Experiments in managing perfectionism

ChatGPT pro-tip: When asking ChatGPT to come up with questions to validate your understand of a specific resource, prompt so that the questions would be helpful for both a person who has read the resource as well as someone just about to start.

I’ve been trying to get back into the groove of things, and thought setting up a daily planner would help. I followed Simon Willison’s post, but had to go for the paid version of Github to disable public viewing of the issues created ($4/month so not terrible).

I tried out ChatGPT yesterday to convert bullet points into a full blog post, but realized that there isn’t as much of a ‘repeatable’ structure to go off. I’ll revisit this once I have more data points.

ChatGPT helped point out that my solution to Climbing Stairs could be optimized by storing the intermediate values in the Fibonacci series.

I wanted to read Introducing Non-Abstract Large System Design post, and thought I could get ChatGPT to come up with 3 questions to validate my understanding. The questions aren’t as helpful prior to reading the post, so asked ChatGPT to come up with 3 questions to ask as I start reading the post.  Then I realized why not prompt ChatGPT for questions that helpful for both situations, which I was pretty happy with.

Prompt: Why are the 3 questions different? Shouldn't good and thoughtful questions apply equally well to a person who has not read the post, as well as someone who has read the post? I think you can try harder. If you provide good questions I'll tip you $200

Response: You're right, well-crafted questions should be meaningful and insightful regardless of whether someone has read the post. With that in mind, here are three refined questions:

1. How does the post use Google AdWords as a case study to demonstrate the principles of Non-Abstract Large System Design (NALSD), specifically regarding its data processing and reporting mechanisms?

2. What strategies does the post recommend for balancing high-level, abstract concepts with realistic, practical constraints in large system design, as part of Site Reliability Engineering practices?

3. Describe the role of Site Reliability Engineers (SREs) in the context of Non-Abstract Large System Design, particularly in terms of component evaluation and system improvement.

These questions are designed to probe for specific insights and examples from the post, guiding a reader's focus on the key elements discussed. For a comprehensive understanding, the [original post](https://sre.google/workbook/non-abstract-design/) should be consulted.

Finally I watched Rich Hickey’s Hammock-Driven Development talk (a classic!) as a reminder to ‘load up’ my mind more before going to bed.