AI 写论文时的问题：面对审稿意见不会辩护，直接放弃
精选理由
Ethan Mollick 说 AI 回应审稿意见时总是直接认怂改稿，他提醒你得自己拿主意，该辩护的要辩护。
Ethan Mollick 指出 AI 撰写学术论文时的一个具体问题：当审稿人提出质疑时，AI 通常直接妥协修改，而不是尝试为自己的观点辩护。而同行评审的意义恰恰在于作者应先尝试回应挑战，只在确实错误时才放弃原有主张。因此用 AI 处理审稿意见时需要人工介入，保留一定的坚持立场。
原文 · Ethan Mollick
Subtle problem with AI-written academic papers is that AI isn't calibrated for feedback. AI often just folds, but if you are sending a paper for review, challenges from peer reviewers should result in some attempt to defend your idea (while still giving up on it if really wrong)
