I Built a Faceless Video Tool. Watching People Use It Made Me More Optimistic, Not Less.
I have spent the last year building an AI video generation tool for faceless creators. Around 90 people pay me every month to use it, and I get to watch what happens after they sign up.
I expected to come out of this cynical. Most people who build in an AI category do. Instead I have ended up genuinely convinced that this is the most accessible moment short-form video has ever had, and that the people succeeding at it are not who I assumed.
The barrier that disappeared
Two years ago, making a faceless channel work meant either forty minutes of editing per video or a monthly budget for an editor. That maths only closed if you were already earning. Everyone else did it as a hobby, posted twice a week, and stalled.
That constraint is gone. Script, voiceover, visuals, captions, and publishing to TikTok, YouTube Shorts, and Instagram Reels now happen in one pass, in minutes. The interesting part is not the time saved. It is what people do with it.
When testing an idea costs almost nothing, you stop protecting your ideas. My most successful users are not producing more polished videos than everyone else. They are running more experiments, and they are willing to be wrong twenty times a week because being wrong is now cheap.
The people who grow are not the professionals
This is the part that surprised me most. I assumed the winners would be people with a media background, or at least an editing background. They mostly are not.
They are people with a narrow obsession. Someone who knows a specific trade inside out. Someone who has spent five years in a niche hobby. Someone who works in an industry and understands what everyone in it quietly worries about. They are not better at video. They just know exactly who they are talking to, and it shows in the first two seconds of everything they post.
Get Brian Millot’s stories in your inbox
Join Medium for free to get updates from this writer.
That is a genuinely democratic filter. Production skill used to gate this. Now the thing that gates it is knowing something real about a group of people, which far more of us have than we think.
Consistency finally became realistic
The advice was always to post daily. It was also always unrealistic, which is why almost nobody followed it and why it kept getting dismissed as guru talk.
It is realistic now. That single change is why I keep seeing accounts go from nothing to a real audience in a couple of months, in niches I would have called saturated. The algorithm was never withholding anything. It just needed enough samples to work out who you were for, and most people never gave it enough samples to decide.
[Optional: drop in one user story here. A niche, a rough starting point, where they got to and how long it took. One concrete example does more work than any argument in this piece.]
What that means if you are thinking about starting
The first thirty days are calibration, not feedback. Post daily and deliberately do not judge the results. Nearly everyone who quits does it in week two, on data that was never going to mean anything yet.
At day thirty, look at your best three videos and make twenty more like them. That loop is simple, it is boring, and it works far more often than people expect once the production cost is no longer fighting you.
Pick your niche by what your audience cares enough about to spend money solving. That decision sets your ceiling before you post anything, and it is the one part of this that is still entirely yours to get right.
Write your hooks before you produce anything. If you can write twenty different openings for the same audience, you understand them well enough to start.
Where I have landed
I set out to remove a chore and I think I have ended up removing an excuse, which is better. The people I watch succeeding are not being carried by the tool. They are just no longer spending their evenings dragging clips around a timeline instead of thinking about their audience.
The creative work did not go away. It moved to the part that was always the interesting part. That feels like the right trade, and it is why I am still building this.
If you want to see the pipeline I built, it is at Blipix. But the honest advice is to start this week with whatever you have. The tooling will keep improving. The thirty days of calibration will not start themselves.
