Week 10, 2024 - Flowers and AI

Spring is here! No doubt now. The other day I was walking around with a light jacket only, and more and more often I can have breakfast on the balcony. And the tulips are starting to show their heads in our garden! Back when my daughters were younger, for a few years, we had a habit of planting tulip bulbs in the autumn, so the arrival of spring could come with the sight of some beautiful flowers. We don’t plant bulbs anymore, but the tulips keep on coming.

I guess there’s a leadership analogy there, about planting the seeds, trusting that the work was not in vain during the winter when seemingly nothing happens, and finally, the results showing up, for years and years to come.

📋 What I learned this week

The article of the week was a book review covering Oliver Burkeman’s 4000 Weeks. Some criticized the book as a bit repetitive of a few concepts, but I found the ideas there so revelatory that I didn’t mind a few reiterations.

As planned, I also worked on my LinkedIn presence, which meant a lot of learning and a little doing. There are some good starting points here, but just checking a bunch of profiles that I liked helped. More next week.

I didn’t plan to, but acting on a quick impulse, I re-arranged the site's front page, moving these weekly newsletters to the right column. I hope this creates a better separation between these more personal updates and the less ephemeral content about engineering leadership.

This resulted in some SLOW_GET_HELPER warnings, so I finetuned the filter to search only the primary tag of the articles, and just to be safe, increased the resources on the instance too. (When I started, I kept CPU and RAM on the bare minimum to see how it would break - I guess this was it.) It’s so much fun to tinker with the engine besides producing content.

🎯 What I want to try next week

I have a few article ideas, and would like to publish at least one of them, but can’t commit to anything yet. The main priority will still be LinkedIn. I might show up on Wednesday at the CraftHub Mobile Meetup in Budapest. During covid, most events like this moved to fully online and just stayed like that after. It’s nice to see a little switch back to in-person. Ping me if you want to chat there!

🤖 Something cool: AI-Supported Book Review

I keep my highlights from books and articles in Readwise, and since I saved a lot from the 4000 Weeks book, I wanted to see if I could use these to help me write a blog post. The idea was that my highlights should reflect what I found to be the best parts of the book, which could provide a great foundation for a review article.

I have my Readwise highlights synchronized to Notion, where I write my articles, so I copied the ones from the book to an empty page, and started with this simple AI prompt:

Write a book review blog post using the highlights above. Use subheadings with emojis to structure the text.

The result was surprisingly good. With a few more instructions, like “merge the last two parts under a single subheading” and “add the aspect about the role of technology in how our focus is fragmented”, I could get to a good foundation to work on.

To ensure I’m not missing something, I created a toggle heading called “Other reviews” below the article and just copy-pasted the first page of Goodreads reviews there. (Literally: CMD+A, CMD+C, I found that AI works well with unstructured text.) Then I used this prompt:

Summarize the other reviews, select three observations that are missing from my article

The results again were surprisingly useful:

1. The book encourages readers to accept their limitations and make peace with the fact that they won't be able to do everything. This perspective is fundamental to dispelling the illusion that perfect time management will lead to accomplishing all our desires and goals.
2. Burkeman's approach to time management is more philosophical than practical. He doesn't offer quick fixes or techniques but rather prompts readers to reflect on the finite nature of life and the choices they make about how to spend their time.
3. Some reviewers found the book to be repetitive, which might indicate that the essential concepts could be condensed. The repetition, however, may also serve to reinforce the key messages and help readers internalize them.

This also gave some good points, some of which I ended up incorporating in the final version of my review.

I’m collecting my ideas on the “efficient way to use AI in writing”, which I’ll publish eventually, but for the moment, I believe this is a good example of the collaborative approach of AI use during the creative process.

That’s it for today, find some blooming tulips near where you live during the weekend,

Péter