How reading philosophy and writing blogs make me a better developer.

2026 Day 173. #PersonalDays #TechDays.

Much of the development work now involves really thinking things through before you type even a single line of code. Prototyping has never been easier, once you have the idea locked in (or not) you write out the prompt and get started.

The daily blog that I have been working on has made some personal progress for me over anything else. My writing has improved, I have gotten better at identifying what reads and what does not, and I find it easier to write down my thoughts in a structured manner.

This has great improved my ability to write prompts. Instead of working with singular one-liners, I can write it down cleanly. With the technical knowledge mixed in I am able to write down ideas that can sound very complex but are simpler for the machine to understand.

Reading philosophy has opened up my mind to think beyond myself. Not that I have become extremely selfless but I can talk to myself about topics that go beyond the self. It has enabled to look at ideas, analyse them, reject or accept them, and explain them to others with a personalised touch.

This is a task that requires me to settle down, rest my mind, and constantly communicate with myself. It can be a little exhausting, although sometimes when one idea gives rise to another, it can be quite invigorating.

When I am thinking about an idea, whether it be function or a design, I have to evaluate more it concretely before writing the prompt. I criticise the function to make it as simple as possible so the complexity is not in the explanation but in the working. I evaluate the design for responsive sizes, some only look good on one screen but I must work to make the responsive ideas interesting as well.

For functional ideas I try to focus on good practices, APIs, security, rate limiting, etc. and the sort which the LLM does not seem to recognise in the first instance. Working with my preferred LLM I have learned a lot more about how it works, what it is good at, and what it misses. This affects the formation of the prompt as I try to maximise the decisions that I make so the LLM only focusses on coding it correctly.

Writing prompts for design ideas can be a little more complex.

Earlier I used to build a design in my mind, wireframe it, draw a responsive version and then translate it into words and get a working prototype. Changes here were rather difficult and complex and often times I worked more on correcting the output of the LLM than it would have taken to simply write it from scratch.

The more evolved workflow involves working on the copy first, at least section by section. I try to write out everything I want to write, or a sensible sentence of expected length. Then I throw the text is some divs and start applying flex or grid styles to arrange them correctly. Once the text has been placed exactly where I want, things get simpler from here.

I simply pass the component I wrote, add what animations I want, the colours (often are just lifted from the theme), and any background that I want to have. The animation can be precisely controlled with specific keywords (I work with GSAP so I include a lot of — the animation should have scrub: true and start at top 80% and end at top 20%) while the backgrounds are where I let my LLM have a little leeway.

Reading & writing are game changers in how to prompt the LLM. As a software developer it has become extremely important that I improve this skills.

They can also be great add ons if you are looking to work in SEO, website copywriting, managerial roles, etc. The learnings go beyond writing with good grammar and thinking clearly.

It enables a creative in you that had not been an opportunity to express itself. Sorting out your thoughts leads to less confusion in the head and more output in whatever you are putting your mind to. These habits enable you to be more of you.

See you tomorrow.