AI – Ushering an Era of Intelligent Machines – Part 1

Intelligence is a highly coveted skill in humans. It is what distinguishes humans from other animals. This smartness has enabled human civilization to create a life of luxury by taming the brute natural forces.

As opposed to us, machines seem stupid. They need to be told what, how and when to do. That is an intimidating amount of effort, right? After all, machines were invented to reduce human labour, not to employ us in their upkeep. This exasperation with machines brought about the idea of intelligent machines. It led to the creation of artificial intelligence (AI), that focusses on creating machines with human-inspired cognitive abilities of problem solving, decision-making and reasoning. 

AI – All Around You

As your day begins, there are multiple decisions to be made. Hence, AI-enabled decision making is ubiquitous. You make use of multiple AI applications throughout your day. Some of them use AI for completing their tasks – like vacuum cleaners and washing machines, while some others act as your Digital Personal Assistants (for e.g. – Siri, Alexa, Google Home). Social media apps such as Instagram, WhatsApp, etc. use the AI algorithms to provide users with a better experience. Video-sharing platforms like YouTube and video editing tools like InShot or Adobe Photoshop also use AI and Machine Learning (ML) models for added features – for example – video recommendation system and photo background remover respectively.

How does the Artificial Intelligence technology work?

AI is called ‘artificial’ because the brilliance of a machine is programmed into it ‘artificially’ as opposed to the natural intelligence that characterizes animals. Hence, AI can be defined as a collection of computer softwares and algorithms that bestow a machine with appropriate intelligence which allows it to complete the given tasks. To illustrate, an AI-enabled washing machine will determine the wash-cycle time, the amount of water and detergent needed and the number of rinse-cycles, after assessing the extent to which the clothes are soiled. This shall automate your task of doing laundry while also saving detergent, water, electricity and cost. Isn’t it a win-it-all situation?

The theoretical concept of machine intelligence is manifested, in its most basic form, by providing the machine with an advanced computer brain. This brain is programmed to sense its surroundings and take an action from a pre-set list of instructions. An advancement over this simplistic AI utilizes the data in order to ‘memorize’ and ‘train’ itself. Data required for this machine-training can either be a pre-fed dataset or data points acquired from its surroundings. This exercise ensures that the next time the machine encounters a situation similar to the one it has been trained for, it is ready with a sequenced list of actions to be performed. This removes the need of a human operator, thus realizing automation. The process of ‘learning’ needs large amounts of information to train our machine. Because data is the backbone of any AI model, it is said – ‘data is the new oil’. As the discovery of oil accelerated the first Industrial Revolution, the collection and analysis of data has ushered in Industrial Revolution 4.0.

Another application of artificial intelligence is in making predictions regarding an entity. The only prerequisite of these prediction models is the availability of adequate data to study past patterns and project the future. This is similar to how a human brain utilizes past memory patterns as well as ambient conditions, in order to decide or predict a future event. For example – YouTube or Facebook recommendation systems analyze the videos and channels you have watched and liked in the past so as to recommend a new list of videos for you.

Evolution of AI

Like many of the modern technologies, AI was at first a figment of imagination. Playwright Karel Capek and novelist Mary Shelly are some of the writers who first created artificially intelligent beings, in the early 19^th century. During the succeeding centennial, Isaac Asimov, the most celebrated sci-fi writer, explored the ideas of artificial intelligence and machine learning (a subset of AI) through his writings. 

As the 20th century progressed, there were leaping advancements in the computing logic, led by Alan Turing’s ‘Theory of Computation’ It is a branch of computer science that aims at solving mathematical problems by using computing models. Coupled with pathbreaking research in neurobiology, information theory and cybernetic loops, researchers fathomed the possibility of creating an ‘Electronic Brain’. What began with the development of ‘Artificial Neurons’ in 1943 by Warren McCullouch and Walter Pitts Jr. in a lab, has come a long way since then with multiple AI products in use today.

Subsets of Artificial Intelligence Technology 

1. Machine Learning (ML) – 

This the branch of AI that can identify and analyze patterns with the help of numerical data inputs in order to make decisions. ML models make use of past trends to make conjecture about future events. For example, the forecasts made regarding the rise of Covid-19 cases during the pandemic employed ML models. The number of past cases, population density, average immunity of the populace, penetration of vaccination amongst the people, behavior of the virus, total deaths in the previous pandemic, etc. were some of the parameters used to prophesize the infection curve.

2. Deep Learning –

It is considered to be a subset of Machine Learning. It mimics the behavior of a human brain, though the tech is far behind in matching its capacity. Like a human brain learns from examples and experiences, so do the deep learning models. These models possess the ability of self-training through capturing data from their surroundings. Opposed to this, ML models require inputs of training datasets which contain large amounts of data points. Higher the amount of data used in training, better is the output of an AI algorithm.

3. Natural Language Processing (NLP) – 

Use of language comes naturally to humans, unlike for machines. When we speak to Amazon Alexa, its processor goes through a set of complex computing, in order to decipher what was said and then acts accordingly. Processing of speech is a highly complex phenomenon for a machine because of the presence of a large number of languages, their variegated accents and multiple pronunciations. NLP models require humongous amounts of data training in order to get a satisfactory output. Hence, NLP embodies highly complex algorithms to recognise the spoken word (speech recognition), make sense of the speech (natural language understanding) and respond back effectively (natural language generation).

4. Computer Vision (CV) – 

When human eyes see an object, the neural networks in the brain instantly recognise it. What is such an effortless response for us, turns out to be strenuous task for the machine. This is remedied by making use of Computer Vision softwares. In order to allow Image Recognition, the software is trained using a great number of photographs, of the object, taken from different angles. The computer then remembers these photographs and later makes use of them to recognize the object. This is done through memory scanning and finding the nearest match. Going farther than image recognition, computer vision models also derive meaningful information about the object so as to allow decision-making and action-taking. The most notable application of CV algorithms is in Self-Driving Cars – for recognizing vehicles, human beings, trees, traffic lights and other objects that are commonly encountered on roads.

Is AI necessary?

Technology skeptics and conservatives, especially in India and other Third-World countries, question the diversion of already scarce resources towards development of newer technologies. Their resistance stems from misinformed reports of AI-caused unemployability and the belief that technology benefits only the rich. Though these points sound logical, they are far from the truth.

Currently, the global market of AI is valued at 136 billion USD. With a compound annual growth rate (CAGR) of a whopping 38% during the period of 2022 – 2030, it becomes impossible to ignore AI. Any economy that aims at thriving and growing, can turn a blind eye towards novel tech only at its own peril. 

With the induction of Synthetic Party, a political entity whose policies are determined by AI, the technology has entered the political sphere too. In such times when the whole world – including businesses, governments and consumers – are adopting artificially intelligent solutions to daily problems, India and other lower income nations cannot lag behind. It is imperative for the third-world nations to innovate AI-enabled systems that will help in alleviation of issues like poverty, hunger and malnutrition, crippled healthcare, inadequate infrastructure, etc. Hence, in the foreseeable future, innovation and incorporation of tech-based solutions that are aimed at benefitting the masses as well as the environment is the need of the hour for all of us.

In the Part-2 of this article ‘AI – Ushering an Era of Intelligent Machines’, we shall address some of the concerns raised against like AI replacing jobs, whether intelligent machines can annihilate humanity and how you can manipulate technology for your own benefit. Stay tuned for next Thursday.

Vishvali Deo

Vishvali Deo is an E&TC (Electronics and Telecommunication) Engineer by education and Software Engineer by Profession. She believes that 'Technology is a Great Democratising and Equalising Force' and hence is on a mission to make the general public understand seemingly complex technologies in a simple manner.

She is convinced that the root of today's world problems lie in the past, hence she has also pursued post-graduation in History. She has a keen interest and a good grip over Economics, Political Science and Environmental Engineering. She has a penchant for working with Women and spreading Digital Literacy amongst them, with the aim of their empowerment. She also strives to provide Free Quality Education to children and counsels young adults. Besides, she is also skilled at Public Speaking, having won many awards in Elocution & Debate Competitions and Technical Paper Presentations.

• Optical Fibres: The Tubes That Bend Light
• Data Science: The Hottest Field of 21^st Century
• Big Data: Powering the Age of Information
• Seeking answers from ChatGPT: Humans & Artificial Intelligence (Part IV)