“MACHINE LEARNING” AND “DEEP LEARNING,” ARE COMMON TERMS THROWN AROUND AND ARE SOMETIMES USED INTERCHANGEABLY WITH ARTIFICIAL INTELLIGENCE. THESE THREE TERMS MAY AT REMS BE VERY CONFUSING.

THE DIFFERENCE.

Artificial Intelligence a term first coined in 1956 by John McCarthy involves machines that can perform tasks that are characteristic of human intelligence. In general it includes things like planning, recognizing objects and sounds, understanding language, learning, and problem solving.

AI can be broadly categorized as general and narrow. General AI basically includes all of the characteristics of human intelligence, including those mentioned earlier whereas narrow AI exhibits some aspects of human intelligence, and can sometimes do extremely well in some aspects but found wanting in others. An example of narrow AI is a machine that’s great at recognizing images, but nothing else. Machine language can be seen as a way of achieving AI.

Arthur Samuel another scholar in 1959 opined that AI was the ability to learn without explicit programming. He saw that you can build AI without using machine learning, but this would require building millions of lines of codes with complex rules and decision-trees.

Instead of using this tiresome method, he saw machine language as a way of “training” an algorithm so that it can learn how to execute a certain set of instructions. This “training” would involve keying in huge amounts of data to the algorithm and allowing the algorithm to adjust itself and improve.

For example, computer vision has been improved drastically by use of machine learning. Computer vision in essence is the ability of a machine to recognize an object in an image or video. Here hundreds of thousands or even millions of pictures are gathered and then humans tag them. For instance, the humans might tag pictures that have a cat in them versus those that do not. Then from a previous experience, the algorithm tries to build a model that can accurately tag a picture as containing a cat or not as well as a human. The machine can be thought of as learned once the accuracy level is high enough.

A MAJOR APPROACH TO MACHINE LEARNING IS DEEP LEARNING.

Some other approaches include decision tree learning, inductive logic programming, clustering, reinforcement learning, and Bayesian networks etc.

The functioning of the brain, particularly interconnection between a number of neurones inspired deep learning. Artificial Neural Networks (ANNs) are algorithms that try to emulate the biological structure of the brain. In an ‘ANN’, there are “neurons” which have discrete layers and connections to other “neurons”.

AI AND IOT ARE CO-DEPENDENT.

The relationship between AI and IoT can be compared to that of the human brain and body. The body collects sensory input like sight, sound, and touch and our brains take that data analyse it and makes sense of it. It turns light into recognizable objects and sounds into understandable speech. Decisions are then made and sent as signals back out to the body to command responses like picking up an object or speaking.

The connected sensors in this case can be thought of as an Internet of things. This ‘internet’ provides raw data on what is going on in the world. Our brain acts as the Artificial intelligence, making sense of that data and deciding what actions to perform. And the connected devices of IoT are again like our bodies, carrying out physical actions or communicating to others.

IN EXTRICATE CO-DEPENDENCY.

As mentioned earlier AI and IoT are dependent on each other. Great strides have been taken for AI because of Machine learning and deep learning. Basically IoT makes better AI even though machine learning and deep learning require a lot of data to work, and this data is being collected by the billions of sensors worldwide.

A lot can be accomplished by improving AI as it makes IoT useful. This can be achieved by adopting the Internet of Things and creating a virtuous cycle in which both areas.
Industrially, AI can be used to determine maintenance days for machines or to analyse manufacturing processes to make big efficiency gains, saving loads of money.

We can make technology adapt to us rather than us adapting to it. Voice commanding an b used to tell the machine what we want rather than clicking, typing, and searching. Things like asking for the weather or tasks like preparing the house for bedtime (turning down the thermostat, locking the doors, turning off the lights, etc.) can be simplified.

CONVERGING TECHNOLOGICAL ADVANCEMENTS HAVE MADE THIS POSSIBLE

Manufacturing and maintenance costs can be cut in half by shrinking computer chips and improving manufacturing techniques. This will not only cut the costs but will provide much more powerful sensors. Revolutionizing battery technology means those sensors can last for years without needing to be connected to a power source.

Smartphones enable us to send large volumes of data at cheap rates using wireless connectivity, allowing all those sensors to send data to the cloud.

The invention of clouds have enabled us to have unlimited storage of that data and virtually infinite computational ability to process it.

Obviously there are some concerns about the impact of AI on our society and our future but as advancements and adoption of both AI and IoT continue to improve, one thing is certain; the impact is going to be profound.