🧠 Real AI collaboration starts here

Everyone’s using AI for content now.

But most are stuck between two extremes: treating it like a magic “write this for me” button or constructing Byzantine automation systems that need constant maintenance. One undercomplicates, the other overcomplicates, and neither produces content worth reading.

When AI works well, it’s not about eliminating the human from the loop—it’s about putting the human firmly in control while using AI for fast thinking iterations, not just fast writing.

Here are five core principles for better AI-assisted content.

Quality comes from context

When I started using AI, I did exactly what the interface encouraged. The chat window design made me treat it like… well, a chat. Short messages. Minimal context.

The AI would dutifully spit out something generic.

It took me longer than I’d like to admit to realize I was being too skittish about front-loading information. The more specific context I provided upfront—audience profiles, brand voice guidelines, competitive differentiators, key messaging pillars—the more intelligent the AI’s questions became, and the better its writing.

All of my writing projects have a set of “knowledge blocks” attached as supplementary materials. They allow the AI to access context-building information for every task.

Constraints work in your favor

The better my AI outputs became, the more selective I got about inputs.

Instead of feeding AI everything, I built focused project libraries with only my signature pieces. Instead of letting it pull from its vast training data, I created blacklists of phrases I’d never use.

The constraint that matters most? Being intentional about what you include.

Think about onboarding a new content writer. You wouldn’t tell them, “go randomly browse our website and figure out our voice.” You’d point them to 10-15 of your best pieces that really capture your style. That curated selection teaches them more than your entire content archive ever could.

Same principle with AI. It’s not about volume—it’s about giving it the specific information it can learn from best.

AI is a great thinking partner, if you tell it to be

This was my biggest shift. I started explicitly telling AI to ask clarifying questions and challenge my thinking.

The AI will ask me to clarify what value the piece brings to readers and how it relates to their pain points. When I present multiple directions for a post, I’ll suddenly realize I’m trying to say too many things at once—and it forces me to focus on one core idea.

That back-and-forth isn’t inefficiency. It’s where the real value lives.

If your AI isn’t occasionally telling you that your idea needs work, you’re not using it right.

Clarity beats speed

We’ve started thinking that AI’s biggest benefit is decreasing the time we spend on tasks. And yes, productivity gains are a great perk.

But the bigger benefit is having a thinking partner who can help you clarify and polish the idea itself.

The act of explaining to AI what you want—really explaining it—forces you to examine whether your idea holds water. Can’t articulate the core insight? Maybe there isn’t one. Can’t explain why your audience should care? Perhaps they shouldn’t.

When set up right, AI acts like a sparring partner that won’t let us post half-baked ideas.

That’s the real win.

Build for iteration, not perfection

The worst thing you can do with AI content? Publish the first draft.

Smart marketers build workflows that assume multiple rounds. They treat first drafts as starting points, not endpoints. They understand that AI’s superpower isn’t writing—it’s rewriting at the speed of thought.

If you can go from prompt to published in under five minutes, you’re not iterating enough. The magic happens in rounds two through five, not round one.

***

These principles work because they are a basis for a structured collaboration—one that respects what AI does well (pattern matching, rapid iteration, tireless revision) while keeping humans in control of strategy, judgment, and authenticity.

If we learn to collaborate better with our AI buddies, we can get a lot more out of them.

Process matters,

📝 Blog post from me

My last newsletter was about positioning and it feels right to continue the theme with the best book recommendation I can give you on that topic. April Dunford’s Obviously Awesome is a practical guide that takes you through the step-by-step of finding your place in the market.

Read the book summary here →

👌 Handpicked stories for you

Here are some interesting things I stumbled upon this week while reading, writing, researching, and being a generally curious marketing person on the internet:

💼 Three for work

Content Type Popularity and Engagement Trends [New Data] | An interesting analysis shows how important it is to talk about your pricing in your content and how to still make TOFU content work for you. While traffic data is useful, the conversion impact may be different.

Extreme brainstorming questions to trigger new, better ideas | This is an awesome set of exercises you can try out, especially with 2026 annual planning coming up.

How to Edit the AI-isms Out of Your Content (No Detectors Needed) | Here’s a great starting point for building your AI clichés blacklist. It’s a well-structured recap of the type of phrases that immediately sound AI-shaped alarm bells in your reader’s mind.

👀 Two for fun

This complex robot trap | This one gets addictive fast! It’s a series of robot captcha tasks with increasing complexity – you need to go to hilarious lengths to prove you’re human!

This beautiful deep dive into Japanese philosophy (alt link) | We may have just four, but Japan has 72 micorseasons, which demonstrate the true connection between human and nature.