AI and CLOUD COMPUTING yana

WHAT IS CLOUD COMPUTING? Deployment Models

★​ Cloud computing is the delivery of

IaaS PaaS SaaS
different services through the Internet.
★​ These resources include tools and rented on a software can subscribe to
pay-as be developed cloud-hosted
applications like data storage, servers, you-go basis and deployed; software; the
databases, networking, and software. only the provider
application manages
and data are everything
managed by except
Advantages the client. service usage.

★​ Cost - Eliminates expenses for hardware,

IaaS PaaS SaaS
software, and on-site data centers.
Applications Applications Applications
★​ Speed - Provides on-demand resources
Data Data Data
quickly with flexibility.
★​ Security - Offers robust policies and tools Runtime Runtime Runtime

to protect data and infrastructure. Middleware Middleware Middleware

★​ Global Scale - Scales resources as needed,
O/S O/S O/S
based on demand and location.
Virtualization Virtualization Virtualization
★​ Reliability - Simplifies data backup,
Servers Servers Servers
disaster recovery, and continuity with
redundancy. Storage Storage Storage

★​ Productivity - Reduces manual IT tasks, Networking Networking Networking

freeing teams for strategic goals.
★​ Performance - Operates on upgraded, ★​ managed by you
efficient hardware for better speed and ★​ managed by vendor
reduced latency.

1
 AI and CLOUD COMPUTING yana

INTRO TO AI ➔​ 1955: Allen Newell and Herbert A. Simon

created "Logic Theorist". It performed

★​ AI refers to the simulation of human automated reasoning and was said to be

intelligence in machines that are the "first artificial intelligence program“.

programmed to think like humans and ◆​ This program proved 38 of 52

mimic their actions. Mathematics theorems, and found

○​ endowed with the intellectual new and more elegant proofs for

processes characteristic of some theorems.

humans, such as the ability to ➔​ 1956: The word "Artificial Intelligence"

reason, discover meaning, was coined by American Computer

generalize, or learn from past scientist John McCarthy at the Dartmouth

experience. Conference
◆​ John McCarthy, widely recognized
as one of the godfathers of AI,
History of AI
defined AI as “the science and

engineering of making
➔​ 1943: The first work which is now
intelligent machines.”.
recognized as AI was done by Warren
➔​ 1966: The researchers emphasized
McCulloch and Walter Pits in 1943. They
developing algorithms which can solve
proposed a model of artificial neurons.
mathematical problems.
➔​ 1950: Alan Turing pioneered Machine
◆​ Joseph Weizenbaum created the
learning in 1950. Alan Turing proposed a
first chatbot in 1966, which was
test. The test can check the machine's
named ELIZA.
ability to exhibit intelligent behavior
➔​ 1972: The first intelligent humanoid robot
equivalent to human intelligence, called a
was built in Japan which was named as
Turing test.
WABOT-1.

2
 AI and CLOUD COMPUTING yana

➔​ 1974 - 1980 was the first AI winter. ➔​ 2002: For the first time, AI entered the
◆​ AI winter refers to the time period home in the form of Roomba, a vacuum
where computer scientists dealt cleaner.
with a severe shortage of funding ➔​ 2006: AI came into the Business world till
from the government for AI the year 2006. Companies like Facebook,
research. Twitter (now X), and Netflix also started
◆​ Expert system was introduced. using AI.
●​ Expert systems – mimic ➔​ 2011: In the year 2011, IBM's Watson won
decision making ; computer jeopardy, a quiz show, where it had to
systems emulating the solve complex questions as well as riddles.
decision-making ability of a ➔​ 2014: In the year 2014, Chatbot "Eugene
human expert. Goostman" won a competition in the

●​ XCON (eXpert CONfigurer) "Turing test."

- production-rule-based ◆​ Eugene Goostman - A computer

system used to assist in the program that pretends to be a

ordering of DEC's VAX 13-year-old Ukrainian boy who

computer systems by passed a Turing test.

automatically selecting the ➔​ 2018: The "Project Debater" from IBM

computer system debated on complex topics with the

components based on the master debater and also performed

customer's requirements extremely well.

(1978). ➔​ 2018: Google has demonstrated an AI

➔​ 1987 - 1994 was the second AI winter program "Duplex" which is a virtual

duration. assistant.

◆​ Again, investors and government ◆​ Duplex is a new technology for

stopped funding for AI research conducting natural conversations

due to high cost but not efficient to carry out “real world” tasks over

results. the phone.

➔​ 1997: IBM’s Deep Blue beats Gary

Kasparov.

3
 AI and CLOUD COMPUTING yana

➔​ ChatGPT, an acronym which stands for ★​ Education: AI tools assist educators and
"generative pretrained transformer” is a students with platforms like Grammarly,
conversational robot with which humans ChatGPT, and Adobe Firefly for learning
will be able to converse in a natural and productivity.
language. ○​ Grammarly, CHATGPT, Curipod,
Magic School, coursehero,
Adobefirefly, Fetchy, Cognii,
Applications
Gemini
○​ Numerous AI lets you create
★​ Healthcare: AI aids in faster, more
ChatGPT prompts in Microsoft
accurate diagnoses, analyzes clinical and
Excel and Google Sheets.
genomic data, reduces human error,
★​ Gaming: AI plays strategic games (e.g.,
enhances medical tests, and supports
chess) using advanced decision-making
wearable tech for monitoring and risk
capabilities.
prediction.
○​ Deep Blue, IBM Watson, Alphago
○​ Best known tech: IBM’s Watson.
★​ Social Media: AI organizes data, identifies
★​ Finance: AI detects fraud, improves
trends, analyzes user behavior, and
trading processes, estimates security
enhances content personalization and
prices, and automates tasks like chatbots
advertising.
and adaptive intelligence.
★​ Entertainment: AI personalized user
★​ Business: Robotic automation handles
experiences on platforms like Netflix, aids
repetitive tasks, AI improves CRM and
marketing, and optimizes searches in
customer service, and job automation is a
media.
growing focus.
★​ Travel: AI streamlines bookings, suggests
★​ Automotive: AI powers virtual assistants,
travel options, and enhances customer
self-driving cars, and advanced tools like
interaction with chatbots.
LiDAR for safety and performance.
○​ TeslaBot, AI assistant for
automotive tech.

4
 AI and CLOUD COMPUTING yana

Machine Learning Machine Learning Algorithms

➔​ Machine learning is a subset of artificial ★​ Supervised learning - The algorithm

intelligence (AI) that provides systems the learns on a labeled dataset, to train
ability to automatically learn and improve algorithms to classify data or predict
from experience without being explicitly outcomes accurately.
programmed. ○​ spam detection, image recognition
➔​ Machine learning algorithms build a ★​ Unsupervised learning- Uses unlabeled
mathematical model based on sample data. From that data, it discovers patterns
data, known as "training data", in order to that help solve clustering or association
make predictions or decisions without problems.
being explicitly programmed to do so. ○​ Clustering, customer
◆​ Training data is the data you use segmentation
to train an algorithm or machine ★​ Deep learning, or deep neural learning -
learning model to predict the a subset of machine learning, which uses
outcome you design your model to the neural networks to analyze different
predict. factors with a structure that is similar to
the human neural system.
★​ AI
○​ ML ○​ DL requires a huge amount of data
■​ DL to train the deep neural networks
●​ Gen. AI
from scratch.
○​ DL is leveraged for complex data
structures especially images
○​ Convolutional Neural Networks
(CNN) are specialized for image
processing

5
AI and CLOUD COMPUTING yana

○​ Generative AI is an artificial
intelligence (AI) that can produce
content such as audio, text, code,
video, images, and other data.
■​ Generative AI is a type of
artificial intelligence that
uses large language
models (LLMs)
■​ Generative AI applications
like ChatGPT, Microsoft
Copilot, and DallE can
produce human-like
responses and generate
original content.
■​ Engineers develop this
technology using several
approaches, including
GANs,
●​ Generative
Adversarial
Network - which
include components
for creating content
and another
component for
evaluating that
content for
authenticity