Hi, just some comments to not discuss this topic to death: Copyright exemptions ================= There are some areas which in most regions are exempt from copyright: Legal contracts are one area. Also we do not want the people submitting a paper with a new wording, prevent any other competing paper to use a standard wording: E.g. my paper suggests adding [[safe]]. But I want to have copyright on that attribute. So any competing paper, even or if mine is not accepted, may not reuse [[safe]], because it is my copyrighted attribute.   AIs trained on papers to generate papers ============================= For AI use, what some others probably do not want at all:     An AI trained on their papers, and then generating new subpar papers. AI, please identify the 10 most imminent missing features in C++ and write papers with at least 20 pages each. At least currently, this would only generate added effort for the committees, as AI is not good enough for such complex tasks (at least if it is about complex features).   On the other hand, paper search engines or automatic feedback for self-written papers can even improve the situation for the comittees. AI, I have written a paper, please provide initial feedback so that I can answer potential questions, before submitting it. As long as the human submitter is knowledgeable about the area and does not just copy AI output as paper without selecting or understanding.   I have used AI in the past to write and improve technical or marketing texts. You need about 10 to 30 iterations for each sentence and take a long time with some manual editing, some discussions and brainstorming (with AI use) for a really good outcome, but then the result can be better, as if I authored it by myself. But that takes a lot of time and you must have full expertise in the subject matter to know what you want to say, and how the current text could be understood by different readers. I have also used AI for unimportant stuff, then you are faster than doing it without and just have to give some rough guidelines and read through it. But that would not be the quality expected from a paper for C++.