Artificial Intelligence

Introduction to Artificial Intelligence


August 23, 2021, Learn eTutorial
1480

Artificial Intelligence, in short AI, is now a fast-growing technology and one of the fascinating and universal fields of the computer science environment and has the greatest scope in the future. We are living in an era where everything in one way or another is related to AI. Nowadays, can you think about a day without your Smartphone? Literally for each and everything we need this simple handset. When AI was incorporated with this simple handset it became a multifunctional system and from ‘phone’ it became ‘Smartphone’. So, what is AI?

What is AI?

What is AI?

In this world, we all are surrounded by machines and smart devices, many of them working based on the AI principle.
Let us think of a self-driving car, the car itself is making decisions without the help of a driver from the data it receives from the various sensors like the human brain. How is that possible? The answer is AI.

Let us take another example as a chess game. Have you ever thought about how the computer or your smartphone is playing perfectly like a human opponent? Again, the answer is AI or Artificial Intelligence. 

Let’s try to understand the meaning of the term artificial intelligence by splitting the term into two: Artificial + Intelligence. The term ‘Artificial’ refers to something produced by human beings rather than occurring naturally, more specifically a copy of something natural. The word ‘Intelligence’ refers to the ability to acquire and apply knowledge and skills. So, in short, Artificial Intelligence is an artificially created Intelligence.

The incredible power of humans to use their brain (intelligence) makes them distinguishable from other creatures. Learning, Reasoning, and Problem Solving are the key factors of human intelligence. Human Intelligence is the mental quality to think rationally, act purposefully, and deal effectively with one’s environment. When this human Intelligence integrates on machines to mimic human behavior, a new technology emerges and is named Artificial Intelligence.If a machine is using AI, there is no need for pre-programming or re-programming for different situations like a normal machine.  AI machines can understand and make decisions for situations or conditions that never happened before.

Definition for AI

How can we define intelligence, so difficult right? Same in the case of AI, AI is not just one technology, it is a set of technologies working together to make a machine to sense, grasp, act and learn and make decisions with intelligence. Hence it is difficult to define AI in one way and so AI holds a different set of definitions. Let’s have a look at how different people define AI:
John McCarthy, one of the men who positioned as the father of AI coined the term ARTIFICIAL INTELLIGENCE in the 1950s. He said “Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it. An attempt will be made to find how to make machines use language, form abstractions, and concepts, solve the kinds of problems now reserved for humans, and improve themselves.”

The English Oxford Living Dictionary gives this definition:
The theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages”.

Another definition for AI is: “Artificial Intelligence is the stimulation of human Intelligence on man-made machines that are well programmed with a set of algorithms which can work with their own Intelligence and act like a human

Why AI is important?

Before going to a deep study about artificial intelligence we must have knowledge about why this new technology is needed? What can AI offer to our world? and much more. Let us discuss some of the uses of AI in the modern world.

  • We can easily solve our real-life problems with the help of some software based on AI with much more precision than humans like medical diagnosis, traffic, weather, share trading, etc.
  • We all have used our virtual assistants like google, siri or alexa etc. all are based on AI.
  • Image processing and pattern analyzing which is an AI application-level program is used in various fields now
  • We are able to create robots with the independent ability to make decisions dynamically which will help humans in critical tasks without any loss of human life.

How AI works?

We have already discussed that AI is not a single technology, in General, we can explain the working of AI as a machine that is able to receive a huge amount of data from different devices and it processes and learns from that data using different intelligent algorithms and making decisions dynamically.
AI makes the machine make decisions using a large number of components of technologies which we can call subdomains. To understand How Artificial Intelligence actually works, you need to get to know about the subdomains of AI and understand how those domains are applied.

  • MACHINE LEARNING:  Machine learning teaches a machine to make inferences and decisions from past experience.
  • DEEP LEARNING: An ML technique teaches a machine to process inputs to classify, infer and predict the outcome.
  • NEURAL NETWORKS: Neural Networks work as similar as human neural cells. A series of algorithms that capture the relationship between various variables and process the data as a human brain does.
  • NATURAL LANGUAGE PROCESSING: A technology of reading, understanding, interpreting a language by a machine.
  • COMPUTER VISION: Algorithms try to understand an image by breaking down the image and studying different parts.
  • COGNITIVE COMPUTING: Algorithms try to mimic a human brain by analyzing text, images, voice, objects like a human does and tries to produce the desired output.

3 stages of Artificial Intelligence

What is AI?

Now we know about AI and the need for AI and how AI works. Now let us check the lifecycle of AI. What are the different stages of Artificial intelligence? 

STAGE 1: - Machine learning

Set of algorithms used to teach a machine to make inferences and decisions from past experience

STAGE 2: - Machine intelligence

The advanced round of algorithms is used by machines to learn from experience. Deep Neural Networks is an example. Currently, Artificial Intelligence technology is at this stage.

STAGE 3: - Machine consciousness

Machines learn themselves from experience without the need for external data.

Types of AI

What is AI?

We mentioned above that AI has a wide range of applications and functionalities in the modern world. There are a lot of areas where these AI applications are used, that starts from a simple game to even in military operations. So generally Artificial intelligence can be broadly divided into two types based on the capability of an AI machine and the functionalities that an AI machine can do.

Based on capabilities AI is classified into three:

  1. NARROW (WEAK) AI: Capable of performing single tasks such as autonomous cars, retail kiosks, etc.;
  2. GENERAL (STRONG) AI: Said to have capabilities and functions like the human mind. e.g.: AI in games.
  3. SUPER AI: Hypothetical agent that possesses capabilities that will exceed human intelligence (and conceivably take over the world).

Based on functionalities artificial intelligence is classified into four:

  1. REACTIVE AI: React only to the existing situations, not past experiences.
  2. LIMITED MEMORY AI: Learn from recent experiences to make decisions.
  3. THEORY OF MIND AI: Machines are able to understand emotions and able to differentiate between various emotions of different people.
  4. SELF-AWARE AI: Systems form representation of themselves.

What contributes to AI?

What is AI?

Artificial Intelligence is not just computer programming, it is not just a technology. We already stated that AI is making a machine to analyze and learn from the data and make decisions from the learned data. Let us think of the intelligence of a human, we know intelligence is a combination of Reasoning, Learning, Problem Solving, Perception, Linguistic Intelligence, etc. like the same way artificial intelligence comprises a lot of computer and mathematical aspects. To achieve the above factors for a machine AI requires the assistance of

  • Computer Science
  • Philosophy
  • Psychology
  • Sociology
  • Maths
  • Biology
  • Neuron Science

Goals of AI

Now we know about artificial intelligence and it's time to discuss what we aim in using artificial intelligence or what are achievements we are going to make using artificial intelligence. 

  • Reasoning and problem-solving.
  • Knowledge representation.
  • Artificial planning.
  • Develop intelligent Machines that could learn on their own.
  • Develop machines that can read and understand human language.
  • Developing machines take input from sensors and react accordingly.

History of AI

1943

The first artificial neuron ‘The Threshold Logic Unit (TLU), or Linear Threshold Unit’ was proposed by Warren McCulloh, a neurophysiologist, and Walter Pitts a Mathematician. The model was specifically targeted as a computational model of the “nerve net” in the brain.

1949

Donald Hebb a psychologist introduced the neuro-scientific concept of Hebbian learning also known as Hebb’s Rule or Cell Assembly Theory, which in turn attempts to connect the psychological and neurological underpinnings of learning.

1950

An imagination game (Turing Test) by Alan Turing, a mathematician, was proposed to test a machine’s ability to exhibit intelligent behavior equivalent to that of a human.

1955

The first Artificial Intelligence program deliberately engineered to perform automated reasoning called Logic Theorist was written by Allen Newell, a researcher in computer science and cognitive psychology, Herbert A. Simon, an economist, political scientist, and cognitive psychologist, and Cliff Shaw, a system programmer. Logic Theorist proved 38 of the first 52 theorems in chapter 2 of Rusell’s Principia Mathematica. One of the proofs was actually more elegant than the proof produced by hand by Rusell

1956

John McCarthy, a computer Scientist organized the Dartmouth Conference, at which the term ‘Artificial Intelligence’ was first adopted.

1956-1959

High-level computer languages such as FORTRAN, LISP, or COBOL were invented.

1966

The first chatbot (AI software that can simulate a conversation with a user in natural language) named ELIZA was created by Joseph Weizenbaum.

1971

The first expert system, DENDRAL, was introduced by Edward Feigenbaum, a computer scientist.

1972

The first anthropomorphic robot called WABOT was developed in the world.

1974- 1980

The duration between the years 1974-1980 was known as the first AI winter. During this period of time, the public interest in AI appears to wane, funding for activities geared toward developing AI in machines is lacking.

 

1980

The first National Conference on Artificial Intelligence was held at Stanford California.

1987-1993

The duration between the years 1987-1993 was the second AI winter. Again lack of funding became a big problem due to high cost. The expert system such as XCON was very cost-effective.

1997

An IBM computer called IBM Deep Blue beat the world chess champion after a six-game match

2002

iRobot launches the Roomba which navigates the floor area of a home and cleans it.

2006

The Dartmouth Artificial Intelligence Conference for the Next 50 Years (AI@50) took place.

2009

Google builds autonomous cars.

2010

Microsoft launched Kinetic for Xbox 360, the first gaming device to track human body movement.

2011-2014

Apple’s Siri, Google’s Google Now, Microsoft’s Cortana.

2015

Google Deepmind’s AlphaGo defeated 3-time European Go champion 2 dan professional Fan Hui by 5 games to 0.

2018

An AI program “Duplex” was demonstrated by Google which was a virtual assistant.

2020

Meeting Owl, Kuri Mobile Robot, etc.

Advantages of Artificial Intelligence

Following are some of the main advantages of artificial intelligence

  • HIGH ACCURACY - The ability to make the right decision is one of the main advantages of AI. AI-powered machines use cognitive computing that makes them capable of making logical decisions without any emotions in real-time.
  • LESS ERROR - AI-based machines help to solve complex real-world problems. Since the machines are programmed for accomplishing a specific task, the chances for error are very less and it depends on how well we design and programmed the machines to carry out the task.
  • HIGH RELIABILITY - AI-based machines help to perform repetitive tasks for a long time without any slowdown. It can be working for an indefinite time, without lacking productivity.
  • DIGITAL ASSISTANCE - AI-powered applications provide digital assistance. Today, most organizations make use of digital assistance to perform automated tasks. There are many practical applications of AI-based digital assistance such as Google Maps, Grammarly, Alexa, and many more.
  • USEFUL FOR RISKY ENVIRONMENTS - AI-based machines can be used for areas where human life might be at risk like defusing a bomb, exploring the ocean floor, etc.
  • HIGH SPEED - AI systems can be of very high speed and also fast decision making.
  • USEFUL AS A PUBLIC UTILITY - Fraud detection, making welfare payments and immigration decisions, etc. are some examples of the public utility application of AI.
  • 24*7 AVAILABILITY - Human employees can’t work for 24 hours as they need time to refresh themselves. AI can provide services without any delay or lack of efficiency.

Disadvantages of Artificial Intelligence

Being so advantageous, AI still has some disadvantages which must be kept in our mind while creating an AI system. Following are some of them:

  • HIGH COST-The creation of AI-empowered machines is very expensive. To meet the demand of this changing world, the hardware and software should be regularly updated. AI-powered devices are built with complex codes, algorithms, software, and hardware. The maintenance of these requires great effort and a large amount.
  • INCREASED UNEMPLOYMENT-AI replaces low-skilled employees. As AI-based machines can work 24*7 With increased accuracy, industries prefer AI rather than employing humans.
  • LACKING CREATIVITY-Still AI machines cannot beat the power of human intelligence to create or imagine new ideas. AI can just perform the task it is programmed for and improve itself by experience.
  • LACKING IMPROVEMENT-AI algorithms are designed to make the machines learn from the data they get. Any redundancy in the data may cause failures in learning and unpredictable results may be generated. They only can work for which they are programmed. Thus, due to the lack of improvement, AI-generated results may have inaccuracy and cause great losses.
  • NO FEELINGS AND EMOTIONS-AI powered machines work on algorithms, Mathematical Computing, and cognitive technologies. They lack judgmental power as they are not aware of ethics, morals, right, the wrong. It would be a great failure if we try to implement AI in places that require strong judgmental ability.
  • INCREASED MACHINE DEPENDENCY-With the improvement of technology, people are getting more machines dependent and hence they are losing their mental capabilities.

Applications of AI

What is AI?

Nowadays there are a lot of areas we are using AI which can be categorized according to the use and functionality of the AI. See the below table to understand the categories and their applications where AI is used at the Application level.

Area

Applications

DEEP LEARNING APPLICATION

 

Fraud Detection.

Autonomous Cars.

Virtual Assistants.

Supercomputing.

Customer Relationship Management (CRM) Systems.

Investment Modeling.

Facial Recognition Systems.

PREDICTIVE ANALYTICS

 

Google Flu Trends (GFT)

Bing

Amazon

Netflix

Coursera

edX

Udemy

TRANSLATION

 

Lingua Custodia

SYSTRAN

SDL Government

Canopy Innovations

Text grabber

The ili

Pilot

Google Translate

Facebook

Skype

CLASSIFICATION AND CLUSTERING

 

Fantasy Football and Sports

Document Analysis

Identifying Fraudulent or Criminal Activity

Classifying Network traffic

Marketing and Sales

Spam Filter

Identifying Fake News

INFORMATION EXTRACTION

 

Business Intelligent

Financial Investigation

Scientific Research

Media Monitoring

Health Care

Record Management

Pharma Research

SPEECH TO TEXT

 

Dragon Anywhere

Google Assistant

Transcribe-Speech to Text

Speech Notes- Speech to Text

Voice Notes

Braina

Speech Texter- Speech to Text

TEXT TO SPEECH

 

Speech Central

Voice Dream Reader

Motoread

Voice Aloud Reader

The Verdict

Murf

Notevibes

IMAGE RECOGNITION

 

Google Lens

Flow

TapTapSee

Leaf Snap

Calorie Mama

Vivino

Screenshop

MACHINE VISION

Electronic component analysis

Signature identification

Optical character recognition

Handwriting recognition

Object recognition

Pattern recognition

Materials inspection

Currency inspection

Medical image analysis

PLANNING, SCHEDULING AND OPTIMIZATION

 

Top-K Planners

Semi-Black box for transportation

Planner 4J

DOCIT

Plan recognition as Planning

Google Calendar

Calendar Hero

ROBOTICS

 

Starship Delivery robots

Pepper Humanoid Robot

Penny Restaurant Robot

Nimbo Security Robot

Shadow Dexterous Hand

Moley Robotic Kitchen system

Flippy Robotic Kitchen Assistant

EXPERT SYSTEM

 

MYCIN

DENDRAL

PXDES

CaDet

Top used applications in artificial intelligence in our daily life

  • Google’s AI-powered predictions (E.g.: Google Maps).
  • Ride-sharing applications (E.g.: Uber, Lyft).
  • AI Autopilot in Commercial Flights.
  • Spam filters on E-mails.
  • Plagiarism checkers and tools.
  • Facial Recognition.