How I'm Building an AI-Native YouTube Channel
Por Miguel Torrez, fundador de Genial Agency
By the end you will have a simple plan for a YouTube workflow where AI helps with ideas, research, editing and tracking, while your own experience stays at the center of every video.
Lo que necesitas
A Notion workspace (notion.so)
Recordings or transcripts of your calls (friends, customers, discovery or sales calls)
A list of YouTube channels in your niche
A camera to film yourself
Paso a paso
Collect your calls as raw material
Gather the calls you already have: conversations with friends, customers, discovery calls and sales calls. I use these as the starting point because the ideas come from your own experience, so you are the one telling the story. One single call can give you tens of ideas.
Break each call into content items in Notion
Create a Notion table (think of it as a spreadsheet) that holds every important thing you said about your subject. Turn each call into separate content items you can later use for LinkedIn posts or YouTube videos.
Tag each idea by type
Label every content item with the kind of post it should become. I use: personal experience, AI, or both. Swap in the categories of your own niche; this works for any niche.
List the channels you follow in your niche
If you are just starting and have no ideas yet, research what is trending instead. Build a "YouTube inspiration" list in Notion with the channels in your niche that have lots of views and videos.
Pull their recent videos and spot the gaps
Set up a system that grabs those channels' videos and tells you what performs well and what is missing from your channel. Use their transcripts as input for your own content. Start with videos from the last one to three months rather than their whole history.
Mine the comments for questions
Also collect the comment sections of those videos. Comments are usually questions, so each one is a content idea tied directly to what people actually want.
Film yourself and hand the editing to AI
Record your video, then give the footage to AI. I say a few spoken keywords while filming so the AI knows where to cut sections and add transitions. The goal: you film, AI handles the editing, the upload and the thumbnail.
Track every video's performance in Notion
Build a table of all your YouTube videos with their views and impressions. Use it to see which hooks and thumbnails do well and to know your baseline. Add a content management system (CMS) view to follow your YouTube subscribers and LinkedIn followers over time.
Keep yourself at the core
Use AI as the assistant, not the face. Your experience and your face stay at the center of what you publish. I choose not to use an AI avatar for now, to keep real contact with people.
Ten cuidado con
Waiting until you have ideas: most people never start because they do not know what to talk about. Your calls already contain the ideas.
Researching too much: I store thousands of videos and it is overkill. The last one to three months of videos is enough.
Publishing without tracking: without views, impressions and a baseline you cannot tell which hooks and thumbnails work.
Replacing yourself with an AI avatar: I avoid it because it weakens contact with real people.
Qué ver después
5:32Turn Calls into Content with AIPull post ideas and quotes from your meeting transcripts, the raw material this workflow starts from.
9:15How You Can Grow Your YouTube Channel with ClaudeConnect Claude to your channel and see which topics to repeat and which videos to repackage.
52:26AI for Content Workflows and Video Editing Webinar July 9thWatch how to hand video editing to AI, either with code or with generative video.
