Banner Ads

How does Artificial Intelligence work? Everything explained!

In today's modern world, gadgets are able to perform a lot more tasks in a very short span of time. Most of you will say ya we already know that this is just because of the advancement in the field of technology, and yes that`s true but it was never possible without Artificial Intelligence (AI).
Artificial Intelligence played a key role in the advancement of technology and the most amazing part is that it is just in its initial phase, there is a long way to go. Artificial Intelligence is opening new possibilities in the field of technology. From Unmanned vehicles to Cloud gaming, AI has made its mark in nearly every field. 

AI is the ability of a computer system to solve problems that are usually done by humans by our human intelligence. Machines with AI can nearly do everything that a human brain can do and can also perform tasks that are beyond our reach.

AI is growing exponentially and according to a report of PWC, AI will contribute $15.7 trillion to the global economy by 2030. After all this, you might be wondering what it is and how does artificial intelligence work?


How does Artificial Intelligence work?

How does artificial intelligence work?

In the past 60 years, AI has progressed a lot. The first computer with AI capabilities was Deep Blue from IBM. The computers which are equipped with AI can do various tasks like Voice recognition, face recognition, learning, planning, and problem-solving.

Nowadays, AI is used in self-driving vehicles to detect obstacles, in the medical sector to provide analytics of your health and even in the film industry to improve graphics. It can also help to save lives by doing tasks that would be dangerous or cannot be done by humans.


Smartwatches like Apple watch 3 has AI features that monitor your health, processes it, and alarms you before any health-related incident occurs. Many industries are adopting AI to do tasks quickly and accurately. It is one of the best future technologies that will change the world. Before moving towards how does ai works, we will first understand 
what does artificial intelligence mean.

What does artificial intelligence mean?

Artificial Intelligence is a sector of computer science that emphasizes the creation of intelligent computers that react and work like humans. Some of the complex tasks computers with artificial intelligence are designed for include:
  • Observing
  • Learning
  • Planning
  • Problem Solving
AI involves 'Deep Learning', which allows computers to process inputs and perform complex functions, including voice or face recognition. 

History of artificial intelligence

The American computer scientist John McCarthy was the first to invent the term "Artificial Intelligence" at the "Dartmouth Summer Research Project on Artificial Intelligence" in 1956. Artificial means something that is made by human beings and Intelligence means the ability to acquire and apply knowledge and skills.

In the 1960s, the US Department of Defence took an interest in AI and started training computers to mimic human intelligence. The projects on street mapping were completed in the 1970s by the Defense Advanced Research Projects Agency (DARPA) 1970s. DARPA`s first personal intelligent assistant was developed in 2003, way before google assistant, Alexa, Siri or Cortana were born. Let`s take a quick look at some of the most important events in the history of AI.

1943 
Walter Pitts and Warren McCullough published "A Logical Calculus of Ideas Immanent in Nervous Activity." The first mathematic model for building a neural network was proposed in this paper.

1949
Canadian Psychologist Donald Olding Hebb proposed the theory also known as cell assembly theory or the Hebbian theory. The Hebbian theory states that neural pathways are developed from experiences and connections between neurons become stronger as they are more frequently used. Donald Hebbian`s learning continues to be an important model in Artificial Intelligence.

1950
Two Harvard undergraduates Dean Edmonds and Marvin Minsky engineered the first neural network computer known as SNARC.

Claude Shannon publishes the paper for Programming a Computer for Playing Chess.

1956
The term artificial intelligence was first coined in 1956 by John McCarthy at the "Dartmouth Summer Research Project on Artificial Intelligence". 

Herbert A. Simon and Allen Newell successfully engineered the first reasoning program called Logic Theorist (LT). This was first of its kind which was developed to mimic the problem-solving techniques of a human. It is also known as the first artificial intelligence program.

1959
The history of Machine Learning lies back in 1959 when an American pioneer "Arthur Samuel" gave the first definition while at IBM. He was in the field of artificial intelligence and computer gaming.

2012
Andrew Ng founded the Google Brain which was based on Deep Learning. It involved building neural networks using deep learning algorithms. Nearly 10 million YouTube videos were used for training purposes. The neural network was able to recognize a cat without being told what a cat is, guiding towards the era of neural networks and deep learning funding.

2014
Google`s self-driving car becomes the first to clear a state driving test.

2016
Google AlphaGo defeated world champion Go player Lee Sedol. The complexity of this ancient Chinese game was seen as a big hurdle to clear in Artificial Intelligence.

Now the main question arises, how does artificial intelligence work?


How does Artificial Intelligence work?

How does Artificial Intelligence work?

The aim of AI is to create such systems that are capable of functioning intelligently and independently. Artificial Intelligence is an intelligent algorithm that combines a large amount of data with fast iterative processing and learns automatically from the patterns or historical data. AI improves its capabilities with use and time. There are some subfields of AI which include Speech Recognition, Machine Learning, Neural network, Deep learning, Natural Language Processing (NLP), Computer Vision.

Speech Recognition-

It is the ability of a program or machine to recognize and translate spoken language into text. It is also called speech to text.

Machine Learning-

Machine Learning is an algorithm or system which uses patterns and dimensions of pattern to distinguish a specific object. This algorithm is trained to perform a specific task. This system works in a constrained environment and cannot be used to perform different tasks. For example, take a system integrated with a camera. If I show some images of dogs to this camera, it will analyze patterns of all these images and accordingly store them into its database. Next time if I show other photos of dogs to this system, it will detect these photos and recognize it as a dog because it has learned from all those previous patterns. But if I show this system a photo of a cat, it won`t be able to answer because it was not trained for it.

Neural Network-

A human brain contains neurons which we use to learn and process new things. If we replicate this structure and functioning of neurons, we will be able to achieve cognitive capabilities in machines. This comes under neural networks. Neural Networks include Deep Learning and Machine Learning. 

Deep Learning-

Deep learning is a way in which we train and use neural networks to solve complex problems. Examples are voice and image recognition.


Computer Vision-

It is the ability of a system to obtain information from a digital image or video. 


Natural Language Processing (NLP)-

The ultimate goal of NLP is to give computers the ability to analyze, understand and generate human language including speech also. Most types of NLP use machine learning to understand the meaning of human language.

Facebook, Twitter, Google translate are some of the apps that use NLP.

Microsoft Word and Grammarly use this technique to find any grammatical errors in the input text.

There are several methods to teach an AI system like Supervised Learning, Unsupervised Learning, and Reinforced learning.

Types of Artificial intelligence-

Types of Artificial Intelligence

AI can be classified into two types which are based on their capabilities and functionality.

Type 1: Based on Capabilities

Weak AI or Narrow AI: Weak AI is a type of AI that is capable of performing a dedicated task autonomously with its intelligence. Weak AI or Narrow AI is the most commonly used AI. 

Systems equipped with Weak AI cannot perform a task that is beyond its reach as it is only trained for a specific task and if it tries to perform, it will fail. The most complex AI systems that use Machine Learning and Deep Learning to teach itself fall under the category of Weak AI.

Artificial General Intelligence or General Intelligence: AGI is the ability of an AI system to analyze, understand and function with human-like intelligence. The idea behind using AGI is to make systems that are smart and capable of replicating multi-functional intelligence to performing tasks on their own. 

Artificial Super Intelligence (ASI): ASI is the level of AI at which the system exceeds the limit of human intelligence and performs tasks better than humans with cognitive abilities. These systems are better at everything because of their abilities like greater memory, faster processing of data and analysis, and capabilities of decision making. ASI systems can think, solve and can make judgments on their own. 

After reading all this, you might be wondering how exciting it would be to see humanoids walk and talk like humans do but let me tell you that if something has pros there will be cons of that also. But to develop ASI systems there is still a long way to go.


Type 2: Based on the functionality

Reactive machines: These machines are the oldest types of AI which have got only limited capabilities. These AI systems have no memory and cannot store past experiences, data to learn new actions. This AI system only focuses on the current situation and can be used for automatically replying to a limited number of questions. An example of this is IBM`s Deep Blue and Google`s AlphaGo.

Limited Memory: These machines can store past data, experience in its memory for a short period of time to improve its response. For example, take an image recognition system that is trained by feeding a large number of images with labels to teach it to name objects it scans. It takes training images as a reference and scans new images. With its learning experience, it labels new images with increased accuracy. Real-Life examples of Limited Memory systems are virtual assistants and self-driving cars.

Theory of Mind: TOM systems are currently in the developing phase and if they become a reality, they would be able to understand our emotions, beliefs, needs, thought processes and would be able to interact socially with us like humans. Researchers are putting a lot of effort to develop such systems and improving other systems as well to complete the whole network.

Self-aware: These machines would be supreme leaders of AI as they will have own consciousness, sentiments, and self-awareness. These systems will be much smarter than humans but to date, it's just a hypothetical concept.

If you want to know more about types of AI in detail,


Checkout- Types of AI


Where is Artificial Intelligence used?

Artificial Intelligence is reshaping our future and in the coming years, it will make our lives much easier than ever. These days AI is used to perform various complex tasks like in search engines, autonomous vehicles, healthcare, logistics, surveillance and many more.


1. Voice assistants- 

Google Assistant, Amazon Alexa, and Apple Siri are some of the voice assistants that use AI to decode voice inputs. These personal assistants use machine learning technology to analyze, understand and answer. Both Google Home, Amazon Echo use speech recognition and NLP to convert speech into desired actions.


2. Autonomous Vehicles-

Self-driving vehicles are those vehicles that have the ability to sense their environment and move slowly with little or no human intervention. These vehicles use sensors (radar, lidar, sonar, GPS, odometry and inertial measurement units) along with AI to interpret this data and accordingly give output. The most popular example of self-driving vehicles is Tesla, Waymo, and Roborace

Waymo`s AI system collects data from lidar to produce control signals that are able to operate the driverless vehicles.

Another famous autonomous vehicle manufactures "Tesla" uses its auto-pilot feature allows you to sit back and let the car drive by itself. If you`re not in a mood to talk, auto-pilot will check your schedule and drive you to your destination.

3. Cogito-

When any customer contacts the helpdesk, they want to get their problem solved as early as possible. Cogito uses Augmented emotional intelligence to elevate human experience and enhance interaction. The company makes use of machine learning and behavioral science to enhance customer interaction with the executive.

4. AI in artificial creativity-

AI is helping us in fields where no one would`ve ever imagined. Nowadays, we`re using AI in creating Music, artworks and much more stuff. MuseNet, a deep neural network takes the advantage of AI to generate 4-min musical compositions with 10 different musical instruments.

5. AI in Social Media-

When we`re talking AI, we cannot forget social media platforms like Facebook, Instagram, twitter who`re taking the full advantage of AI to add more capabilities in their platforms. Since the invention of these platforms, these are becoming our daily necessities. Sites like Facebook, Instagram process a large amount of data and it is not possible to handle this data manually. Facebook uses AI for face verification whereas machine learning and deep learning is used to detect facial features. Deep Learning extracts complex facial details from the image with the help of neural networks. Machine Learning is used to design feed based on your interest.

Twitter uses a mix of Machine Learning, Deep Learning and NLP to detect hate speech and to stop it from spreading. Twitter recently discovered and banned 300,000 terrorist-linked accounts in which 95% of the work was done by AI systems.

6. AI in Space Exploration-

This is one of the most interesting fields where we`ve used AI. Space expedition requires vast amounts of data to be analyzed. After long research, astronomers used AI to diagnose years of data obtained from the Kepler telescope to identify eight-planet in our solar system.

NASA is also using AI in its next rover mission to Mars. NASA`s AEGIS can handle cameras autonomously and can also choose what to investigate.

7. AI in Gaming-

The gaming industry is one of the fastest-growing industries in the world and AI has played a key role in its growth. Nowadays almost every game uses AI to enhance its gameplay.

DeepMind`s AI-powered AlphaGo software was the first of its kind to beat a professional human player. 

Another example in the gaming industry is FEAR, which is actually is a First-person video game but what makes this game more interesting is the use of AI in it. The actions performed by the opponent AI are designed in such a way that no action and mistakes are repeated. The system learns from its past experiences and gets better with time and use.

For more applications of AI, watch this from Edureka.


Conclusion:

Stephen Hawking once said "Success in creating AI would be the biggest event in human history. Unfortunately, it might also be the last, unless we learn how to avoid the risks".

That was all about AI and how does artificial intelligence work

Post a Comment

2 Comments