Here's something I can't stop noticing:
Everyone's "using AI" now. Agencies. Brands. Startups. Enterprises. Your competitor. Your cousin.
And yet... most aren't getting real value from it. The ones using AI for creative work? Their output all looks the same. Same aesthetic. Same language. Same "optimized" ideas floating slightly above real breakthrough creative work.
Different problems, same root cause.
The businesses struggling to unlock value and the teams drowning in generic output are making the same mistake: they're treating AI like a vending machine.
The vending machine problem
Most teams treat AI like a vending machine. Prompt in, asset out. Done.
That's fast. It's also why everything looks generic.
AI is incredible at pattern replication — it's very good at what has already worked. But that means if everyone's prompting the same way, everyone's getting the same output.
The stats back this up: only 6% of companies are seeing meaningful AI impact (McKinsey). Over 70% of marketers cite "generic or bland content" as their top AI concern. And only 23% are even training AI on their brand guidelines.
The sameness isn't an AI problem. It's a relationship problem.
Relationship, not prompts
I've spent months building a real collaboration system with AI. Documented working styles. Communication shortcuts. Feedback loops that compound over time.
And here's what I've learned: the gains don't come from better prompts. They come from treating AI like a partner you're building a relationship with — not a tool you're operating.
That means investing in the back-and-forth. Following up. Challenging. Steering the conversation. Not accepting the first draft.
Most teams skip this entirely. They want the output without the relationship. And then they wonder why everything feels flat.
Taste is the filter
Here's the other piece: AI can generate anything. The skill isn't generating. It's selecting.
Knowing what's off. Pushing back. Recognizing when something is close but not right.
Taste isn't preference — it's judgment. It's knowing what not to make. What to hold back. What feels risky but right.
When everyone has the same tools, competitive advantage shifts from execution to selection. From production to restraint. From scale to discernment.
The brands winning with AI aren't asking "what can AI make for us?" They're asking "what should exist — and what shouldn't?"
Slower than you'd think
One more thing people get wrong: even with an efficient AI partner, work doesn't become instant.
Discovery time. Brainstorming. Iteration. Thinking. These don't disappear.
The shape of work changes. The existence of work doesn't.
The teams producing distinctive work aren't moving faster by skipping steps. They're moving smarter by investing in the relationship and applying taste to the output.
The way out
The sameness problem has a fix. It's just not the fix most people want to hear.
It's not a better model. It's not a magic prompt. It's changing how you work with AI — building collaboration infrastructure, developing your own taste filter, and treating AI like a partner worth training.
I've been documenting what actually works. The methodology is real. And the teams that figure this out first won't just make better work — they'll make work that's unmistakably theirs.
In a world of infinite AI content, being unmistakable is the only competitive advantage left.
I write about AI and business here. If you're experimenting too, I'd love to hear what you're learning.
Sources: McKinsey Global Survey on AI (2024), ContentMarketing.ai State of AI Report (2024), Content Marketing Institute (2024).
Note: Observations here are based on publicly available AI tools and models as they existed when tested. Outputs may differ materially as systems update or as prompts and inputs change. AI doesn't replace taste, judgment, or creative intent — it amplifies whatever you bring to it.