Happy New Year, and goodbye to 2025‌ — ‌when Merriam-Webster crowned “slop” as its word of the year. I enjoy a sloppy joe sandwich as much as the next guy, but this type slop is decidedly less appealing, defined as” the avalanche of low-effort, low-quality content you can now generate with AI practically for free.

Use of the AI in the workplace is growing rapidly, but simply making loads more content won’t help your team increase productivity or make an impact if it’s all kind of … blah. With AI tools, I — a person who can’t draw a stick figure — can whip up illustrations, videos, slide decks, and some theme music to go with them at the push of a button.

The challenge isn’t resources, time, or headcount — ‌it’s execution, style, substance, and trust.

So how do you use AI to actually get work done without drowning in a sea of mediocrity? I recently turned a 38-page research PDF into a full multi-channel campaign in under ten minutes. Not by typing a single prompt into a chatbot, but by following five simple rules that separate signal from noise.

Let’s talk about what they are‌ — ‌and why you might be doing this all wrong.

Rule #1: Your prompt isn’t a one-night stand

Here’s what most people do when they get started with AI: they write a prompt, hit enter, copy the output, and call it a day. Here’s what actually works: treating your prompts like living documents that evolve and improve over time.

At WRITER, we use something called Playbooks‌ — ‌think of them as prompt templates on steroids. They let you refine your instructions iteratively, add variables for repeatability, easily share improvements with colleagues, and build something that gets better every time you use it. In the video, I show you exactly how Playbooks work and why they’re the secret to consistent, high-quality outputs.

Rule #2: Generic is the enemy of great

As soon as readers see a bullet-point list, each section topped with an emoji for emphasis,  dozens of em dashes dotted across the page, they recognize it as generic AI content and tune out.

Most default AI voices sound  like a very polite robot trying not to offend anyone. Neutral. Safe. Boring. If you want content that actually lands, you need to teach your AI to speak in voices that match your brand, your audience, and your context.

There’s a world of difference between “executive communication for pharma clients” and “playful product marketing for startups.” The question is: how do you encode that? I show you how to fine-tune voices and how Playbooks make it easy to apply the same agents and automated workflows to different topics with different voices, all without rewriting a single line of your prompt.

Rule #3: The iceberg principle

What you see in a prompt is just the tip. What powers great AI outputs is everything beneath the surface: your knowledge bases, your style guides, your previous campaigns, your legal guidelines, your product messaging.

In the video demo, I’m working with knowledge graphs and connectors‌ — ‌layers of context that pull from your team’s ground truth and external tools like HubSpot or Salesforce. It’s like the difference between asking a newcomer for directions and asking someone who’s lived in the neighborhood for years. Context transforms generic into good.

Rule #4: Never ship the first draft

Think of your AI as a very eager intern. Enthusiastic, capable of following instructions, but definitely not ready to publish without supervision. The first output is raw material, not the finished product.

In my demo, the AI generated a presentation that was… functional. Then I gave it our brand guidelines and previous slide decks, and it created something that actually looked like it came from WRITER. That iteration, from “meh” to “nice”, is where slop goes to die. I walk through that transformation in the video, but the principle is simple: review, refine, repeat.

Rule #5: Keep doing what you love

Here’s the existential question: as AI gets better at so many tasks, what work should you keep, and what should you give to AI?

For me? I love deep research, interviewing fascinating people, and writing essays. I do not love remixing that essay into twelve different formats for SEO and sales enablement. Luckily, AI is phenomenal at the latter.

Decide what energizes you and what drains you, then architect your workflow accordingly.

The video walks through a complete example of this philosophy in action‌ — ‌turning one asset into dozens while keeping human creativity and control at the center.

In the full video, you’ll watch me take that 38-page PDF and transform it into LinkedIn posts, Twitter threads, email sequences, and a presentation‌ — ‌all while applying these five rules in real time. You’ll see what playbooks actually look like, how knowledge graphs supercharge your outputs, and why iteration beats perfection every time.

If you want to learn more about how we use Playbooks or the suite of agentic solutions we’ve built for marketing teams.