THINK about it: you woke up this morning, checked the time, got ready while reviewing your day ahead and planning and prioritising your work.
You can do this seamlessly, without really analysing the process and wondering how that wonderful organ called your brain can string all these random processes together.
This ability to be given a large amount of information and to analyse the important connections between them based on past information, then reach a logical conclusion happens because of the many different layers of nerve connections in our brain.
Our neural networks are created and broken down as we go through our lives, depending on our experiences and knowledge.
Now, technology has progressed to a point where computer generated artificial neural connections are being used by electronic networks, recreating this awesome biological process in our brains.
These networks emulate a biological neural network of creating and deleting connections through learning and decision making based on previous knowledge and experiences, emulating our own brain!
Artificial Neural Networks (ANNs) gather their knowledge by detecting the patterns and relationships in the given data and learn through experience, not from human programming!
In 1943, Warren S. McCulloch and Walter Pitts published a research that replicated the human brain's ability to produce logical answers to complex patterns through connected neurons using binary logic system of 0/1 or true/false statements.
The first artificial neural network, called Perceptron, was invented in 1958 by psychologist Frank Rosenblatt.
In 1959 at Stanford, Bernard Widrow and Marcian Hoff developed the first neural network that was successfully applied to real world problems.
This system was named MADALINE after their use of Multiple ADAptive LINear Elements, and was specifically designed to eliminate noise in phone lines; it is still in use today.
By 1990's, Neural networks began to catch the imagination of scientists, finally developing into what is now called artificial intelligence (AI), which allows the computer to learn on its own.
Many of today's highest-performing AI systems, like speech recognisers and automatic translation, are produced using deep learning through parallel multilayered neural networks just like the human brain, which has been given the name ANN.
In the modern era, neural networks are assisting humans in education, medicine, finance, aerospace and other sectors. The first of these were in visual pattern speech recognition, and text-to-speech.
Automated and robotic factories are monitored by ANNs that control machinery, adjust temperature settings, diagnose malfunctions and more.
ANNs make successful stock predictions in real time with Multilayer Perceptron. Large financial institutions are using ANNs to analyse bond rating, credit scoring, target marketing and loan application evaluations, as well as credit card transactions to detect likely instances of fraud.
Facial Recognition Systems are used in crowd surveillance, airport check-ins and other areas where IDs are studied, comparing it with information present in its database.
ANNs are used to study the behaviours of social media users. Data shared everyday via virtual conversations is tracked and analysed for competitive analysis, linking it to people's spending habits, for targeted marketing.
In Aerospace engineering, ANNs are used in fault diagnosis, high performance auto-piloting, and securing aircraft control systems. Neural networks are also used in air and maritime patrol, and for controlling automated drones.
Healthcare: Convolutional Neural Networks are employed for X ray detection, CT Scan and ultrasound analysis. ANNs are used in Oncology to create algorithms that can identify cancerous tissue at a microscopic level with the same accuracy as trained physicians.
Many rare diseases that manifest in physical characteristics can be identified in their early stages by using Facial Analysis of patient photographs.
Personal Assistants: speech recognition that uses Natural Language Processing to interact with the users and formulate a response accordingly are used in Siri, Alexa and other personal assistant systems.
Weather Forecasting: traditional ANN and multilayer models can be used to predict climatic conditions 15 days in advance, with greater accuracy than human predictions.
The biggest disadvantage of ANN is that it doesn't have a clue of why and how, no idea about the implications of its answers. Perhaps a true "I don't care" unbiased report.
Artificial neural networks require very strong processors with huge processing powers and large storage spaces.
Technological unemployment: Will super-intelligent computers take over the planet one day? Or will robots help create a golden age of human leisure and prosperity?
I feel human skills can be complemented by automated processes, with machines taking charge of the routine aspects of the job; leaving humans with more time for face to face interactions and increased humane communication skills, so lacking in our interactions today.
The Fourth Industrial Revolution of AI may create new jobs that will require new skills and necessitate significant investment in upskilling and reskilling young people and adults.
Today's companies need to invest in developing their employees' soft skills that ANNs cannot replicate.
In this changing world, the value of creativity and emotional intelligence will increase, and leveraging these qualities into organisational cultures is critical to enable workers to remain meaningful in an AI world. The future is now.
The writer is Assistant Professor at the Faculty of Medicine, AIMST University
The views expressed in this article are the author's own and do not necessarily reflect those of the New Straits Times