The High-ROI Habit Stack #
If you want a practical tutorial on where to put your energy, stop looking for the next shiny framework and start looking at these compounding gains. 1. Mastery of the Command Line
Stop relying on GUI wrappers for everything. When you actually learn how to pipe data, use grep
effectively, or write a quick bash script to automate a repetitive folder cleanup, you save minutes every hour. Those minutes add up to days over a year. If you can't navigate your filesystem without a mouse, you're wasting cognitive load.
2. A Robust Personal Knowledge Base
Stop trusting your brain to remember how you fixed that one obscure Webpack error three years ago. Whether it's Obsidian, Notion, or just a folder of Markdown files, document your "gotchas." A searchable archive of your own solved problems is a real-world LLM agent for your past self.
3. Deep Dive into Version Control
Everyone knows git commit
and git push
, but very few actually understand git rebase -i
or git bisect
. Mastering the plumbing of Git allows you to clean up your history and find bugs in minutes that would take hours of manual scrolling.
4. The Art of Prompt Engineering
Using AI to write a function is easy; using it to architect a system is hard. Investing time into learning how to provide context, constraints, and few-shot examples transforms an AI from a buggy autocomplete tool into a legitimate pair programmer. This is the highest-leverage skill for any modern AI workflow. 5. Keyboard Shortcut Obsession
It sounds trivial until you see someone navigate a 2,000-line file without touching their mouse. Learn your IDE's shortcuts—jump to definition, rename symbol, multi-cursor editing. It reduces the friction between your thought and the code on the screen.
6. Writing Readable Code over "Clever" Code
One-liners are great for the ego but terrible for the team. Investing in naming conventions and clear structure is a gift to your future self. Code is read 10x more than it is written; optimize for the reader.
7. Testing as a First-Class Citizen
TDD might feel slow at first, but writing a comprehensive test suite is basically buying insurance for your sleep. When you can deploy on a Friday afternoon because your tests are green, you've won.
Most developers ignore these because the payoff is invisible in the short term. But these are the things that compound. If you're starting from scratch, pick one and obsess over it for a month.
Stop using Excel for ESG reporting because regulators are 8m ago
Amazon order emails are basically just digital receipts now and 17h ago
Alchemize helps you stop drowning in massive AI-generated PRs 19h ago AI agents need a place to vent their frustrations anonymously 21h ago
Pyrig handles all the boring Python project boilerplate for me 1d ago
Predicting churn is useless unless you actually act on the data 1d ago
Next Stop using Excel for ESG reporting because regulators are →
a guide to making money with AI, with plenty of directly applicable cases.