Introduction to Artificial

Intelligence
Dr. K. Ganesan
Professor Higher Academic Grade,
School of Computer Science
Engineering and Information Systems,
VIT University, Vellore – 632 014
[email protected]
Ph : 6382203768
• Why one should learn AI?
• It is Disruptive for Software Development and many
low handing fruit kind of jobs
– Web Site Design (E-commerce sites) - Figma
– Software Testing (unit, Integration)
– Mobile Application (for all platforms) - ChatGPT
– Auto Code Generation (Programmer) - CodeGeeX
– Debugging (Error handling) – Github Copilot
– Documentation (Manual preparation) - Codeguim
– UI / UX design (front end)
– Logo creation (Branding) - BrandCrowd
– Audio and Video Editing – CreatorKit
– Remove unwanted objects in video - Runway
– Social Media Posting (Marketing) – Gravity Write
– Content Creation (Advertising)
– Maintenance ( Providing solutions – chatbot)
– Reception (FAQ)
– Animation Creation (Education) – Neural Frames
– Image Editing (Photographer) – Adobe Firefly
– Summarize You Tube Video – VidRapid
– Improving image quality – replicate
– Converting Text into Video – Lumen5
– Music Track Generator – Mubert
– Movie Creation
– Changing Voice – Dubbing (voice.ai)
• What is Artificial Intelligence and Why It Matters in 2024?
• AI can introduce new sources of growth and change the
way work is done across industries.
• AI could contribute $15.7 trillion to global economy by 2035.
• China & US are accounting for nearly 70% of global impact
• What Is Artificial Intelligence?
• Learning, reasoning, problem solving, perception,
language comprehension are examples of cognitive
abilities
• Artificial Intelligence is a method of making a computer, a
computer-controlled robot, or a software think intelligently
like the human mind.
• AI is accomplished by studying the patterns of the human
brain and by analyzing the cognitive process.
• Outcome of these studies leads to intelligent software & system
Kinds of AI
Weak AI (Narrow AI)
Weak AI refers to AI systems that are designed
to perform specific tasks and are limited to
those tasks only.
These AI systems excel at their designated
functions but lack general intelligence.
Examples of weak AI include voice assistants
like Siri or Alexa, recommendation algorithms,
and image recognition systems.
Weak AI operates within predefined
boundaries and cannot generalize beyond
their specialized domain.
Strong AI (General AI)
It refers to AI systems that possess human-level
intelligence or even surpass human intelligence
across a wide range of tasks.
Strong AI would be capable of understanding,
reasoning, learning, and applying knowledge to
solve complex problems in a manner similar to
human cognition.
However strong AI is still largely theoretical and not
achieved to date.
Strong AI will be able to evaluate a scenario and
choose an appropriate action, even if it deviates
• Types of Artificial Intelligence
1. Purely Reactive - These machines do not have any memory or
data, specializing in just one field of work.
For example, in a chess game, the machine observes the moves
and makes the best possible decision to win.
2. Limited Memory - These machines collect previous data and
continue adding it to their memory.
They have enough memory or experience to make proper
decisions, but memory is minimal.
For e,g, machine can suggest a restaurant based on location.
3. Theory of Mind - This AI can understand thoughts and
emotions and interact socially – This machine is yet to be built.
4. Self-Aware - Self-aware machines are future generation ones
using new technologies.
They will be intelligent, sentient, and conscious.
For e.g, “I’m hungry” becomes “I want to eat dosa because it’s my
favorite food.”
 • Deep Learning vs. Machine Learning
• Machine Learning: ML develops algorithms and models for
computers to learn from data & make predictions /
decisions without programming.
• Key characteristics of machine learning: Feature Engineering:
• In ML, experts manually select relevant features from input
data to aid the algorithm in making accurate predictions.
• Supervised and Unsupervised Learning:
• ML algorithms are categorized into supervised learning,
where models learn from labeled data with known
outcomes, and unsupervised learning, where algorithms
discover patterns & structures in unlabeled data.
• Broad Applicability: ML techniques find application across
various domains, including image and speech recognition,
natural language processing & recommendation systems.
• Deep Learning:
• It is a subset of ML that focuses on training artificial neural
networks inspired by human brain's structure & functions.
• Key characteristics of deep learning:
• Automatic Feature Extraction:
• Deep learning algorithms automatically extract relevant
features from raw data, without using feature engineering.
• Deep Neural Networks:
• Deep learning employs neural networks with multiple layers
of interconnected nodes (neurons), enabling learning of
complex hierarchical representations of data.
• High Performance: Deep learning are widely used in domains
such as computer vision, natural language processing, and
speech recognition (shows better results than ML).
• How Does Artificial Intelligence Work?
• AI systems merge large data with intelligent, iterative
processing algorithms. This allows AI to learn from patterns
and features in the analyzed data.
• Each time an Artificial Intelligence system, tests and
measures its performance and uses results to develop
additional expertise (feedback)
• AI emphasizes three cognitive skills of learning, reasoning,
and self-correction, skills that the human brain possess.
• Three cognitive skills of learning are:
• Memory: Ability to store, retain, and recall information is
crucial for learning. Includes short-term & long-term memory
• Attention: Focusing on relevant information while filtering
out distractions is needed for effective learning - Attention
is needed for processing & understanding new information.
• Problem-Solving:
• Capacity to analyze situations, identify problems,
and develop solutions is a key cognitive skill.
• Problem-solving involves critical thinking, logical
reasoning, and creativity.
• We define these in the context of AI as:
• Learning: The acquisition of information and the
rules needed to use that information.
• Reasoning: Using the information rules to reach
definite or approximate conclusions.
• Self-Correction: The process of continually fine-
tuning AI algorithms and ensure that they offer
accurate results.
• Applications of AI Natural Language Processing (NLP)
• It is used to analyze and understand human language.
• It powers speech recognition, machine translation,
sentiment analysis, virtual assistants (Siri, Alexa).
• Image and Video Analysis
• Computer vision, enable analysis and interpretation of
images and videos.
• It is used in facial recognition, object detection & tracking,
content moderation, medical imaging, autonomous
vehicles.
• Robotics and Automation
• Robots equipped with AI can perform complex tasks in
manufacturing, healthcare, logistics, and exploration.
• They can adapt to changing environments, learn from
experience, and collaborate with humans.
• Recommendation Systems
• These are used in e-commerce, streaming platforms (Netflix),
and social media to personalize user experiences.
• They analyze user preferences, behavior, and historical data to
suggest relevant products, movie, music & content
• Financial Services
• AI is used in the finance industry for fraud detection,
algorithmic trading, credit scoring, and risk assessment.
• ML models can analyze vast amounts of financial data to identify
patterns and make predictions.
• Healthcare
• AI applications in healthcare include disease diagnosis, medical
imaging analysis, drug discovery, personalized medicine, and
patient monitoring.
• AI can assist in identifying patterns in medical data and
provide insights for better diagnosis and treatment.
• Virtual Assistants and Chatbots
• AI-powered virtual assistants and chatbots interact with users,
understand their queries, and provide relevant inf.
• They are used in customer support, information retrieval,
and personalized assistance (based on FAQ)
• Gaming - AI is used for creating virtual characters,
opponent behavior, intelligent decision-making and can
optimize game graphics, simulations, and game testing.
• Smart Homes and IoT
• AI can automate tasks, control devices, learn user
preferences (controlling A/C,fan based on sleeping pattern)
• Cybersecurity
• AI helps in detecting and preventing cyber threats by
analyzing network traffic, identifying anomalies, and
predicting potential attacks.
• AI Examples
• ChatGPT – It is an advanced language model, capable of
generating human-like responses and engaging in natural
language conversations and can understand and generate
coherent text for customer support, chatbots, virtual assistants.
• Google Maps - Provide real-time navigation, traffic updates, and
personalized recommendations (3D view, EV charging stations)
• It analyzes vast amounts of data, including historical traffic
patterns and user input, to suggest the fastest routes, estimate
arrival times, and even predict traffic congestion.
• Smart assistants – SAs like Amazon's Alexa, Apple's Siri,
Google Assistant employ AI to interpret voice commands,
answer questions, and perform tasks.
• These assistants use natural language processing and ML
algorithms to understand user intent, retrieve relevant
information, and carry out requested actions.
• Snapchat Filters
• Snapchat's augmented reality filters incorporate AI to
recognize facial features, track movements, and overlay
interactive effects on users' faces in real-time.
• Self-Driving Cars
• Self-driving cars use AI for perception, decision-making,
and control. Using many sensors, cameras, and ML
algorithms, these vehicles can detect objects, interpret
traffic signs, navigate complex road conditions
autonomously, enhancing safety and efficiency on roads.
• Wearables (Wellness)
• These devices such as fitness trackers and smart watches,
utilize AI to monitor and analyze users' health data.
• They track activities, heart rate, sleep patterns, and more,
providing personalized insights and recommendations to
improve overall well-being.
• Artificial Intelligence in Social Media
• Instagram – Here AI considers our likes and the accounts we
follow to determine what posts to show on our explore tab
• Facebook AI is used with a tool called DeepText.
• Facebook understand conversations better.
• It can be used to translate posts from many languages
automatically
• Twitter AI is used by Twitter for fraud detection, for removing
propaganda, and hateful content.
• Twitter uses AI to recommend tweets that users might enjoy,
based on what type of tweets they engage with.
• Artificial Intelligence in Marketing
• Using AI, marketers can deliver targeted and personalized ads
with the help of behavioral analysis.
• It helps with retargeting audiences at right time to ensure
better results & reduced feelings of distrust & annoyance.
• AI can help with content marketing in a way that matches
the brand's style and voice.
• It can be used to handle routine tasks like performance,
campaign reports, etc.
• Chatbots can analyze the user's language and respond in
the ways humans do.
• AI can provide users with real-time personalizations based
on their behavior and can be used to edit and optimize
marketing campaigns to fit a local market's needs.
• AI in Astronomy
• Given the huge amount of data, researchers created a system
that trained itself to locate merging galaxies.
• Night sky every night, gathers over 80 TB of data.
• One can study how stars and galaxies in cosmos change
over time.
• Using AI's time-series analysis, one can analyze data as and
identify planetary signals with up to 96% accuracy.
• AI in Travel and Transport
• Heavy Goods Transportation
• Truck platooning, which networks HGV (heavy goods vehicles),
for example, might be extremely valuable for vehicle transport
businesses or for moving other large items.
• The lead vehicle in a truck platoon is steered by a driver.
• As all of the trucks in the platoon are linked via a network, one
can activate the actions done by human driver.
• So, if the lead driver comes to a complete stop, all of the
vehicles following him do as well.
• Traffic Management
• Alibaba's City Brain in China uses AI such as predictive
analysis, big data analysis, and a visual search engine to
track road networks in real-time to reduce congestion.
• Ride-Sharing
• Uber and OLA use AI to improve user experiences by connecting
riders and drivers, improving user communication, messaging,
optimizing decision-making.
• For example, Uber uses a platform called Michelangelo that can
anticipate supply and demand, identify trip abnormalities like
wrecks, estimate arrival timings.
• Route Planning
• AI-enabled route planning is an approach for businesses, particularly
logistics and shipping industries, to construct a more efficient supply
network by anticipating road conditions and optimizing vehicle
routes.
• Predictive analytics in route planning uses many road usage
parameters such as congestion level, road restrictions, traffic
patterns, consumer preferences, and so on.
• Logistics companies use this to reduce delivery costs, accelerate
delivery times, manage assets & operations.
• AI in Automotive Industry
• Manufacturing
• Used in smarter factories, boosting productivity & lowering
costs.
• AI may be utilized in supply chain optimization, employing
robots on manufacturing floor, improving performance
using sensors, designing cars & post-production activities.
• Supply Chain
• AI helps firms identify the hurdles by forecasting and
replenishing supply chains as needed.
• AI may assist with routing difficulties, volume forecasts, …
• Passenger and Driver Experience
• AI may assist drivers in remaining focused by decreasing
distractions, analyzing driving behaviors.
• Passengers can benefit from customized accessibility as
well as in-car delivery services (JioMotive)
• Inspections
• Inspecting an automobile is subjective and manual. With AI, we
can identify where flaws are, and produce a status report.
• Quality Control People prefer premium vehicle and
experience. They want to predict in advance about the vehicle
failure (tires, oil, filter, clutch plate, break pedal changes)
• AI enables accurate predictive monitoring, fracture
detection, and other functions.
• Applications of Artificial Intelligence in Agriculture
• Identify defects and nutrient deficiencies in the soil using
computer vision, robotics, and machine learning applications,
• AI can analyze where weeds are growing.
• AI bots can help to harvest crops at a higher volume and
faster pace than human laborers.
• Artificial Intelligence in Education
• Administrative Tasks Automated to Aid Educators
• AI can do non-educational tasks like sending automated
personalized messages to students, back-office tasks like
grading paperwork, arranging for parent / guardian
interactions, feedback collection, managing enrollment,
course allotment, timetable , etc.
• Creating Smart Content
• Digitization of content like video lectures, conferences, and
textbook guides can be made using Artificial Intelligence.
• We can apply different interfaces like animations and
learning content through customization for students from
different grades.
• AI can create a rich learning experience by generating and
providing audio and video summaries & lesson plans.
• Voice Assistants
• Without even the direct involvement of a teacher, a student
can access extra learning material via Voice Assistants.
• Through this, printing costs of temporary handbooks and
provide answers to very common questions easily.
• Personalized Learning
• Hyper-personalization techniques can be used to monitor
students’ data thoroughly, and habits, lesson plans,
reminders, study guides, flash notes, frequency or
revision, etc., can be easily generated.
• AI in HR
AI helps with blind hiring - we can examine applications
based on specific parameters. AI can scan candidates'
profiles, and resumes to provide recruiters an
understanding of the talent pool of the candidate.
 Definition - Intelligence
• Intelligence is always a controversial topic.
• Are ants intelligent? Ant colonies / rats
intelligent?
• May be, rats are more intelligent than ants.
• Humans are more intelligent than ants and
rats.
• Many adults speak, write and play chess –
activities no animal can perform.
• Playing chess is not considered as intelligent
but if a one year old child does, it is super-
intelligent.
• If a dog does it, we call it a genius.
• Intelligence is related to our expectations.
 Some definitions of AI
• Ability to carry on abstract thinking(L.M.
Terman);
• Having learned or ability to learn to adjust
oneself to the environment (S.S. Colvin);
• Ability to adapt to new situations in life(R.
Pintner);
• A biological mechanism in which effects of
complex stimuli are brought together and given
a unified effect in behavior (J. Perterson);
• The capacity to acquire capacity (H. Woodrow);
• Capacity to learn by experience (W.F.
Dearborn).
• Some mention the environment, some don’t.
 Commonsense notations
• The ability to think is mentioned as an essential
characteristic of intelligence.
• Thinking in commonsense, includes problem
solving and logical reasoning but also less mental
activities such as those we use in our everyday lives.
• To measure intelligence conduct an experiment to
see whether something is thinking or not.
• For e.g, we can give a horse an arithmetic problem
in some form and if it comes up with right answer,
say by knocking the ground with correct number
of times, we may say that it is thinking.
 Learning and Memory
• Problem solving is closely related to thinking.
• Problem solving tasks  Finding a bug in a
computer program, diagnosing the disease of a
patient, finding a solution to a physics problem,
designing an experiment with animals to test a
hypothesis.
• Good students are usually perceived as the ones who
learn easily.
• We say that they have a good memory.
• They study the words once and they know them
and they do not forget them.
• Many people view learning as the core property of
intelligence.
• Learning to learn appears to be the key point.
 Language
• The capacity to communicate in natural language,
is considered to be the hallmark of intelligence.
• Natural language requires a high level of
intelligence.
• The ability to talk to one another, to read and write,
is one of the human’s distinguishing features.
(soft skills insist this).
• Those who speak multiple languages are regarded
as intelligent.
• Their ability is the combination of good learning,
memory capacity and talent for languages.
 Intuition and Creativity
• Einstein, Beethoven and Picasso were creative.
• They had a lot of intuition.
• Leaders and managers have intuition.
• It means arriving at conclusions without a train of
logical thought that can be traced to its origins.
• Creativity not only includes individual but also the
society as a whole.
• It cannot be defined for an individual in isolation
but must be discussed with respect to a particular
society’s value criteria.
• Many regard creativity as the highest form of
human intelligence.
 Emotions
• Humans have emotions.
• Most people think that higher mammals such as
apes, dolphins, dogs, and cats have emotions.
• People now talk about emotional intelligence.
• It refers to the ability to recognize emotions and
using emotions to support thinking and actions.
• Humans are jealous; they can be ashamed or feel
guilty.
• A lion is not guilt when it killed an animal whereas
a human is likely emotional when he kills another
human.
 AI and Computer Science
• AI is the study of computations that make it
possible to perceive, reason and act.
• AI differs from psychology because of the
emphasis on computation.
• AI differs from computer science because of
emphasis on perception, reasoning & action
• AI can be viewed as part engineering and part
science.
• The engineering goal of AI is to solve real-
world problems to represent knowledge,
using knowledge and assembling systems.
 Turing test approach
• Turing test (1950) was designed to define intelligence.
• Here, interrogator asks a test subject a series of
questions. Each party is in a separate area (no physical
contact)
• Responses given by test subject are evaluated to see
whether answers were given by a human / not.
• Computer passes the test if a human interrogator, cannot
tell whether written responses come from a person or not.

• Computer need to possess the following capabilities:

• NLP to enable it to communicate in English
• Knowledge representation to store what it knows or hears.
• Automated reasoning to use the stored information to
answer questions and to draw new conclusions.
• Machine learning to adapt to new circumstances and to
detect and extrapolate patterns.
 Turing Test as an imitation game
• It is played by 3 people, a man (A), a woman (B), an
interrogator (C) who may be of either sex.
• Interrogator stays in a room apart from other two.
• The object of the game is to determine which of
other two is the man and which is woman.
• At the end of the game he says either “X is A” (man)
and “Y is B” (woman) or vice versa.
• Interrogator is allowed to put questions to A & B such
as:
• C : Will X please tell me the length of your hair?
• Suppose X is actually A, then A must answer.
• A’s object is to confuse C to make wrong identification.

• His answer might be: “My hair is shingled (means

using a curly hair product, like a curl cream, hair gel),
• Turing test example questions
• What is your most memorable childhood event and how has
that impacted you today?
• Describe yourself using only colors and shapes.
• What historical event changed you the most and where were
you when it happened?
• What is the different being time flying and an airplane flying?
• There are variations to the Turing Test. Such examples include:
• Reverse Turing Test - where a human tries to convince a
computer that it is not a computer.
• An example of this is a CAPTCHA.
• Total Turing Test - where the questioner can test perceptual
abilities as well as the ability to manipulate objects.
• Minimum Intelligent Signal Test - where only true/false and
yes/no questions are given.
• Conversation between interrogator and computer is like this:
C(Interrogator): Are you a computer?
• A(Computer): No
• C: Multiply a large number to another, 158745887 * 56755647
A: After a long pause, an incorrect answer!
• C: Add 5478012, 4563145
A: (Pause 20 seconds and then give an answer) 10041157
• If the interrogator is not able to distinguish answers given by
both humans and computers then computer passes test and
machine(computer) is considered as intelligent as a human.
• In other words, a computer would be considered intelligent if its
conversation couldn’t be easily distinguished from a human’s.
• The whole conversation would be limited to a text-only channel
such as a computer keyboard and screen.
 Total Turing Test
• Turing test avoid direct physical interaction between
interrogator and computer.
• However total Turing test includes a video signal so
that the interrogator can test subject’s perceptual
abilities.
• Here verbal behaviors are not the sole standard for
intelligence: other behaviors are examined too.
• To pass the total Turing test, the computer will need:
• Computer vision to perceive objects and Robotics
to manipulate objects and move about.
• The quest for “artificial flight” succeeded when the
Wright brothers and others stopped imitating birds
and learned about aerodynamics.
 Cognitive Science
• Cognitive modeling approach: Cognitive science brings
together computer models from AI and experimental
techniques from psychology to try to construct precise
and testable theories of workings of the human mind.
• Thinking rationally: Greek philosopher Aristotle attempted
to codify “right thinking”.
• His syllogisms provided patterns for argument
structure that always yielded correct conclusions
when given correct premises.
• Ex: Socrates is a man; all men are mortal; therefore,
Socrates is mortal.
• All TVs use energy; Energy always generates heat;
therefore, all TVs generate heat.”
• Their study initiated the field called logic.
 Cognitive Science
• By 1965, programs existed that could, solve
any solvable problem described in logical
notation.
• There are 2 main obstacles to this approach.
• First, it is not easy to take informal
knowledge and state it in formal terms
required by logical notation, particularly
when knowledge is less than 100% certain.
• Second, even problem with just a few dozen
facts can exhaust computational resources
of any computer unless it has some
guidance as to which reasoning steps to try
first.
 The rational agent approach
• An agent is just something that acts.
• But computer agents are expected to have other attributes that
distinguish them from “programs” such as operating under
autonomous control, perceiving their environment,
persisting over a prolonged time period, adapting to
change, and being capable of taking on another’s goals.
• (intelligent computer agents include some spam filters, game
bots, and server monitoring tools, Search engine indexing bots)
• A rational agent is one that acts to achieve best outcome or
when there is uncertainty, the best expected outcome.
• Rational agents in AI: Self-driving cars make decisions
based on sensor data & optimize for safety and efficiency.
• Game-playing AI - AlphaGo, makes decisions based on the
game's rules and the board's current state to maximize the
chances of winning.
 Laws of Thought
• In the “laws of thought” approach to AI, the emphasis
was on correct inferences (process of running live
data through a trained AI model to make a
prediction or solve a task – signature verification
of a signed cheque).
• Making correct inferences is part of being a
rational agent.
• But correct inference is not all of rationality, because
there are often situations where there is no
provably correct thing to do, yet something must
still be done.
• For example, pulling one's hand off of a hot stove
is a reflex action that is more successful than a
slower action taken after careful deliberation.
 Perfect rationality
• Achieving perfect rationality – always doing the right thing –
is not feasible in complicated environment.
• Computational demands are too high. Limited rationality –
acting appropriately when there is not enough time to do all
computations is a good starting point for analysis
• Dave is a web developer. He gets a client call to discuss the
work details in 5 minutes. He cannot let them go, stating he
does not have time to quote a price for his services. As a web
developer, he has to discuss many details on the call.
• Developers usually quote based on the work done. Dave has
less time and he obviously would quote a higher price than
his previous assignment to make a quick decision.
• It is not that he has no experience in his field or does not know
how to price the services he offers. It is the time that presses
him to make a quick decision. In short, decision may not be
rational, as this assignment may demand more work from him.
• History of AI
• Following are milestones in the history of AI which
defines journey from the AI generation to till date.
• Maturation of Artificial Intelligence (1943-1952)
• Year 1943: The first work which is now recognized as AI was
done by Warren McCulloch and Walter pits in 1943. They
proposed a model of artificial neurons.
The first part, g takes an input,
performs an aggregation and based
on the aggregated value the second
part, f makes a decision.
The inputs are all boolean i.e., {0,1}
and output variable is boolean {0,1}
• Year 1949: Donald Hebb demonstrated an updating rule for
modifying the connection strength between neurons. His
rule is now called Hebbian learning.
• It is a single layer neural network, i.e. it has one input layer with
n units and one output layer.
• Year 1950: Alan Turing pioneered Machine learning in 1950.
Alan Turing published "Computing Machinery and
Intelligence" in which he proposed a test.
• The test can check the machine's ability to exhibit intelligent
behavior equivalent to human intelligence, called a Turing
test.
• The birth of Artificial Intelligence (1952-1956)
• Year 1955: An Allen Newell and Herbert A. Simon created
the "first artificial intelligence program“ which was named
as "Logic Theorist".
• This program had proved 38 of 52 Mathematics theorems,
and find new and more elegant proofs for some theorems.
• Year 1956: The word "Artificial Intelligence" first
adopted by American Computer scientist John
McCarthy at the Dartmouth Conference.
• At that time high-level computer languages such as FORTRAN,
LISP, or COBOL were invented.
• The golden years-Early enthusiasm (1956-1974)
• Year 1966: The researchers emphasized developing algorithms
which can solve mathematical problems. Joseph Weizenbaum
created the first chatbot in 1966, named as ELIZA.
• (https://web.njit.edu/~ronkowit/eliza.html)
• Year 1972: First intelligent humanoid
• robot was built in Japan which was
• named as WABOT-1. It consisted of a
• limb-control system, a vision system and
• a conversation system.
• The first AI winter (1974-1980)
• AI winter refers to the time period where computer scientist dealt
with a severe shortage of funding from government for AI
researches. During this time interest on artificial intelligence was
decreased.
• A boom of AI (1980-1987)
• Year 1980: AI came back with "Expert System". Expert
systems were programmed that emulate the decision-
making ability of a human expert.
• In 1980, first national conference of the American Association
of Artificial Intelligence was held at Stanford University.

• The second AI winter (1987-1993)

• Again Investors and government stopped funding for AI
research due to high cost but not efficient result.
• Emergence of intelligent agents (1993-2011)
• Year 1997: IBM Deep Blue beats world chess champion, Gary
Kasparov, first computer to beat a world chess champion.
• Year 2002: AI entered the home in the form of Roomba, a
vacuum cleaner.
• Year 2006: AI came in Business world till the year 2006.
• Companies like Facebook, Twitter, and Netflix started using AI.
• Deep learning, big data & artificial general intelligence
(2011+)
• Year 2011: In the year 2011, IBM's Watson won a quiz show,
where it had to solve complex questions as well as riddles.
Watson proved that it could understand natural language
and solve tricky questions.
• Year 2012: Google has launched an Android app feature
"Google now", which was able to provide information to
the user as a prediction.
• “Google Now is an intelligent personal assistant. Google Now is
available within the Google Search mobile application for
Android and iOS, and Google Chrome web browser on
personal computers. Google Now uses a natural language user
interface to answer questions, make recommendations, and
perform actions by delegating requests to a set of web services.
• Along with answering user-initiated queries, Google cards
proactively delivers to users more information that it predicts
(based on their search habits) they may want.”
• Year 2014: In the year 2014, Chatbot "Eugene Goostman"
won a competition in the infamous "Turing test."
• Year 2018: The "Project Debater" from IBM debated on
complex topics with two master debaters and also
performed extremely well.
• Google has demonstrated an AI program "Duplex" which was
a virtual assistant and which had taken hairdresser
appointment on call, and lady on other side didn't notice
 Old History of AI
• Philosophy (428 B.C – Present)
• Can formal rules be used to draw valid conclusion?
• Where does knowledge come from?
• How does knowledge lead to action?
• Aristotle developed a system of syllogisms for proper reasoning
which allowed one to generate conclusions mechanically, given
initial premises.
• Syllogism - An instance of a form of reasoning in which a
conclusion is drawn from two given or assumed
propositions (premises);
• a common or middle term is present in the two premises
but not in the conclusion, which may be invalid
• (e.g. all dogs are animals;
• all animals have four legs;
• therefore all dogs have four legs ).
 Automation of computation
• The automation of computation started around 1500 by Leonardo
da Vinci who designed but did not build a mechanical calculator.
• The first known calculating machine was constructed around 1623
by the German scientist Wilhelm Schickard (1592-1635), although
Pascaline, built in 1642 by Blaise pascal is more famous.
• Gottfried Wilhelm Leibniz built a mechanical device intended to carry
out operations on concepts rather than numbers, but its scope
was rather limited.
• Logical Positivism
• Rudolf Carnap says that all knowledge can be characterized by
logical theories connected, to observation sentences that
correspond to sensory inputs.
• Carnap and Carl Hempel says knowledge can be acquired from
experience.
• Goal based analysis does not say what to do when several actions
will achieve the goal, or when no action will achieve it
completely. Antoine Arnauld gave a quantitative formula for
deciding what action to take in case like this.
 Mathematics (c.800 – present)
• What can be computed? How do handle uncertain information?
• There are 3 fundamental areas of mathematical formalization of
AI: logic, computation, and probability.
• George Boole (1815-1864), who worked out the details of
propositional or Boolean logic.
• In 1879, Gottlob Frege extended Boole’s logic to include objects
and relations, creating the first-order logic that is used today as
the basic knowledge representation system.
• Alfred Tarski introduced a theory of reference that shows how to
relate the objects in a logic to objects in the real world.
• Algorithm: The first algorithm is Euclid’s algorithm for
computing greatest common denominators.
• In 1931, Kurt Godel proposed the incompleteness theorem
and showed that in any language expressive enough to
describe the properties of the natural numbers, there are
true statements that are undecidable in the sense that their
truth cannot be established by any algorithm.
 Computable functions
• Chruch-Turing thesis states that the Turing machine is
capable of computing any computable function, is
accepted as providing a sufficient definition.
• Turing also showed that there were some functions that no
Turing machine can compute.
• For example, no machine can tell in general whether a
given program will return an answer on a given input or
run forever.
• Tractable Problems
• A problem is called intractable if the time required to
solve instances of the problem grows exponentially
with the size of the instances.
• It is important because exponential growth means even
moderately large instances cannot be solved in any
reasonable time.
• Therefore we should try to divide the overall problem into
tractable sub problems rather than intractable ones.
 Neuroscience
• It is the study of the nervous system, particularly the brain.
• Of all animals, man has largest brain in proportion to his size.
• The brain is consisted of nerve cells or neurons.
• Each neuron consists of a cell body or soma, that contains a cell nucleus.
Branching out from the cell body are a number of fibers called dendrites and a
singe long fiber called the axon.
• The axon stretches out for a long distance.
• They are 1cm long (100 times the diameter of the cell body), but can reach up
to 1 meter. A neuron makes connections with 10 to 1000,000 other neurons at
junctions called synapses.
• Signals are propagated from neuron to neuron by a complicated
electrochemical reaction.
• The signals control brain activity in the short terms and also enable long term
changes in the position and connectivity of neurons.
• These mechanisms are thought to form the basis for learning in the brain.
• There is no theory on how an individual memory is stored
The measurement of brain activity started by Hans Berger of
electroencephalograph (EEG).
Recently, functional magnetic resonance imaging (fMRI) are giving detailed
images of brain activity.
Brains and digital computers perform quite different tasks and have different
properties
•Properties Computer Human Brain
•-------------------------------------------------------------------------------------------
•Computational units 1 CPU, 10^8 gates 10^11 neurons
•Storage units 10^10 bits RAM 10^11 neurons
• 10^11 bits disk 10^14 synapses
•Cycle-time 10^-9 sec 10^-3 sec
•Bandwidth 10^10 bits/sec 10^14 bits/sec
•Memory updates/sec 10^9 10^14
•------------------------------------------------------------------------------------------
Probability
Gerolamo Cardano (1501-76) framed the idea of probability,
describing it in terms of the possible outcomes of gambling
events. Probability is, helping to deal with uncertain
measurements and incomplete theories
 Brain vs computer
• From the above table, we have 1000 times more neurons in
the human brain than are the gates in CPU of a high-end
computer.
• Moore’s law predicts that the CPU gate count will equal the
brain’s neuron count around 2020.
• The difference in storage capacity is minor compared to the
difference in switching speed and in parallelism.
• Computer chips can execute an instruction in a nanosecond,
whereas neurons are millions of times slower.
• But in brains all the neurons and synapses are active
simultaneously, where as most computers have only one or
at most a few CPUs.
• Thus even though a computer is a million times faster in raw
switching speed, the brian ends up being 100,000 times
faster at what it does.
 Linguistics
• Modern linguistics and AI were “born” at the same time and
grew together, intersecting a hybrid field called computational
linguistics or natural language processing.
• Understanding language requires an understanding of the
subject matter and context, not just an understanding of the
structure of sentences.
• Much work in knowledge representation (the study of how to
put knowledge into a form that a computer can reason with)
was tied to language.
• LISP
• In MIT AI lab, McCarthy defined the high level language LISP,
which become the dominant AI programming language.
• Lisp is the second oldest major high-level language in use, one
year younger than FORTRAN.
• But Lisp needed scarce and expensive computing resources.
• To resolve this he invented time sharing.
 AI in translations
• For example, when Sputnik was launched in 1957, the US National
Research Council wanted to speed up the translation of Russian
scientific papers.
• It was thought initially that simple syntactic transformations based on
grammars of Russian and English, and word replacement using an
electronic dictionary, would be enough to preserve the exact meaning
of sentences.
• The fact is that translation requires general knowledge of the subject
matter in order to resolve ambiguity and establish the content of the
sentence.
• The second kind of difficulty was the intractability of many of the
problems that AI was attempting to solve.
• Early AI programs solved problems by trying out different
combinations of steps until the solution was found.
• This worked initially as microworlds contained very few objects and
hence few possible actions and short solution sequences.
• People that scaling up to larger problems was a matter of faster
hardware and larger memories. But it was not the case.
 Multilayer networks
• The third difficulty arose because of some fundamental
limitations on the basic structure being used to generate
intelligent behavior.
• For example, a two-input perceptron could not be trained to
recognize when its two input were different.
• Subsequently the network propagation algorithms for
multilayer networks were discovered.
• Game Playing - Autonomous control
• The ALVINN computer vision system was trained to steer a
car to keep it following a lane.
• It was navigating about 2850 miles.
• Here a human steered only 2%, mostly at exit ramps.
• NAVLAB has video cameras that transmit images to ALVINN,
which computers the best direction to steer, based on
experience from previous training runs.
 MYCIN
• Feigenbaum started the Heuristic Programming Project
(HPP), to investigate the extent to which the new
methodology of expert systems could be applied to other
areas of human expertise.
• His team developed MYCIN to diagnose blood infections.
• With about 450 rules, MYCIN was able to perform as well as
some experts, and considerably better than junior doctors.
• Here the rules were acquired from extensive interviewing of
experts, who in turn acquired them from text books, other
experts, and experience of cases.
• Rules also reflected uncertainty associated with medical
knowledge.
• MYCIN used a calculus of uncertainty called certainty
factors.
Emergence of intelligent agents
• The so-called situated movement aims to
understand the working of agents embedded in
real environments with continuous sensory
inputs.
• One of the most important environments for
intelligent agents is the Internet.
• AI systems are common in web based
applications that the “-bot” suffix has entered
everyday language.
• AI technologies are used in every Internet tools
such as search engines, recommender systems
and web site construction systems.
 Diagnosis
• Medical diagnosis programs based on probabilistic analysis
were able to perform at the level of an expert physician in
several areas of medicine.
• Here creator of the program suggests he ask the computer for
an explanation of the diagnosis.
• The machine points out the major factors influencing its
decision and explains the subtle interaction of several of the
symptoms in this case.
• Eventually, the expert agrees with the program.
• Logistics Planning
• During Gulf war in 1991, US deployed an automated logistics
planning and scheduling for transportation.
• This involved 50,000 vehicles, cargo and people and had to
account for starting points, destinations, routes and conflict
resolution among all parameters. It did planning in hours that
would have taken weeks with older methods.
 Robotics
• Many surgeons use robot assistants in microsurgery.
HipNav is a system that uses computer vision
techniques to create a 3-D model of a patient’s
internal anatomy and then uses robotic control to
guide the insertion of a hip replacement prosthesis.
• Language understanding and problem solving
• PROVERB is a computer program that solves
crossword puzzles between than most humans, using
constraints on possible word fillers, a large database
of past puzzles and a variety of information sources
including dictionaries and online databases such as
list of movies and actors that appear in them.