Part one: AI is not your problem. Judgment is.
- Amanda Downie
- 5 days ago
- 3 min read
Right now, the biggest risk with AI in marketing is when AI replaces human judgment—when we stop thinking.
“If AI can do this faster and cheaper, why wouldn’t we use it for everything?”
“If AI can create content in seconds, why do we still need a marketing team?”
If you work in marketing and haven’t heard those questions yet, you will. And to be fair, they’re not unreasonable—at least when framed around cost. Time is money. And money is… money.
But let me stop there before it sounds like I’m coming for marketing teams. I’m not. I am a marketer. I believe deeply in the work and the people doing the work. We need them!
Yes, AI can generate endless drafts, campaign concepts, social posts and more. Used well, it can do much more than speed up productivity.
But here’s the shift some are missing: Volume isn’t a constraint anymore. Judgment is.
Others, like Tahnee Perry—AI expert and author of the AI at Work newsletter—have made similar arguments recently. The real differentiator isn’t access to the tool. It’s how thoughtfully you use it.
At a time when seemingly everyone has access to AI models, the real competitive advantage isn’t who can produce more content. It’s who can decide what’s worth saying.
When AI content works
Look at the AI content that resonates:
The Kalshi NBA Finals ad. It was weird. (Okay, 'weird' is doing some heavy lifting here. It was fully unhinged, and I loved it.)
Why did it work? Not because of the tool—but because of the decisions behind it. There was a clear concept. A point of view. A willingness to lean into its own strangeness in a way that aligned with the product. The creators owned the idea of an odd ad for an odds trading company.
Another example I love: Heinz AI Ketchup campaign. Sure, its 'old' AI now, hailing all the way back to July 2022. But guess what? It worked then, and it works now, even without the latest and greatest in agentic AI.
When asked to 'draw ketchup,' AI kept generating images resembling a Heinz bottle—across styles, prompts and iterations. They tried 'ketchup street art,' 'ketchup synthwave' and so on.
The results weren’t just fun and remarkably similar outputs. It was brand dominance, made visible.
And importantly, AI wasn’t the idea. It was the proof—used effectively by a team that chose to ask the right questions.
That’s judgment at work.
When AI content doesn’t work
Now think about the opposite.
The hyper-polished posts that feel completely empty.
The campaigns that miss cultural context.
The content that just feels… off.
We’ve all seen these examples. We usually scroll right past them.
AI fails when it’s used to replace thinking, not support it.
When outputs are accepted without critique. When teams optimize for speed and scale over resonance and efficacy. When we surrender our thinking for the sake of AI productivity.
The results are generic. Forgettable. Risky.
That’s why it is essential for teams to look beyond skilling their marketers to just build an AI agent or learn how to write better prompts.* We cannot lose our ability to think strategically, critically and confidently.
In the last two years, I’ve built AI assistants, experimented with AI agents, and taught others how to do the same. I can tell you this: prompting is quicker to learn than great editorial judgment.
And yet, most teams have focused on the former—not the latter.
What actually separates good AI content from the bad
If you zoom out, the difference between strong AI-driven content and weak AI-driven content isn’t subtle.
For me, it often comes down to a few consistent factors:
Intentional ideas (not just generated outputs)
Clear understanding of the audience
Active human judgment at every stage
AI can help you execute. But it doesn’t decide what’s worth executing.
Questions around judgment
So, if judgment is becoming the real differentiator, the next question is obvious:
Where does judgment show up in the work—and how do you build it into your process?
That’s where things get more interesting.
*Prompting is a valuable skill. No question, better prompts = better outputs. If you want to get some great prompting tips, I’d recommend this free guide from Tina Huang and Hubspot or go to YouTube and search “learn to prompt”. Here’s one I did on AI tools for writing.
Comments