When scientists first came up with the concept of Artificial Intelligence people thought that they would build machines that can imitate human beings. But it took around five decades for them to make a huge change in the field of artificial intelligence as we see today.
We humans have cognitive abilities that is what makes us different. Some of the capacities include sustained attention, selective attention, divided attention, long-term memory, working memory, logic and reasoning, auditory processing, visual processing etc.. These are brain based abilities that help us to carry out any task from the simplest to the most complex ones. In order to make machines do task like humans we need to develop cognitive skills in machines. This is why researchers are working hard to find a complex design which can make machines do tasks like humans.
One way to develop cognitive abilities in machines is to recreate deep neural networks but this is a great challenge in the field of artificial intelligence. For years scientists have been researching on human brain to understand how it functions, still there are some unsolved puzzles about human brain. Expanding neural networks will help machines to emulate humans as well as develop cognitive skills. Scientists are inspired by specific cognitive abilities of humans so they are implementing algorithms for the recreation of cognitive abilities in machines. Also this is producing promising results.
Along with all these researches too the interaction between machines and humans is remarkable. Watson, Siri, Cortana are best example for this. In order to make a transformational effect on artificial intelligence, artificial neural networks should be used by the support of human native intelligence. Anyway it is a huge challenge for scientist to replicate neural networks in machines since technology need time to get developed. It took decades for advancing human brain in overcoming survival instincts, utilising intellectual abilities etc.. The evolution or transformation of cavemen to people thinking about getting humans to Mars was slow. When a brain is finding way to overcome our physical limitations or to expand ideas in computational methods, statistical methods the pursuit of science gets fuelled.
It is a difficult process to teach machines but we can strengthen their cognitive abilities by introducing algorithms in them. However, there is fear of distortable future in one side. People are curious about machines with cognitive abilities. When the concept of brain computer interface came the curiosity got accelerated. By the introduction of start-ups like Neuralink (a seamlessly implanted device which help to solve important brain and spine problems), by Elon Musk and Kernal, people are hoping that machine capabilities can be enhanced through interlinking. Exploration of human cognitive abilities will help in the formation of artificial intelligence brain.
The three important human cognitive abilities are:
1.Thinking
A machine get distracted whenever a sound is heard or a new thing is found other than its knowledge. Similarly we humans get distracted to certain things in our surroundings. Also we have the ability to stay focused on things for a long time and only get distracted if necessary. But in case of machines they cannot forget distractions and continue their work. This attention mechanism of humans motivated scientists. They are strengthening deep generating models or convolutional neural networks through deep learning methods.
2. Memory
For humans there is no need to put effort to find solutions since we have ability to understand conversations (speech recognition) and recognise objects (object recognition). These abilities work continuously so there is no need to spend time to think or arrive at solutions. But in the case of machines it is difficult for them to arrive at a solution. A theoretical model of the human brain can be built which helps to study functions like vision, motion, sensory control learning etc.. This will help to create a memory information process, object- speech recognition capability in artificial brain.
3.Reproducing sense
Object recognition is difficult for machines.
In order to make them recognise we need to fed them with similar data. For example, to make it
recognise a monkey we need to show it thousands of images of a monkey. Thus by
giving similar data's, a machine become able to recognise objects whether it is
placed at different angles or in different lighting conditions.