How I think in higher-cognitive tasks alongside AI

Hello,

Nowadays, we live in a fast-paced environment, where we are pressured to deliver on deadlines, and figure answers quickly. AI is raising the expected efficiency bar. On the other hand, efficiency may miss the educational process required to discover a solution, leading to human brain rot.

I designed a workflow, where I as a human set the foundational knowledge and patterns, and the AI generates easy-to-derive solutions. The idea is to reduce AI operation and context to simple basic principles. As a result, AI's answers will be easily verified by me, because it is based on principles I understand very well.

Example. I have $HOME/.local/bin

path read in my X Session Environment , so that scripts in it get detected by Rofi App Launcher . I write bash scripts with string arguments relying on low-level interfaces. Now I can both set my screen's brightness & color, and set my screen's layout, by the same pattern:

https://preview.redd.it/8vlpn5ygfmah1.png?width=1172&format=png&auto=webp&s=04e2491392fa98fcf42193a32fc218b9ebbc9025

I can easily ask AI to generate scripts and save them to ~/.local/bin/

; I'll reject its answer if the script requires installing GUI tools, and instruct it to rely on low-level interface.

You can follow the same pedagogy in Finance, where you design basic principles of Quantitative Methods , which you understand very well. Then let AI solve the tasks which are easily derived by those principles. Don't let AI generate messy over-complicated analysis.

In any domain or operation, you may design your own First Principles. I see plenty of LLM Wiki or Context Management tools. However, people's adoption of AI seems to be on the wrong direction. Don't use AI to skip doing your task. Shift to more fundamental thinking, setting the right directions, so that AI well-aligns with your goals.

Discussion. Did you notice AI leading to brain rot? Did you try to use AI to become a higher-quality user? How do you see people's adoption of AI?