Let’s first see what is computing, in general it can define as any goal-centered activityneed or creating computers.
So it includes designing and developing software/hardwaresystems for a goal, configuration and managing diverse kind of information ordoing scientific studies using computers.Computing can be conventional or intelligent.Simply we can say a conventional computing is based on particular sets ofinstruction (algorithm) to solve given problem and it can’t solve any problemfor which it haven’t instructions.
Whereas intelligence computing usually refers to the ability of the computer (machine) tolearn a definite task or solve a problem from data or trials and it can makenew algorithm to solve new problems. An intelligent system firstobserve its environment for collecting information using sensors or data givento it then plan how it will execute to solve the problem in best way andfinally it learn and update its information from its mistake in solving problemfor next solution. For example chess played by computer, in this it havelearned that much now that it can beat any human in chess and all this islearned by improving its moves from preceding mistakes. So to see if the given solution is intelligent or not allwe need to check is if it can learn from its previous mistake to solve newproblem on its on.
If it repeat the same mistake again and again then it is notintelligent.Also forchecking if the system is AI or not we have Turing tests but since now a day itcan easily be fooled we have some new methods to check AI, for example WinogradSchema Challenge. In this it presents anumber of multiple choice questions within specific format like: Q: The trophywould not fit in the brown suitcase because it was too big (small). What wastoo big (small)? ( form https://io9.
gizmodo.com/8-possible-alternatives-to-the-turing-test-1697983985)Answer 0: the trophyAnswer 1: the suitcaseThe answers to these questions seem easyfor human. But for answering this correctly requires skills that remain vaguefor computers, such as spatial and interpersonal reasoning or common sense. Agent is a computer program that executes variousactions autonomously and continuously on behalf of an individual or anorganization. For example, an agent may archive numerous computer files orretrieve electronic messages on a systematic schedule. There are manytypes of agents, two of them are utility based agent and learning agent. The utility based agents are used for achieving agoal and also to determine the degree of how efficiently our goal is completed. Whereas learningagent has a performance component that decides which actions to take and alearning component that can change the performance component to be more effectiveas the agents learns.
For example a car which is a utility agent driveby a human and then driverless car which uses learning agent and it drive itsown which doesn’t require any human help but this is still in experiment phase.