There should at least be a code of professional conduct where authors state the extent to which LLMs were used. (This would also help not wasting time by asking some “authors” about “their” paper.)
Journals themselves should make policies about the extent to which they allow the use of LLMs. In some areas it might be considered more benign than in others.
There is. NLP conferences, like the ACL family (and the ARR) require you to disclose in the paper if you used LLMs for writing and coding, which ones and how. Whether every author is honest is a different matter.
Not sure if you're just talking about past design, but currently you can use arrow keys and enter in the menubar, tapping first letter of options to hops there, and can activate menubar with a keyboard shortcut of your choosing. You can also hop directly to help, start typing and get matches to select, all without a mouse.
Yes! I did this and love it. My layouts are heavily customized so no standard painted keycaps would do me any good anyway. I use a coherent set of key colors on various layers to provide reminders of what various key clusters are about. Looks great too.
You’re a ”developer“, i guess, but not a coder (anymore), which is what your interlocutors are probably asking about. You’ve migrated to a middle manager job, not something they probably can just start doing competently. Essentially you’re agreeing with their initial sentiment, that coders will be made irrelevant.
This brings up an often overlooked aspect of the role of laws in society, that it it’s important that there exist an ability to break laws. It’s critically important to the growth and flexibility of a society that laws are never perfectly enforced, that there remain ways to evade persecution. It is healthy. Faced with this situation, societies have to think further about what might have been missed in existing law that would cause ongoing skirting of the law and find better ways to structure its mutual responsibilities that we each impose on each other, often unjustly. It would be a terrible thing if the snapshot of laws at any given moment in time was allowed to be perfectly enforced. Laws are not moral documents. Their creation is fraught with unjust power grabs and non-universal moral codes. They are also created knowing that they will not be perfectly enforced and are given exaggerated cruelty when enforced to discourage others. Perfect enforcement would require a full rewrite of all laws.
Pardons should be changed to be subject to congressional approval (and rate limited). It’s clearly been demonstrated that the temptation to abuse it is too great.
As a young engineer, I was once visited at my work desk by my CEO and the HR team because of all the Dilberts I had up on my cubicle wall. They felt they were harming morale. The engineers around me loved them, but they made fun of management, the real issue. I was surprised it merited the attention. I won a short battle over the issue and was allowed to keep them up. I still have a photo of that cubicle with them up.
Once, before the web existed, I emailed Scott and joked that perhaps he was someone at my company, looking over my shoulder. The comics were often absurd but also so accurate. He replied something friendly, I forget what.
Journals themselves should make policies about the extent to which they allow the use of LLMs. In some areas it might be considered more benign than in others.
reply