AI in real life · Personal essay
Where AI actually
fits into my day.
Five practical uses across design, storytelling and product work — and the decisions I deliberately keep human.
AI entered my life through making things. I was already interested in painting, visual design, stories and digital products. The new tools did not create that curiosity; they gave it more surfaces to move across.
The useful question is not “Can AI do this?” It usually can produce something. The better question is “Which part becomes clearer, and which decision still needs me?” That is the filter I use in daily work.
1. Turning a vague idea into a visual direction
I use AI to explore layouts, moods, typography and the relationship between images and copy before committing to a site or visual concept. The output is a conversation starter, not a finished identity. I choose what fits the story, remove what feels generic and keep accessibility and mobile behaviour in the final design.
2. Developing stories without outsourcing the voice
A short scene, poem or character can begin as a few lines. AI helps me test structure, alternative dialogue or visual descriptions. I keep the emotional premise and edit the language until it sounds intentional. A fluent paragraph is not automatically a true or memorable one.
3. Moving from characters to scenes
For visual storytelling, I can turn a written character into image references, then explore expressions, locations and motion. Consistency is the difficult part: the same face, body, clothing and environment must survive multiple frames. Every scene needs visual quality control before it becomes part of a film or Reel.
4. Organising complex information
AI is useful when a subject arrives as a wall of terms. I use it to map questions, compare explanations and identify what a beginner is likely to misunderstand. I still verify facts and remove confident claims that the source cannot support.
5. Shaping products around real people
This habit eventually influenced Rocket. I was interested in how a system could translate dense market analysis into a guided Telegram flow. The aim was clarity, not certainty: the product should show information in a more approachable way while leaving the decision and risk with the person using it.