UNDERSTANDING HOW THE HUMAN BRAIN AND NEURON STRUCTURE IS THE BASIS OF ALL ARTIFICIAL INTELLIGENCE .

“FOOD HAS NO IMPORTANCE WITHOUT HUNGER ,”RIGHT?”FOOD HAS NO IMPORTANCE WITHOUT HUNGER ,”RIGHT? BEFORE GETTING TO DAUNTING TERMS LIKE ALGORITHMS, ARTIFICIAL INTELLIGENCE, MACHINE LEARNING, DATA, LETS ADDRESS THE VERY NEED OF UNDERSTANDING NEURAL NETWORKS. WERE THE TRADITIONAL COMPUTERS NOT GOOD ENOUGH? AND WHY DO MACHINES NEED TO “LEARN”?

“DECISION MAKING” IS SOMETHING THAT YOU DO SO OFTEN IN EVERYDAY LIFE,THAT SOMETIMES YOU DON’T EVEN REALISE IT.

LIKE YOU DECIDED THIS READ A BLOG ON NEURAL NETWORKS/AI .YOU DECIDED TO USE A CELLPHONE OR COMPUTER FOR THE SAME. AND HOW GOOD A PERSON’S DECISIONS ARE DEPENDS ON HIS INTELLIGENCE , WHICH IN TURN DEPENDS ON HIS ” LEARNING EXPERIENCE” IN LIFE. AND WE ARE GETTING LAZIER DAY BY DAY. WE WANT ONLINE RECOMMENDATIONS WITHOUT DOING ANY RESEARCH, WE WANT FINGERPRINT RECOGNITION IN CELL PHONES INSTEAD OF WRITING THOSE LONG PASSWORDS, OUR PHONES TO RESPOND TO OUR VOICES .WE WANT CARS TO DRIVE ON THEIR OWN.

SO THIS “LAZINESS” IS DEPENDING ON SOMEONE ELSE FOR TAKING DECISIONS ON YOUR BEHALF. AND IN THIS CASE MACHINES. YOU AS A HUMAN USE YOUR BRAIN ,YOUR PAST EXPERIENCES AND YOUR DEVELOPED “INTELLIGENCE” TO TAKE A CERTAIN DECISION. HOW DOES A MACHINE RELATE TO THESE TERMS. THIS IS WHERE THE TERM “TRAINING ” COMES IN. JUST LIKE AN INFANT’S BRAIN IS TRAINED BY ALLOWING IT TO LEARN FROM ITS ENVIRONMENT, BY SCOLDING IT WHEN IT TAKES A WRONG DECISION. AND THEN WHEN ITS ALL GROWN UP, ALLOWING IT TO TAKE ITS OWN DECISION.

UNDERSTANDING “INTELLIGENCE”

NOW LETS ADDRESS THE QUESTION AS TO WHY SUCH ARTIFICIAL INTELLIGENCE IS REQUIRED WHEN WE ALREADY HAVE ADVANCED COMPUTERS IN HAND. THE THING IS THAT YOU CANNOT HARD WIRE A COMPUTER PROGRAM TO DECIDE WHETHER A CERTAIN USER WOULD LIKE A CERTAIN MOVIE ,JUST AS YOU CANNOT HARD WIRE A PERSON’S VOICE COMMAND . AS A RESULT WE NEED TO FIND A DIFFERENT APPROACH TO PROBLEM SOLVING.

IN SUCH CASES THE MACHINE MUST LEARN FROM ITS ENVIRONMENT AND “EVOLVE”. AND WHAT BETTER A ROLE MODEL THAN THE VERY HUMAN BRAIN WHEN IT COMES TO LEARNING . OUR BRAIN’S BASIC STRUCTURE CONSTITUTES OF CELLS CALLED NEURONS, THERE ARE APPROXIMATELY 86 BILLION NEURONS IN THE HUMAN BRAIN . THESE NEURONS WORKING TOGETHER CONSTITUTE OUR “INTELLIGENCE” AND “REFLEXES”.

NOT ALL NEURONS ARE RESPONSIBLE FOR A CERTAIN FEELING/RESPONSE. A PARTICULAR PORTION OF THE BRAIN IS MORE ACTIVATED CORRESPONDING TO A PARTICULAR STIMULUS. CERTAIN WORDS MAKE YOU HAPPY ,CERTAIN SAD, YOUR BRAIN HAS LEARNT TO RESPOND TO A CERTAIN SIGHT. THIS IS THE BASIS OF REPLICATING AN ARTIFICIAL “NEURON” WHICH IMITATES THE LEARNING PROCESS OF A BRAIN.

JUST AS YOUR BRAIN HAS A TREMENDOUS NUMBER OF NEURONS ,A COLLECTION OF SUCH “MATHEMATICAL NEURONS IS CALLED A NEURAL NETWORK. AND THE DECISION MAKING CAPABILITY THAT THEY ATTAIN AFTER BEING TRAINED IS TERMED A ARTIFICIAL INTELLIGENCE.

HERE BELOW YOU CAN SEE ONE SUCH NEURAL NETWORK, THE BLACK DOTS CORRESPOND TO CERTAIN NUMBERS AND THOSE MESSY LINES ARE KNOWN A “WEIGHTS” AS THEY SHOW HOW IMPORTANT THE PARTICULAR INPUT IS FOR DECIDING WHETHER A NEURON HAS TO BE ACTIVATED OR NOT.