The Top 10 Anti-AI Slop Skills
Top Insights
The Top 10 Anti-AI Slop Skills For Better AI Writing
At this point, your customers can smell AI writing in one sentence, and every "it's not X, it's Y" tells them you didn't write it. AI dev Juampi Hernandez of Firecrawl ranked his top ten skills that kill those tells. Each are free. And you can drop them into any harness like Codex or Claude Codex to instantly make your AI sound less like AI.
💡 Key Insights:
- Stop Slop grades your draft before you publish...it scores five traits, from directness to density, out of ten each, and a total under 35 sends it back for another pass.
- Humanizer makes the rewrite sound like you...paste a few paragraphs of your writing and it copies your rhythm, word choice, and punctuation quirks, without inventing a fact to fill a gap.
- Slopbeth catches the sentences that read clean but say nothing...it keeps every sourced fact, cuts unsourced claims, and flags the promises an edit changed, since a line made less true to fool a detector is worse than slop.
The New AI Divide: Who’s Using It and Where
WeWork's research team went through the Census Bureau's survey of American businesses to answer one plain question: nearly four years after ChatGPT arrived, how many companies use AI, and for what? The answer is one in five, mostly to write emails.
💡 Key Insights:
- Anyone who builds a real AI workflow is already ahead of most adopters...only 21.7% of U.S. businesses used AI in July, 85.4% of them for writing and editing, and 64.3% changed nothing about how they operate.
- Marketing is the door AI walks through first...Sales & Marketing leads every business function at 14.3%, and 62.8% of companies planning to adopt expect to start there.
- Most holdouts think AI doesn't apply to them...61.6% of the businesses sitting it out say it isn't relevant to their work, and 22% admit they don't understand what it can do. That's the objection any AI pitch has to answer.
🚀 WATCH: How These AI Copy Bots Are Producing World-Class Sales Copy 50X Faster Than Even The "BEST" Copywriters On The Market…
(Plus… They Don't Get Sick, Miss Deadlines, Or Ask For Raises Either!)
Watch the full AI Copywriting Tell-All Video Here
Prompt of the Week
This Week's Feature: 3 Astra Accelerators
GPT-6 Astra ships with its own instruction manual. The prompting guide OpenAI published alongside the model shows you the exact phrasing the model was trained to respond to, so you can get the top-tier version of Astra on day one.
We pulled three snippets that deliver the biggest upgrade. The first turns Astra into a finisher... hand it a task and it carries the whole thing through to done.
The second gives you cleaner, plain writing with less AI speak. The third makes sure your instructions always win, even when a skill or template says otherwise.
You can check out OpenAI's full Astra guidance here.
THE 3 FULL PROMPTS:
PROMPT #1 - Make GPT More Autonomous
"You should infer the user's intent and task scope from the instructions and prior conversation context. Your job is to bias towards action and carry the user's intended task to completion.
When the user expresses intent to perform new work or fix an existing issue, persist until the user's intended goal is complete. Progress autonomously towards the user's goal (e.g. creating isolated worktrees / checkouts if needed, resolving merge conflicts, read-only actions, creating draft PRs etc.) unless they are clearly destructive or irreversible."
PROMPT #2 - Eliminate AI Speak & Slop
"Avoid using slop words or phrases like "Bottom Line:" in conclusions, "delve," "foster," "leverage," "it's worth noting," "importantly," "Question? Answer." or "This isn't about X. It's about Y.", "genuinely" or hyphenated compound descriptions and adjectives. Do not use concluding summary statements such as "In short:..", "The simplest mental model is:...".
State the intended action directly. Avoid adding what you won't do, what will remain unchanged, or how you'll separate or categorize results. Do not use contrastive framing such as "X, not Y" or "X—not Y" that introduces an unprompted alternative that the user didn't ask about. Avoid invented compound labels like "exact-head checks" and "editorial-row layouts", vague qualifiers, and canned transitions; use plain verbs and prepositions to state the actual relationship directly."
PROMPT #3 - Better Instruction Following
"The user's instructions take precedence over guidelines provided in a skill. If explicit user instructions conflict with a skill's instructions, prioritize the user's instructions."
Tool to Try
Muse is Meta's new personal AI agent, a free app for iPhone and Android that also lives inside WhatsApp. Text it a task the way you would text a friend and it goes and does the thing... booking the appointment, filling out the form, sitting on hold with customer service... on its own secure computer, then keeps working in the background after you close the app.
AI Tool Highlights:
🖥️ A Computer of Its Own: Muse runs on a persistent, dedicated virtual machine with its own browser... so it books appointments, fills out forms, and handles customer service chats without ever touching your phone.
💬 Just Text It: Chat with Muse in the app or straight from WhatsApp... no dashboards, no prompt engineering, you simply say what you want done.
✅ You Approve the Big Moves: Muse asks before it sends an email, makes a purchase, or shares information with a connected app, and you can allow once, always allow, or deny... every action it has taken and everything it plans to do sits in an audit trail you can read at any time.
🔐 Your Cards and Logins Stay Out of Its Reach: Passwords live in a secure credential store the agent cannot read, with 1Password integration coming... at checkout it generates a one time card number so neither the merchant nor Muse sees your real card
🎯 Hand It a Goal, Not Just a Task: Give Muse something bigger, like planning a trip or fixing your inbox, and it builds the action plan, tracks progress, and quietly knocks out steps along the way.
One More Need-to-Know News Story
Three xAI Staffers Will Build a Company Live in 72 Hours Using Grok Bot
Tuesday morning, three SpaceXAI employees sit down with a blank whiteboard and one coworker, Grok Bot. By Thursday evening, Matt Palmer, Lauren Tan, and Roshan Sadanani should have a company, idea through working code... built on camera, nothing decided before the stream starts.
📋 The Details: Grok Bot is the agent product SpaceXAI put in beta on August 11, separate from the Grok chatbot on X. Each Bot gets a computer of its own, signs into your tools, and works until something needs your sign-off, and SuperGrok and Cursor subscribers have it.
The stream runs September 15 to 17, 8:30 to 6 Pacific, registration on Luma, and it keeps handing off to department sessions... marketing's land Thursday.
🎯 Why You Need to Know: It's a vendor demoing its own tool live and unedited, so the stalls air beside the wins. The department blocks are the useful part if you don't code, since sales, support, and marketing each get the agent on real work.
⚡ Your Move: Register and bring one campaign task you'd hand a junior hire to judge the demo against.
Mind Fodder
Thanks for reading.
Until next time!
The AI Marketers
P.S. Help shape the future of this newsletter – take a short 2-minute survey so we can deliver even better AI marketing insights, prompts, and tools.
[Take Survey Here]
