One of the essentials of good research is training your own communication skills—being able to elicit a response from the environment and have others recognize you. For readers, typically they start from the beginning of an article; if a sentence doesn’t interest them or feels off, they begin to criticize or abandon reading. For researchers in the AI era, AI cannot replace human decision-making. If all your research experience comes from what AI tells you, you have no judgment, and then you stubbornly insist on your own view, you will naturally be more Lost. Moreover, all fragmented copy-paste work has no value; work that has not been digested by humans and re-expressed is hard to contribute to innovation.
Researchers should build a self-centered system that filters and organizes the problems they want to solve under a self-centric mechanism. Specifically, during conversations with others, you should be able to express your own understanding and requests simply, concisely, and clearly, view things from the other party’s perspective, capture the other party, and consider their limited patience and bandwidth — you cannot send to collaborators the results of your unfocused, headless AI conversations. In concrete field research and learning, you should be able to think clearly yourself and re-describe the clues in your own words — you cannot treat AI-organized fragments as knowledge you have already indexed into your knowledge system. It’s especially important to be wary of how many papers you have read that AI didn’t read, how much time you spent thinking in depth, and how many answers weren’t AI-guided.