Updated August 13, 2025
This is how I use Artificial Intelligence (Large Language Models – LLMs)
AI never writes for me; it only assists me in improving what I have already written. Every sentence of every article, comment, or email is mine. I feel it would be intellectually dishonest to put my name under something that isn’t authentically me.
Trust, but verify
Don’t take my word for it. If you want to, you can copy a segment of my article and paste it into an AI detection tool. For example here is what GPTZero says:
However if the scan sometimes shows a small percentage of AI-generated text, it’s probably because of edits I do using LLMs or a mistake within the error margin.
How I use LLM’s?
That said, I still use AI because it’s an incredibly useful tool:
- Proofreading:
AI highlights areas to improve, but I manually make the changes. - Finding words:
As English isn’t my first language, AI helps me find the right words when I struggle. - Organizing notes:
I ask AI to create an outline out of my chaotic notes, which then I refine before writing. - Quotes:
When I vaguely remember a quote, AI helps me locate it and find the source. - Tone:
Lately I started asking AI to review my text to help me avoid misunderstanding and communicate clearly.
All images are from real events with real people – I’ve used only one or two AI-generated images, and only for topics related to AI.
I mainly use the free versions of ChatGPT, Perplexity, Gemini and the models included on Duck.ai
Check out my article: My AI usage policy: Robots are Confused About Tango
