Progressing AI with Neural Networks: A progress towards the benefit of mankind or its disruption?

Source: Deep Learning on Medium

Artificial Intelligence or AI is the current “HOT TOPIC”. Everyones buzzing about it, some in favor and some in against. And those in favor are constantly working to take it to a level beyond imagination.

In the current scenario, the so-called Turing test is defined as the qualifying test for a robust and efficient AI model. A Turing Test is a method of determining whether or not an AI model is on par with a human mind. This test is named after Alan Turing, the founder of the Turning Test and an English computer scientist, cryptanalyst, mathematician and theoretical biologist. A machine or an AI model that passes this test is said to be on par with a human mind.

The foundation of any AI model is the machine learning subset. A robust AI model has a robust ML subset that constantly updates its dataset and ongoing algorithm.

Neural Networks: The fuel of AI models

There have been several recent advancements in the field of computing infrastructure, storage and computation power that has led to the evolution of ML into a more complex model of Deep Learning(DL), Generative Adversarial Network (GAN) and Reinforcement Learning (RL), all with the help of Neural Networks. Neural Networks are some very specific and efficient algorithms that can differentiate and make judgments on an image or facial patterns. They are called so as they mimic the neurological functions of a neuron in the human brain and the study of ML models which imply Neural Networks on the complex datasets is known as Deep Learning.

There are two well-known types of Neural Network models that serve the purpose of mimicking a human brain. These are the Convolutional Neural Network (CNN) model, which is widely used in different image related applications like autonomous driving, robots, image search, etc., and the Recurrent Neural Network (RNN) model, which is the base of most of the Natural Language Processing-based (NLP) text or voice applications, such as chatbots, virtual home and office assistants and simultaneous interpreters.

Now, this has worked as a fuel for AI models. Neural Networks have helped AI models to perform some mindboggling tasks like image processing, facial recognition, auto-driving, NLP, etc. Such AI models can be build using some open source libraries such as TensorFlow and Keras

NLP or Neural Language Processing is an algorithm which is directed to work on voice and word-based speech recognition.

Real-world applications of Neural Networks and some views on it

All these real-world applications like self-driving cars, facial recognition, etc. are obviously viewed as a boon for mankind by every individual.

But, the biggest successful application of Neural Networks in an AI model in the real-world is ‘ Sophia the robot’.The bot was dreamed up by the brains at Hanson Robotics, lead by AI developer David Hanson. This was the point where a little fear of AI raised in the minds of some.

“ With Artificial Intelligence we are summoning a demon” — Elon musk

But actually that’s not the case. If regulated by the Human Mind, it’s will always be a boon to mankind.