As a child Psychologist I always wonder… · How do young children learn so quickly and how does there brain works? · There reasoning power and questioning us? AND · Can we train computer machine or robots the same way? These are the questions that I am curious to know and I am not alone. The psychologists and now even computer scientist are observing and trying to know how young children can learn so much so quickly like alphabets, numbers, plants, animals, clocks, colors , shapes, etc, with such limited exposure in life. Now Computer scientist are creating an AI that will work as effectively as child brain is pretty exciting and challenging as well .It will require intensive training of the machine more then human, but will be worthwhile at the end to explore. Let’s understand how its done in layman terms. There are three machine learning strategies that we will see to train our AI . Supervised Learning Unsupervised Learning. Reinforcement Learning. As we parents and teachers supervise and teach our kids alphabets, shapes colors, etc, the same way we feed supervised data to our AI machine. In Supervised learning, the system is fed with the data, then the machine starts going through the fed data . The systems then makes predictions about what the new data should look like. It is like the systems revise its data to give us the output. In Unsupervised Learning we do not feed any trained or supervised data, or give any guidance to the computer. We do not give any directions. Here the computer has to figure out the hidden pattern, shapes, color, size and learn on its own. It is like we adults learning from our surroundings, the same way machine learns too. Reinforcement is like a child learns to walk on its own by crawling, falling, and taking few steps and adapting the pattern by hit and trial way, the same way machine adapts by reading its data and giving us the output. Hope you got some layman idea by reading my blog. Consider training AI computer machine as you would train and teach a preschooler.