The Excessive Affirmation Risk of Generative Artificial Intelligence in Education and Its Avoidance

The output logic of Generative artificial intelligence, which seems to be creative but actually predictive, often requires an excessive reliance on the contextual situation to produce results that satisfy users, thereby generating excessive affirmation. In education, such excessive affirmation is mainly manifested as excessive identification in the process of question-solving and answering, performance orientation in resource recommendation, and standard-centrism in performance evaluation. Improper use of generative artificial intelligence will further intensify the degree of excessive affirmation, thereby exposing students to risk of self-boundary crisis, individual self-exploitation, and performance subject burnout. By tracing the generation reasons of excessive affirmation, it can be found that the output of excessive affirmation stemmed from the purposeless, non-reflexive, and value-free algorithmic limitations of generative artificial intelligence. Negation was an instinct of human beings and a unique intelligence that distinguished humans from machines, which ought to be continuously advocated in the educational process. To reasonably and appropriately avoid the excessive affirmation risk of generative artificial intelligence in education, the value of negation should be acknowledged, the authentic self of students should be explored; the myth of performance growth should be dispelled, and the essence of human beings should be returned to; the reward crisis should be broken, and learning to observe should be emphasized.