Problems and Algorithms:A Study on the Limitations of Artificial Intelligence from the Logic of Discovery
Proposing and solving questions are pivotal to scientific discovery and knowledge development.Currently,there is limited exploration in the difference between artificial intelligence and human cognition in the realms of"questioning"and coping with problems.Therefore,it is necessary to exam-ine the connection between"questions"and"algorithms".Upon examining the processes such as the formation of scientific questions,coping mechanism,and theory development,it reveals that AI struggles to pose creative scientifc questions.Vice versa,scientific problems are challenging to resolve solely through algorithms.The primary basis for this lies in phenomena where assumptions precede observations in certain significant scientific discovery proces-ses,and the processes of problem formation and cognitive leaps are often non-inductive and non-algorithmic.Therefore,concerning the cabability for scientific discovery,algorithm-based AI may encounter significant limitations.
万方数据
