Hi. FYI, I’ll be cross-posting to my blog karlyang.net, where you can find other writing that won’t appear on Substack. The basic theory here is that, most of the time, dependencies slow you down. There are ways to ac/decelerate the process that are problem-agnostic. Usually, you are communicating back and forth with other people until you have reached some form of mutually acceptable agreement, and there are ways to do that faster.
I agree with all of this (but still am working on it).
I would note that, as someone recovering from consultant mode, clarity and focus in text are now available as often-good-enough outputs of LLMs. There's a lot of leverage in making your default response to, "I'm too tired to write this more clearly" be "I'll ask chatGPT to draft it, and I can tweak and send" rather than "I guess I'll do it tomorrow."
Should have provided a trigger warning about the solo “hi” Slack message 😂
I agree with all of this (but still am working on it).
I would note that, as someone recovering from consultant mode, clarity and focus in text are now available as often-good-enough outputs of LLMs. There's a lot of leverage in making your default response to, "I'm too tired to write this more clearly" be "I'll ask chatGPT to draft it, and I can tweak and send" rather than "I guess I'll do it tomorrow."