The thinking still has to be yours.

There is a futurist called Marina Gorbis who said something I keep coming back to. Powered by data, AI can only look backward. We need people to imagine better futures.

She said it factually. I read it as something deeper.

Data is the past. It is everything that has already happened, already been seen, already been decided. Creativity is the opposite. It is the thing that does not exist yet. And creative direction — the real kind — is not a technical skill. It is an instinct. A point of view about what should exist that no dataset can generate because it has not happened yet.

That is why handing creative direction to AI does not work.

What I observed

When AI entered creative work, I was paying attention. The early images were limited. The early videos were stranger. But the signal was clear. This was not a trend to observe from a distance. It was a shift that required a decision about how to lead through it, not whether to use it.

I led the team into it early. There was no framework to follow, no established best practice in our category. We built our own understanding through months of real work, real briefs, real failures, and real solutions.

What came out of that process was not just a set of tools. It was a point of view on how AI should and should not be directed in brand creative work.

What we found was this. AI works brilliantly for categories with a lot of data behind them. Fashion. Skincare. Beauty. Products the world has photographed a million times. Our planters were different. Unique shapes, copyrighted designs, not in any training data anywhere. Every model produced something generic. It took four to five months of prompt engineering and model training before we could render the actual product accurately.

That is the problem nobody tells you about when they say AI saves time. It does. Once you solve the hard problem first.

What is happening across the industry

Most brands skipped the hard problem entirely. They opened the tool and started producing. And now everything looks the same.

The campaigns look the same. The same kind of model holding the same kind of product in the same kind of light with the same composition. Because AI is working from the same data everyone else is using. It is giving people the average of what has already been done.

This is not AI failing. This is humans failing to do their job.

AI is an amplifier. Bring it clear thinking and it amplifies that. Bring it nothing and it amplifies the average. The brief still has to be written. The standard still has to be set. The judgment about what is on-brand and what is not — that stays with the creative director.

What AI actually unlocks

There is one thing AI has done that does not get talked about enough. Not cost cutting. Creative freedom within budget.

Most D2C brands do not have the budgets luxury houses operate with. The location, the props, the crew — all of it gets decided by a number. Creativity gets tamed to fit what is available.

AI changes that math for smaller brands. You can try ten set directions in the time it used to take to build one. You can test a premium visual before committing to the cost of shooting it. You can show a founder what the vision actually looks like instead of a cheaper version of it.

This is not about replacing the shoot. We still photograph the actual product. We take a clean image, bring it into the AI workflow, build the set around it, finish in Photoshop. The product stays honest. The environment becomes what the budget could not afford.

The model that works

Human thinking first. AI directed at scale.

The creative standard, the positioning judgment, the brief — these remain the CD's responsibility. AI serves the vision. It does not generate it.

The brands and creative teams that come out of this period strongest are not the ones who used AI the most. They are the ones who were clearest about what they were building before AI touched it.

AI saves time. It saves money. It does not save you from thinking. You still have to do that part.