Conventional and Intelligent ComputingConventional computingConventional computers follow a set of instructions and accordingly process data.Major components include memory, processing, and bandwidth.It is a machine that can do anything we like if given a set of instructions called programs. They store numbers in memory and perform operations on numbers stored in memory. Example adding and subtracting. They can perform complex tasks by executing series of instructions called algorithms.These computers calculate using circuits that are called Logic gates.
Logic gates compare patterns or bits stored in temporary memory that are called registers. The limitation to conventional computing is that it depends on these conventional transistors, ie. the more memory we require more is the number of transistors that we need to process 0’s and 1’s and to perform more complex operations.Intelligent computingUnlike conventional computing that uses set of instructions or algorithms to perform operations, intelligent computing can also compute images and concepts and process them as information.It focuses on attempting human-like intelligence through symbols or speech or the ability to make decisions based on certain situations. By analysis of these symbols, images and using various methodologies for making a decision.Its main principles include :Fuzzy LogicNeural NetworksEvolutionary ComputationLearning TheoryProbabilistic MethodsWhen is a solution Intelligent?When it is the outcome of various factors such as its ability to interact with real-world problems, Reason and Plan, Learning and Adaptation to the given problems and situation. And if the solution solves the problem and adapts to the changing environment of the problem and gives a better solution to the new problem.
AgentAn Agent is anything that helps perceive its environment through sensors. Used to obtain useful information through information gathering and exploration.Utility Agent: Estimates how close the solution to a given state is and is aware of the function to be performed.Learning Agent: Agents that can observe and take actions. Uses perception of its environment and takes decisions.Learning agent uses sensors to perceive its environment and make decisions based on its situation, whereas a utility agent’s outcome is fixed based on its current state.
Therefore learning agent is better since it adapts to a changing state or environment. Which is better?Intelligent computing is better since its solution is a result of deeper analysis than a conventional system whose outcome only follows a given specific set of instructions.Example, using conventional computation to play a game of chess would mean to use only specific instructions to play a move irrespective of environment change whereas intelligent computing analyzes the current situation and decides what best move to play based on the changing nature of its environment.ConclusionSmart/Intelligent computing has a bigger scope in the future than its conventional counterpart, with advancements in AI, Intelligent computing is the need of the hour to see more innovations in various fields such as medical, transportation, robotics etc which cannot be achieved through conventional means of computing.