out
this seminar work under the supervision. We express our sincere thanks for the
encouragement, support and the guidance. We would further like to thank all the faculty
member for their cooperation and extended support in undergoing the project work. We
also would thank the department of F i r s t Y e a r Engineering to provide the
infrastructure and facility in carrying the seminars like these.

Mr.Akash Honrao

3
CSE FY 2023-24 SEM II

INDEX
Sr. Content Page
no. no.
1. Abstract 5
2. Introduction 6
3. History 7 to 8
4. Different fields under AI to clear common 9 to 11
misconceptions
5. How AI stands out in different industries. 12
6. The current state of AI-based software systems. 13
7. The Future of AI. What to expect from AI in the 14
next few years or decades
8. Concerns surrounding the advancement and usage 15
of AI
9. Conclusion 16

4
CSE FY 2023-24 SEM II

Abstract:
Artificial Intelligence (A.I.) is a multidisciplinary field whose goal is to automate activities
that presently require human intelligence. Recent successes in A.I. include computerized
medical diagnosticians and systems that automatically customize hardware to particular user
requirements. The major problem areas addressed in A.I. can be summarized as Perception,
Manipulation, Reasoning, Communication, and Learning. Perception is concerned with
building models of the physical world from sensory input (visual, audio, etc.). Manipulation
is concerned with articulating appendages (e.g., mechanical arms, locomotion devices) in
order to effect a desired state in the physical world. Reasoning is concerned with higher level
cognitive functions such as planning, drawing inferential conclusions from a world model,
diagnosing, designing, etc. Communication treats the problem understanding and conveying
information through the use of language. Finally, Learning treats the problem of
automatically improving system performance over time based on the system's experience.
Many important technical concepts have arisen from A.I. that unify these diverse problem
areas and that form the foundation of the scientific discipline. Generally, A.I. systems
function based on a Knowledge Base of facts and rules that characterize the system's domain
of proficiency. The elements of a Knowledge Base consist of independently valid (or at least
plausible) chunks of information. The system must automatically organize and utilize this
information to solve the specific problems that it encounters. This organization process can
be generally characterized as a Search directed toward specific goals. The search is made
complex because of the need to determine the relevance of information and because of the
frequent occurence of uncertain and ambiguous data. Heuristics provide the A.I. system with
a mechanism for focusing its attention and controlling its searching processes.

5
CSE FY 2023-24 SEM II

Introduction to artificial intelligence

Artificial intelligence applications are all around us, but what does it really mean? In this
article, Kumar Abhishek explains the history and progress of artificial intelligence.

Artificial intelligence (AI) is the ability of machines to replicate or enhance human intellect,
such as reasoning and learning from experience. Artificial intelligence has been used in
computer programs for years, but it is now applied to many other products and services. For
example, some digital cameras can determine what objects are present in an image using
artificial intelligence software. In addition, experts predict many more innovative uses for
artificial intelligence in the future, including smart electric grids.

AI uses techniques from probability theory, economics, and algorithm design to solve
practical problems. In addition, the AI field draws upon computer science, mathematics,
psychology, and linguistics. Computer science provides tools for designing and building
algorithms, while mathematics offers tools for modeling and solving the resulting
optimization problems.

Although the concept of AI has been around since the 19th century, when Alan Turing first
proposed an “imitation game” to assess machine intelligence, it only became feasible to
achieve in recent decades due to the increased availability of computing power and data to
train AI systems.

To understand the idea behind AI, you should think about what distinguishes human
intelligence from that of other creatures – our ability to learn from experiences and apply
these lessons to new situations. We can do this because of our advanced brainpower; we have
more neurons than any animal species.

Today’s computers don’t match the human biological neural network – not even close. But
they have one significant advantage over us: their ability to analyze vast amounts of data and
experiences much faster than humans could ever hope.

AI lets you focus on the most critical tasks and make better decisions based on acquired data
related to a use case. It can be used for complex tasks, such as predicting maintenance
requirements, detecting credit card fraud, and finding the best route for a delivery truck. In
other words, AI can automate many business processes leaving you to concentrate on your
core business.

Research in the field is concerned with producing machines to automate tasks requiring
intelligent behavior. Examples include control, planning and scheduling, the ability to answer
diagnostic and consumer questions, handwriting, natural language processing and perception,
speech recognition, and the ability to move and manipulate objects.

6
CSE FY 2023-24 SEM II

History of AI and how it has progressed over the years

With so much attention on modern artificial intelligence, it is easy to forget that the field is
not brand new. AI has had a number of different periods, distinguished by whether the focus
was on proving logical theorems or trying to mimic human thought via neurology.

Artificial intelligence dates back to the late 1940s when computer pioneers like Alan Turing
and John von Neumann first started examining how machines could “think.” However, a
significant milestone in AI occurred in 1956 when researchers proved that a machine could
solve any problem if it were allowed to use an unlimited amount of memory. The result was a
program called the General Problem Solver (GPS).

Over the next two decades, research efforts focused on applying artificial intelligence to real-
world problems. This development led to expert systems, which allow machines to learn from
experience and make predictions based on gathered data. Expert systems aren’t as complex as
human brains, but they can be trained to identify patterns and make decisions based on that
data. They’re commonly used in medicine and manufacturing today.

A second major milestone came in 1965 with the development of programs like Shakey the
robot and ELIZA, which automated simple conversations between humans and machines.
These early programs paved the way for more advanced speech recognition technology,
eventually leading to Siri and Alexa.

The initial surge of excitement around artificial intelligence lasted about ten years. It led to
significant advances in programming language design, theorem proving, and robotics. But it
also provoked a backlash against over-hyped claims that had been made for the field, and
funding was cut back sharply around 1974.

After a decade without much progress, interest revived in the late 1980s. This revival was
primarily driven by reports that machines were becoming better than humans at “narrow”
tasks like playing checkers or chess and advances in computer vision and speech recognition.
This time, the emphasis was on building systems that could understand and learn from real-
world data with less human intervention.

These developments continued slowly until 1992, when interest began to increase again.
First, technological advances in computing power and information storage helped boost
interest in research on artificial intelligence. Then, in the mid-1990s, another major boom was
driven by considerable advances in computer hardware that had taken place since the early
1980s. The result has been dramatic improvements in performance on several significant
benchmark problems, such as image recognition, where machines are now almost as good as
humans at some tasks.

7
CSE FY 2023-24 SEM II

The early years of the 21st century were a period of significant progress in artificial
intelligence. The first major advance was the development of the self-learning neural
network. By 2001, its performance had already surpassed human beings in many specific
areas, such as object classification and machine translation. Over the next few years,
researchers improved its performance across a range of tasks, thanks to improvements in the
underlying technologies.

The second significant advancement in this period was the development of generative model-
based reinforcement learning algorithms. Generative models can generate novel examples
from a given class, which helps learn complex behaviors from very little data. For example,
they can be used to learn how to control a car from only 20 minutes of driving experience.

In addition to these two advances, there have been several other significant developments in
AI over the past decade. There has been an increasing emphasis on using deep neural
networks for computer vision tasks, such as object recognition and scene understanding.
There has also been an increased focus on using machine learning tools for natural language
processing tasks such as information extraction and question answering. Finally, there has
been a growing interest in using these same tools for speech recognition tasks like automatic
speech recognition (ASR) and speaker identification (SID).

8
CSE FY 2023-24 SEM II

Different fields under AI to clear common misconceptions

Artificial Intelligence is the most trending field of computer science. However, with all the
new technology and research, it’s growing so fast that it can be confusing to understand what
is what. Furthermore, there are many different fields within AI, each one having its specific
algorithms. Therefore, it’s essential to know that AI is not a single field but a combination of
various fields.

Artificial Intelligence (AI) is the general term for being able to make computers do things
that require intelligence if done by humans. AI can be broken down into two major fields,
Machine Learning (ML) and Neural Networks (NN). Both are subfields under Artificial
Intelligence, and each one has its methods and algorithms to help solve problems.

Machine learning

Machine Learning (ML) makes computers learn from data and experience to improve their
performance on some tasks or decision-making processes. ML uses statistics and probability
theory for this purpose. Machine learning uses algorithms to parse data, learn from it, and
make determinations without explicit programming.

9
CSE FY 2023-24 SEM II

datasets. Machine learning algorithms are designed to strive to establish linear and non-linear
relationships in a given set of data. This feat is achieved by statistical methods used to train
the algorithm to classify or predict from a dataset.

Deep learning

Deep learning is a subset of machine learning that uses multi-layered artificial neural
networks to deliver state-of-the-art accuracy in object detection, speech recognition and
language translation. Deep learning is a crucial technology behind driverless cars and enables
the machine analysis of large amounts of complex data — for example, recognizing the faces
of people who appear in an image or video.

Neural networks

Neural networks are inspired by biological neurons in the human brain and are composed of
layers of connected nodes called “neurons” that contain mathematical functions to process
incoming data and predict an output value. Artificial neural network learns by example,
similarly to how humans learn from our parents, teachers, and peers. They consist of at least
three layers: an input layer, hidden layers, and an output layer. Each layer contains nodes
(also known as neurons) which have weighted inputs that compute the output.

The performance of traditional machine learning models plateau and throwing any more data
doesn’t help improve the performance. Deep learning models continue to improve in
performance with more data.

10
CSE FY 2023-24 SEM II

These fields have different algorithms, depending on the use case. For example, we have
decision trees, random forests, boosting, support vector machines (SVM), k-nearest neighbors
(kNN), and others for machine learning. For neural networks, we have convolutional neural
networks (CNNs), recurrent neural networks (RNNs), long short-term memory networks
(LSTMs), and more.

However, classifying AI according to its strength and capabilities would mean further
subdividing it into “narrow AI” and “general AI.” Narrow AI is about getting machines to do
one task really well, like image recognition or playing chess. General AI means devices that
can do everything humans can do and more. Today’s research focuses on narrow AI, but
many researchers would like machine learning to eventually achieve general AI.

11
CSE FY 2023-24 SEM II

How AI stands out in different industries.

AI is a booming technology that the global community has accepted. It has been
revolutionizing the industry from various sectors for quite some time. It is a comprehensive
technology that is being applied in almost every industry. This section discusses how AI is
impacting service delivery in various sectors.

Fully self-driving cars are now a reality. Tesla is the first company to make a car with all of
the sensors, cameras, and software needed for a computer to drive itself from start to finish.
Trucks may be the next primary target for autonomy: self-driving trucks will enormously
impact road safety and infrastructure and save companies money by reducing labor costs.

A few other industries are also implementing AI. For example, in finance, AI helps with
forecasting and supports hedge-fund investment decisions. Predictive analytics (or
forecasting) applies artificial intelligence using machine learning and statistical techniques to
make predictions about future events based on previous data. For example, you can use
forecasting to predict product sales, customer demand, or even stock prices. One popular
example of predictive analytics is Amazon’s product recommendations engine (also known
as “Customers who bought this item also bought”). It uses past purchase data from millions
of customers to recommend products based on the users’ preferences.

In healthcare, AI is helping doctors to diagnose diseases by gathering data from health

records, scanning reports, and medical images. This helps doctors to make faster diagnoses
and guide the patient for further tests or prescribe medications. In addition, AI can be used in
the treatment process by monitoring patients and alerting their doctors when something goes
wrong. According to Forbes, AI will save over 7 million lives in 2035.

In retail, AI does everything from stock management to customer service chatbots. As a

result, many businesses are taking advantage of AI to improve productivity, efficiency, and
accuracy. In addition, companies find new ways to use AI to make life easier for their
customers and employees, from product design to customer service.

12
CSE FY 2023-24 SEM II

The current state of AI-based software systems.

The recent advancements in AI have led to the emergence of a new type of system called
Generative Adversarial Networks (GANs), which generate realistic images, text, or audio.
Due to their remarkable capabilities, some people are concerned that this technology could
replace humans in the future.

GANs are just one example of how AI is changing our lives. This section explores more
current AI examples and its applications in software systems such as GPT3, DALL.E, and
virtual reality/augmented reality (VR/AR).

AI-based software systems are comprised of many layers such as foundational models,
advanced algorithms, and automated reasoning tools. Some of the most popular AI-based
systems that use these layers include GPT3, DALL.E, AlphaGo, RoBERTa, and many others.

DALL.E and GPT3 are large-scale models that have achieved remarkable results in computer
vision and natural language processing (NLP).

The GPT3 model is an NLP model based on a deep learning algorithm called transformers. It
was trained on a corpus of text from Common Crawl and published in 2020. GPT3 uses a
large dataset trained in the English language to produce outputs based on the inputted
information. The model can be trained to perform any task imaginable, from generating text
to solving math problems. Also, we can use GPT3 to generate text, translate between
languages, answer questions about images, and more.

The DALL.E model is an image generator based on a deep learning algorithm called
variational autoencoders (VAEs). Similarly, DALL.E can be trained using an image dataset
to produce images based on the inputted text descriptions. It was trained on datasets such as
ImageNet and published in 2021. We can use DALL.E to generate images that match
captions or URLs given by users. These models have been developed by OpenAI, which has
close ties to the US government and military-industrial complex (MIC).

DeepMind created AlphaGo as a program that would play the ancient game Go without
anyone’s help. The game is similar to chess but much more complex due to its simple rules
and many possible moves per turn. AlphaGo used reinforcement learning to learn how to play
the game better over time by playing against itself repeatedly until it mastered every possible
situation that could occur in a game of Go with 100% accuracy.

RoBERTa is an algorithm from Facebook AI Research (FAIR) that uses deep learning
techniques to solve problems in natural language processing (NLP), such as sentence
classification or machine translation.

13
CSE FY 2023-24 SEM II

The Future of AI. What to expect from AI in the next few years or
decades
Artificial intelligence has come a long way, but it’s about to make a huge leap. Artificial
general intelligence (AGI), the kind of AI capable of doing any intellectual task that a human
being can do, is still a ways off, but we’re already starting to see plenty of progress in other
areas of AI. Here’s what you can expect soon:

Artificial Intelligence will make more jobs obsolete as it takes over more and more tasks

The reason why is simple: if you can replace one person with an AGI system, you don’t need
one computer to do the work – you can spread it out across thousands or millions of
computers. That’s only possible because a general AI system can learn from past experiences
and improve itself, meaning that it doesn’t have to be reprogrammed for every new task. In
fact, there’s no reason why an AGI system would need humans at all – once it learns enough,
it could design its own machines or find ways to automate entire industries.

The advent of AI is transforming the business landscape and changing people’s lives for the
better. In the coming years, most industries will see a significant transformation due to new-
age technologies like cloud computing, Internet of Things (IoT), and Big Data Analytics. All
these factors profoundly influence how businesses operate today and are also finding
applications in other areas like military, healthcare, and infrastructure development.

To build an engaging metaverse that appeals to millions of users who want to learn, create,
and inhabit virtual worlds, AI must be used to enable realistic simulations of the real world.
People need to feel immersed in the environments they participate in. AI is helping to achieve
this reality by making objects look more realistic and enabling computer vision so users can
interact with simulated objects using their body movements.

14
CSE FY 2023-24 SEM II

Concerns surrounding the advancement and usage of AI

AI is a very powerful idea, but it’s not magic. The key thing to remember about AI is that it
learns from data. The model and algorithm underneath are only as good as the data put into
them. This means that data availability, bias, improper labeling, and privacy issues can all
significantly impact the performance of an AI model.

Data availability and quality are critical for training an AI system. Some of the biggest
concerns surrounding AI today relate to potentially biased datasets that may produce
unsatisfactory results or exacerbate gender/racial biases within AI systems. When we
research different types of machine learning models, we find that certain models are more
susceptible to bias than others. For example, when using deep learning models (e.g., neural
networks), the training process can introduce bias into the model if a biased dataset is used
during training.

However, other machine learning models (e.g., random forests) can be less sensitive to the
bias in the data during training. For example, if a dataset contains information about many
different variables but only one variable is used to make decisions (e.g., gender), this model
will tend to be more biased toward that variable than random forests that consider all
variables equally weighted by default.

Other concerns need to be taken into account with the advancement and usage of AI. These
include data availability, computational power, and privacy, such as health data. People’s
data is needed to develop models, but how do we get such data given how protected health
data needs to be.

As artificial intelligence becomes more common, it’s only natural that there are increasing
requirements for processing power. As a result, AI researchers use supercomputers to develop
algorithms and models on a massive and complex scale.

This is especially true of deep learning, a type of machine learning that uses algorithms to
recognize patterns in large data sets like images or sound. The main issue with DL is that it
requires enormous computational power. To train a neural network using DL, you need to
feed vast amounts of data into the system — for example, thousands or millions of pictures
— and then let it figure out how to tell one from another on its own. This training process is
complex and laborious, but it is also computationally expensive. Model development can take
days or even weeks on a single high-end GPU or CPU capable of delivering lots of
computational power. To make matters worse, once you train the model, you need a
supercomputer to execute the model at full capacity. Google’s investments in TPUs (Tensor
Processing Units) attempt to solve this problem using state-of-the-art hardware technology.

15
CSE FY 2023-24 SEM II

Conclusion
Artificial intelligence (AI) is the intelligence of machines and the branch of computer science
that aims to create it. AI is today’s dominant technology and will continue to be a significant
factor in various industries for years to come. As AI systems become more advanced, they
are not only poised to disrupt multiple industries with their impact but also raise concerns
about how we should handle such incredible power.

This field has evolved a great deal over the years. It has gone from being a subject of popular
science fiction to a significant part of our lives today. By examining AI from its past, it is
possible to better understand its present and predict its future, as we have done in this article.

16
CSE FY 2023-24 SEM II

Reference

1)https://www.red-gate.com/simple-talk/development/data-science-
development/introduction-to-artificial-intelligence/

2) https://www.simplilearn.com/tutorials/artificial-intelligence-
tutorial/what-is-artificial-intelligence#:~:text=Artificial
%20Intelligence%20is%20a%20method,develops%20intelligent
%20software%20and%20systems.

3) https://www.techtarget.com/searchenterpriseai/definition/AI-
Artificial-Intelligence

17