Common, Uncommon Sense: Visiting the Human‐Versus‐AI Argument
通过系统回顾1996-2024年AI文献，讨论人类与AI之争，指出AI虽高效但缺乏可解释性，未来人机可协同工作。
ABSTRACT Contemporary developments have empowered artificial intelligence (AI) to take a generalist approach: a single AI like ChatGPT can now solve problems from diverse fields. Here, we discuss the long‐debated human‐versus‐AI argument through a systematic review of AI literature from 1996–2024. While the strengths of AI are undeniable—it can perform various tasks from different bodies of knowledge more accurately and efficiently than humans—its output is hindered by the explainability of its information processing mechanism also known as explainable AI (XAI). This shortcoming of AI has reduced its feasibility. In the future, humans and AI can work synergistically: AI can support humans by doing tiring work, while humans assume the role of expert by maintaining AI performance and approving (and/or improving) its output.