I’ve been using LLM to help me write emails, captions, video scripts, ideas, code, etc. For almost every day for the past 3 years since the introduction of ChatGPT.
Since then I’ve been trying and experimenting with other models from OpenAI, Anthropic, Google, DeepSeek, Moonshot, Z.ai, GLM, Mistral, Grok, and more.
I switch around depending on the free tier limit and use them for the basics of tasks as simple as a short email reply.
As fast and convenient as it is, I felt the need to do something. I don’t want to delegate every single thought, word, sentence, and paragraph and become completely reliant on a super amazing and complex “linear algebra”.
I want to regain the originality and thought of writing. I want to have a REAL voice again.
So with this blog/article, I’m challenging myself to re-learn how to write.
I won’t fix any grammar or typo and want to use mixed language, formal-semi-formal slang, any.
Of course, mid-writing this I realized that technically I can put a lot of my “original” writing and tell the LLM to learn from my writing style. BUTT I guess that’s for a future project ;)