Ideas and topics: how to avoid the obvious
AI for video · Lesson 2 / 20
The idea decides more than the production
A video with flawless editing on a dull topic won't find viewers, while a simple phone recording that lands precisely might. So the idea is where effort belongs first.
Why the first list is always generic
Asking for "ideas for videos about [topic]" returns what's already obvious: "5 beginner mistakes", "how to start", "myths about…". The model produces the most probable, and the most probable is the most worn out.
What produces the non-obvious
- Your data instead of general knowledge. "Here are 40 questions people ask me in the comments: [list]. Which topics suggest themselves?" — ideas from real audience questions rather than from the air.
- Axes instead of a list. "Give me 20 ideas across four groups: demolishing a common misconception, dissecting my own mistake, comparing two approaches, answering an awkward question."
- The reverse move. "What does everyone in this niche avoid discussing, and why?" — that's usually where the unoccupied topic sits.
- Other people's comments. Comments under competitors' popular videos are a ready list of what's been left uncovered.
Selection matters more than generation
Twenty ideas are useless until one is chosen. The criteria that work: can I say something on this that others won't; do I have my own example or data; would I watch such a video to the end myself.
The third question eliminates the most. If you wouldn't watch it through, nobody will.
An idea bank rather than a one-off session
Once a week, add ideas from comments, questions and observations to one document. Within a month you have a queue, and "what should I make a video about" stops existing — and that's the question that most often stops people filming.
Cheat sheet
- The first list is always generic — you need your own data.
- Generation axes rather than a plain list.
- Selection matters more than quantity.
- An idea bank removes "what shall I film about".
Gather your own real data: at least 20 questions from comments, messages or conversations with your audience. Using them, get 20 video ideas across the four axes given. Separately ask what everyone in your niche avoids discussing. Select three ideas against the lesson's three criteria and explain the choice, including an honest answer to "would I watch this myself".