Comment by latexr

> Blank PR descriptions are a tax on your reviewers. I paste my diff summary and ask:

LLM-generated PR descriptions are a bigger tax than blank descriptions.

> Instead of a normal rubber-duck session, I want pushback

If you get pushback, it’s no longer rubber ducking; rubber ducks don’t talk back. The whole point of rubber ducking is to work it out yourself as you reason through your thought process out loud. If you talk to another person or a chat window and are getting feedback, that’s just a discussion.

> When an incident closes, the postmortem doc gets written under pressure. I paste the timeline and ask:

The purpose of a postmortem is to reflect. It’s to understand what went wrong, why it went wrong, and what and how you need to change so it doesn’t happen again, in addition to exploring where else problems could occur and fix them early. A timeline alone doesn’t tell you that. An LLM can never say what lessons you learned.

> Long Slack threads where decisions got made are black holes. I paste the thread and ask: (…) Three months from now, your future self will thank you.

Or curse you, when it happens that the LLM summarised it wrong and you wasted your time on something you weren’t supposed to.

> None of these are "write my code for me" prompts. Every one is about getting to clarity faster — tighter scope, better context, cleaner handoffs. The value isn't outsourcing thinking; it's compressing the parts of the job that drain time without demanding creativity.

And yet, you are outsourcing thinking with several of those. It also seems (though I hope I’m misinterpreting) you are more concerned with making the job easier for yourself first without fully considering how those actions will affect the people on the receiving end.