But I am addicted to technology since I can remember. My parents had the care to manage the vices by delaying to give me free access to technology, I had my first own computer with around 13 years old, and my first smartphone only around 16.

With a computer, I sticked into games. The whole day in the computer, stopping just because of hungry, sleepiness or headache. I was aware that it was a idiotic behavior but I didn't had much discipline to handle.

But then around my 15 years old, with the pressure of my father to make something smart, I initiated my journey on Linux and programming through a expensive yet shallow course about pentesting. It made me like how software works, but without understanding the development I would never get rid of the label of "script kiddie".

So after getting a minimal adaption in that Debian distro (Kali Linux), I decided to get rid of my vice of games: Abandoned the Windows 8.1 and the Kali Linux at VMWare, installing a pure Debian as second boot in my notebook.

After some time I removed Windows from the disk, and the ganes weren't a problem anymore. But then… Linux became my new game.

Warping time back to now.

Distractions are hard to handle: Pleasing to the moment but packed with a high potential for damage.

LLM is bringing convenience to intellectual jobs, considerably decreasing the required efforts. We're being invited to outsource our reasoning.

This is not new, the concept of ruler and the servant is well established in humanity, sad, but this is how society works.

Before LLMs, the servants and the rulers also outsourced intellectual efforts by trust or fear, to whom they knew or had power over them:

• State, which defines what's punishable or not.
• Specialists, which brings their thoughts.
• Companies, which provides services.
• Schools, which shapes your children.

And this is not necessarily good or bad, it's just how it is. What determines the good or bad is the your decision on what and to whom outsource your intellectual efforts.

There's a couple of implications in letting someone set the direction of your life and resposibilities:

• Specialists can lie and your trust might be result of propaganda.
• No one really cares on you, unless you're useful for them.
• Companies may be looking for monopoly.
• State and schools are not a person, but a pack of them, which you don't know well.
• Good decisions requires good context.

But this always happened to occur, this outsourcing lets both the success of current world-uniting capitalism, and also the the success of the great tyrants.

Because lifetime is short, studying everything in depth is unfortunately infeasible, but key decisions are weighted by reversibility, and the more irreversible, more study it deserves.

Unfortunately we was trained by propaganda to conveniently trust:

• A small group of so called authorities for healthcare and personal life advising.
• State to micromanage the citizens.
• Companies to store our data and memories.

Training you to obey without seeking study and confront them. And now another agent enters in the game, LLMs are your own specialist to trust, it answers all your doubts without requiring any study!

They say to "fact-check" and not trust, but if you seek solution with LLM, you don't care with the source.

This is not about LLMs, the danger only increased now, but it always existed.

Because our parents and grandparents had to grind to build the foundations, they outsourced the intellectual effort to educate us, and now we feel more identified with friends and job partners than our parents, the bounds didn't grown thick.

Not even talking about the huge harm for privacy which SaaS LLM brings. The same way you'll have "your own specialist as pet", you'll have your own "spy as pet" as well. It knows your intrinsic patterns and interests. A perfect scenario for the twist of the friend's betrayal.

Writing this because I got too deep with LLMs, getting lazy and avoiding reading the codebase to understand. Started get the habit of avoid reasoning. I had enough, this is not the right way to use a tool.

It's hard to get rid of it. I am telling this after deliver a large full refactor of a legacy React web app in ~3 months. LLM indeed assisted, but how much I am willing to pay for deliver speed?

So now, until I get the enough computation power to run a conversational LLM locally, I will use the same strategy I use for Google, YouTube, social media and any kind of threat for my time and privacy: Sacrifice the convenience.

I still rather spend some extra minutes reading documentation and learning normally as I always did: Facing frustrations, getting stuck and then get relieved and proud when overcome the problem. This builds a solid experience and makes me more efficient to solve future problems.

This flow shapes the mind for better decisions, all friction counts and you know the costs with high precision.

So now, I rather search at Kiwix snapshots for solutions and run ripgrep on cloned repos for keywords that lights up solutions for my problems.

Tough decisions are being done, so here's the complete timeline tracking my usage of LLMs in my workflow and personal life.

title: LLM Usage
timeline:
  - summary: Playing with Google Bard and Nim
    date: 2023-07-18T00:33:57+00:00
  - summary: Playing with ChatGPT and Nim
    date: 2023-08-21T19:43:20+00:00
  - summary: Setupped Ollama local server with Llama 3.1 7B
    date: 2024-07-26T10:03:38-03:00
  - summary: Installed Ellama at old Doom Emacs config repo for Ollama
    date: 2024-07-27T09:52:59-03:00
  - summary: Installed Copilot Mode at old Doom Emacs config repo
    date: 2024-12-20T16:11:17-03:00
  - summary: Configured Aider at old dotfiles repo
    date: 2025-01-30T10:07:17-03:00
  - summary: Configured opencode
    date: 2025-08-05T10:24:02-03:00
  - summary: Tried Claude Code
    date: 2025-08-14T09:45:15-03:00
  - summary: Removed Copilot Mode completion from Emacs
    date: 2025-09-28T08:57:46-03:00
  - summary: Removed Aider in old dotfiles repo
    date: 2025-10-17T20:51:11-03:00
  - summary: Removed opencode from dotfiles
    date: 2025-11-13T16:30:08-03:00
  - summary: Tested gemma4:26b in a RTX 5070 Ti 16GB with Opencode
    date: 2026-04-05T19:05:45-03:00
  - summary: Started to use Claude Opus 4.6/4.7 at job
    date: 2026-04-17T15:56:10+03:00
  - summary: Played with Claude Opus 4.7 in personal projects
    date: 2026-04-28T14:08:51+00:00
  - summary: Suspending LLMs after addicting again
    date: 2026-05-09T21:34:25+00:00

• https://blog.fogus.me/meta/LLMe.html
• https://gist.github.com/richhickey/ea94e3741ff0a4e3af55b9fe6287887f
• https://wiki.gentoo.org/wiki/Project:Council/AI_policy
• https://www.netbsd.org/developers/commit-guidelines.html