In AI era, natural language processing is one of several important research fields of artificial intelligence. In theory, linguistics can make a great contribution to the computer processing of natural language, but in fact it is not. There are three main reasons for this situation: First, the main research results of linguistics are lack of the maneuverability and correspondence to solve practical problems in terms of the technical processing requirements of natural language. Second, linguists don’t have an in-depth and comprehensive grasp of the structural rules and operation mechanisms of natural language yet. Third, linguists’ exploration of language meaning often lacks depth and insight. If linguists want to make a great contribution to the study of artificial intelligence, especially the computer processing of natural language in AI era, they must give full play to their strengths and circumvent their weaknesses, and come up with theoretical explanations and theoretical systems about the generation, structure, hierarchy and understanding of language meaning that can meet the needs of artificial intelligence research. In AI era, not only meaning should become the most important research object of linguistics, but also the urgency of pragmatic meaning research is steadily on the increase, the pragmatic meaning should be put in the core position of meaning research. In a period from1980s to the end of last century, Chinese linguists have conducted characteristic researches on pragmatic phenomena and pragmatic meaning. But unfortunately, after entering this century, these researches have not been carried forward. In a word, linguists must conscientiously sum up the past positive and negative experience, grasp the correct research ideas, find a suitable research direction, and adopt accurate and effective research methods. Only in this way, can they make a great contribution to the study of artificial intelligence.