A pomegranate takes some work to open. Amy Hunt, the British sprinter, said as much this summer when she helped Puresport develop a pomegranate flavour for its electrolytes. What the company did next is the part I keep thinking about. It skipped the product shoot and commissioned four painters to make still lifes of the box.
Puresport's head of creative explained the choice to Creative Review in August. The easy move in 2026 is to put some product shots through an AI workflow and get hundreds of assets back, so the team asked what the opposite of that would be. The answer was four oil paintings by four different people, each with their own hand in it.
I was reminded of it reading Brooke's newsletter, The Brand Blueprint, which this week rounds up brands pushing back against AI slop. It's a good list. What interests me more is why the approach works, because I don't think the answer is nostalgia.
My read is this. When everyone has the same tools, the output starts to converge. The tools aren't bad. They're built to produce the likely answer, and the likely answer is the one everyone else is getting too. A brand that looks like the average of its category is hard to remember and easy to replace. Effort is one of the few signals left that reads as a decision. You can see that somebody chose.
J.Crew made the same bet at a larger scale. For the barn jacket's 40th year it photographed 160 people across New York, Los Angeles, Austin, Boston and Chicago. A model could generate 160 faces in an afternoon. It couldn't generate 160 people someone in Austin recognises from down the street.
Neither brand is giving up AI to make the point. The pattern is narrower than that. The tools carry the production work out of sight, and a person stays visibly responsible for the parts a customer is meant to believe: the image on the box, the faces in the campaign.
Two questions I'd put to any team right now. Where would someone's hand make this better than the fastest version? And what part of how you make things are you keeping out of sight that people might actually find interesting?
Neither needs a bigger budget. Both need time, and a willingness to let the work look a little less polished than it could. As machines make the average version cheaper, the version a person made is the one people notice.
Sources
- Creative Review, on Puresport's commissioned still lifes: creativereview.co.uk
- Puresport launch release for Pomegranate Ultra Electrolytes: endurancesportswire.com
- FashionUnited, on J.Crew's 160-person barn jacket campaign: fashionunited.com
- The Brand Blueprint, "How Brands Are Fighting AI Slop": thebrandblueprint.substack.com