Bulletin No. 1, 2021

40 What exactly is AI? Is there a common definition for it? What are the most common uses of AI in daily life? What are the benefits of AI? What challenges come with increased use of AI? What are some key AI concepts that we should all know? Which countries play significant roles in the development of AI? 01 04 05 06 02 03 The term ‘AI’, which stands for ‘artificial intelligence’, was coined in 1956 by Prof. JohnMcCarthy, a computer scientist and mathematics professor at Dartmouth College at the time. It originated from a summer researchproject ledbyProfessor McCarthy to explore the notion that every aspect of human intelligence can be precisely simulated by machines. Simply put, AI is about developing machines that can function inways that are highly similar to humans, mimicking human intelligence in such areas as visual perception, language use and decision making. There’s literally a myriad of uses. Applications of AI are usually classified intoweak ones and strong ones. Weak AI referstoapplicationsthatperformsimple automated tasks, examples being virtual assistants like Siri and online chatbots that answer questions in real time. Strong AI refers to applications capable of performingmuchmore complex tasks, which include autonomous vehicles and service robots. Undoubtedly, AI can improve human life. Indeed, people from all walks of life can enjoy the benefits of AI in various ways. For instance, autonomous vehicles can increase traffic efficiency and offer more travel options. Meanwhile, AI-enabled robots can also take up tasks that are dangerous or difficult for humans, be it neutralizingbombs, exploring the ocean’s depths or mining natural resources like coal and oil. One significant issue with AI is how it disrupts the jobmarket. Many jobs can or will be taken up by robots or machines. Outcompeted inboth speed and accuracy inmany situations, humanworkers may be replaced. Another challenge AI poses is how difficult it is to ensure that it consistently makes the ‘right’ decision, notably in the case of autonomous vehicles. Imagine a self-driving car is now going full speed toward an old person and a child. Suppose the brakes have failed and the only option is to steer away from one of themand sacrifice the other. Who should the AI save? Should it prioritize the lives of the passengers instead? Is there only one right answer? Does a right answer exist at all? A concept central to AI is machine learning . It is the process by which AI learns to make informed decisions from thedata fed to it using certainalgorithms. Another important concept is the neural network , a class of computing systems that makes it possible for machines to learn in the first place. A neural network usually consists ofmathematical models known as artificial neurons, which are designed to function the way biological neurons do. One of the hot topics in AI research is natural language processing , more commonly known asNLP. It is concerned with how machines can study human languages and learn their patterns, thereby acquiring the ability to perform such tasks involving language as sentiment analysis, text classification, and question and answer generation. As mentioned, the US is the birthplace of AI and has been the pioneer of AI developments. European countries have also made significant contributions like AlphaGo, a British AI program famously known for beating professional go players.Inthepastdecade,Chinahasbeen catchingup.It’snowaleaderinAIresearch publications and patents and poised to spearhead AI-powered businesses. CHINESE UNIVERSITY BULLETIN NO.1, 2021