
An explorer watches raw video artifacts fuse into a massive machine.
I’ve worked with ‘let’s make some cool’ video for most of my career.
In shoots.
Writing scripts.
Working with brands.
Agencies.
And creating music videos and Transmedia films.
But I’ve spent even more time with something less visible:
Video systems.
And now, video systems with AI.
And the uncomfortable truth I keep running into is this:
Many teams don’t actually have a video problem.
They have a video systems problem.
They already believe in video.
They already have tools.
They already have people who ‘know video.’
What they don’t always have is anything repeatable.
So every new video becomes a small crisis.
Someone writes a brief.
Someone edits.
Someone waits for approvals.
Someone asks for a tweak.
Someone worries about brand.
Someone worries about compliance.
Someone asks for last-minute changes.
Someone wonders where the file lives.
Someone asks why live audio failed again.
By the time it ships or goes live, the moment has passed.
Video doesn’t fail because teams lack talent.
It fails because it’s treated like a one-off artistic artifact instead of infrastructure.
There’s another truth I’ve learned the hard way:
If you have a big budget, you can do impressive things.
CGI.
Sets.
Studios.
International shoots
A-list actors.
You can throw money at production value.
You can hire agencies.
And honestly, for a while, that works.
But even then, video stays messy.
Manual.
Held together by people and heroics.
Most teams don’t live in that world anyway.
They don’t have a dedicated video team.
They don’t have endless agency hours.
They don’t have someone whose full-time job is chasing approvals and fixing captions at midnight.
That’s the moment when video quietly changes.
It stops being about ‘making content.’
And starts becoming a system.
Not because it’s elegant.
But because it’s necessary.
Training materials need to update without reshooting.
Summaries need to exist without an editor.
Highlights need to be cut automatically.
Captions need to be right every time.
Compliance needs to be checked before anyone posts.
When resources are tight, repeatability stops being optional.
That’s also where AI enters the story.
Not as magic.
Not as a shortcut.
As leverage.
I’ve watched teams pile on AI tools hoping for relief.
Avatar tools.
Editing tools.
Automation tools.
And without a system underneath, all that happens is chaos gets faster.
More videos.
More variants.
More confusion about what goes where, when, and why.
That’s why, when I look at a video setup now, I don’t start with the video.
I start with questions that feel almost boring:
What’s the input?
What triggers creation?
Where does approval actually happen?
What breaks when volume increases?
What can’t be vague because risk is involved?
Most teams don’t like this part.
It feels slow.
Unsexy.
But it’s the difference between video as effort and video as infrastructure.
I’ve been in systems and discussions where:
A spreadsheet quietly becomes hundreds of videos.
Text turns into video without anyone opening an editor.
A missed demo triggers a contextual follow-up automatically.
Compliance isn’t a panic moment, it’s just part of the flow.
When it works, video stops being ‘content.’
It becomes something closer to plumbing.
Predictable.
Invisible.
Reliable.
Here’s the tradeoff I’ve had to accept myself:
I can’t do this by being creative-first.
And I know that because I was once a filmmaker, and a big part of my brain still thinks that way.
I have to be constraint-first.
Clear scope.
Clear handoff.
Clear ownership.
That’s not how most people want to think about video.
But it’s how video finally starts working.
When teams don’t have infinite money,
infinite time,
or infinite patience.
"Most teams don’t fail at video because they lack creativity, they fail because they never turned it into a system."3 Lessons:
Video becomes a system when manual work stops being affordable.
AI only creates leverage when structure exists underneath it.
Treating video like infrastructure feels boring, until it finally works.
Follow me on LinkedIn: Robert Figueras
