Anyone can make a hundred ads before lunch now.
Copy, image, video, voiceover. Type a sentence, get a campaign. The tools are that good and they get better monthly.
Your nephew who spent two years in college and has very specific coffee preferences could probably do it.
But it won't work.
Weights and counterweights. When something becomes this accessible, the window of demand moves.
After ChatGPT launched, new book releases on Amazon nearly tripled in three years. Two economists, one at Cornell and one at Minnesota, tracked it. Three times the books, and readers didn't get three times the good ones. They got a bigger haystack.
More stuff means more bad stuff.
Art learned it in about 48 hours after chatGPT launched its first image model.
Publishing learned it next.
Advertising is learning it right now.
Because the machine will hand you a hundred options and it has no opinion about any of them. It doesn't care if the ad runs or dies. Ask for a hundred more and it hands you those too, just as cheerfully.
Somebody still has to look at the hundred and pick the one.
Somebody has to know why option 47 stops a thumb and option 12 gets scrolled past by every human alive.
Nobody automated that part. If anything it's worth more now, because the pile it has to sort got a hundred times taller.
For most of history, execution was the hard part. You could imagine something great and lack the hands to build it. That flipped. The hands are free now. The judgment is what's scarce.
What's going on now
We use the machines, heavily. A creative test that used to take a week of production takes an afternoon.
But nothing runs just because it exists:
- Generate wide. Twenty concepts where we used to afford three.
- Kill fast. Most die before a client ever sees them.
- Put real money behind the survivors and let buyers vote.
If your feed is full of ads that all look competent and none of them are pulling, that's a taste problem. Book a strategy call and we'll tell you which ones deserve your budget.