How to AI

AI repeats the same reframe trick in 12 sentence shapes. Here's how to audit any draft in under 2 minutes: The 4-step workflow: Step 1. Open the AI output you want to audit. Step 2. Scan it against these 12 shapes. Step 3. Find the audit prompt that runs this automatically → https://lnkd.in/gb-6qd57 The 12 patterns AI uses: ✦ "This isn't X. This is Y." ✦ "Not X. Y." ✦ "Forget X. This is Y." ✦ "Less X, more Y." ✦ "Not only X, but also Y." ✦ "It's not just about X, it's about Y." ✦ "X? No. Y." ✦ "Stop thinking X. Start thinking Y." ✦ "X is dead. Y is the future." ✦ "The question isn't X. The question is Y." ✦ "You don't need X. You need Y." ✦ "X is overrated. Y is what matters." 1. For every match, delete the rejected half. Example: Bad: "It's not about the prompt. It's about the context." Step 3a: "It's about the context." 1. Rewrite the positive claim as a direct sentence. Example: Step 3a: "It's about the context." Step 4: "Context controls the output." Run this on every long-form AI output. The same workflow works on your own writing. ♻️ Repost this so your network can stop sounding like AI.

You're prompting Claude for slides the wrong way. Here's the (exact) skill I built before I prompt: Step 1. Open Claude. Step 2. Go to "Customize." Step 3. Select "Skill" and click "+". Step 4. Choose "Create skill." Step 5. Pick "Write skill instructions." Step 6. Paste the deck-builder skill. Step 7. Add the title and instructions, then save. Claude writes the full outline, sends it to Gamma. Gamma generates the entire presentation. 13 slides. Designed. Laid out. Ready to present. You don't even open Gamma.  Claude handles it through the Gamma connector. You give the topic. Gamma gives you the deck. To copy-paste this skill (and 4 others):  ↳ Go to how-to-ai.guide  ↳ Don't pay anything. It's free in the welcome email. ↳ Open the 'How to AI' library. ↳ Click "Claude Cowork". ↳ Pick from 5 skills. ↳ Save them & share with your team. ♻️ Repost to help anyone make slides in Claude.

You already know the obvious "It's not X, it's Y" pattern. Here are 3 sneaky versions that get past most people. 1. "While X might seem right, Y is actually..." Real example: "While prompt engineering might seem critical, context is actually what controls the output." Scan for "while" + "actually" in the same sentence.  2. "Sure, X works. But Y is where the real..." Real example: "Sure, ChatGPT works. But Claude is where the real depth lives." Scan for "Sure," at the start of a sentence followed by "But" in the next. 3. "X gets all the attention, but Y is what actually..." Real example: "Models get all the attention, but context engineering is what actually moves the needle." Scan for "gets all the attention" plus "but" plus "actually." Find the full list of 12 patterns and the audit prompt → https://lnkd.in/gb-6qd57 Here's the 2-step fix: Step 1: Delete the rejected half. Step 2: Rewrite the positive claim as a direct sentence. Worked example: Bad: "It's not about the prompt. It's about the context." Step 1: "It's about the context." Step 2: "Context controls the output." Apply this to any draft. AI output, your own writing, copy you're editing. ♻️ Repost this so your network can stop sounding like AI.

Claude can turn any document into a slide deck. But there's a better way to do it. Here's how: → Paste your source content into a markdown file. → Go to claude .ai/design and open "From template" mode. → Upload the file. Send Prompt 1 below. → Wait for the animated video to generate. → Send Prompt 2 in the same chat. → Export the deck to PPTX or PDF. Find the whole process here → https://lnkd.in/eGnArUbB Prompt 1: Make a 30-second animated video that summarizes this [blog/report/document] for a first-time viewer. Prompt 2: Now convert that video into a slide pitch deck. Slides built from a video inherit its structure.  You get better hierarchy, cleaner pacing, sharper frames. Asking for slides first skips all of that. ♻️ Repost this so your network doesn't miss this workflow.

How to find the 8 AI writing habits most people miss: Each one takes 30 seconds to fix. 1. AI lists 3 things by default. "Speed, efficiency, and innovation." Use 1 thing if 1 thing matters. Use 2 or 4 if that's true. 2. AI inflates everything. "A pivotal moment." "A seismic shift." Say what happened. Let the reader decide if it's a big deal. 3. AI tacks fake analysis onto sentences. "Highlighting its importance." "Underscoring its significance." Delete the phrase. Or rewrite it as a full sentence with a real point. 4. AI uses fake ranges. "From ancient traditions to modern innovations." If you can't name what's in the middle, the range is empty. Cut it. 5. AI renames the same thing 4 times. "Claude" becomes "the model," then "the assistant," then "the chatbot." Just say Claude again. 6. AI avoids the word "is." "The report serves as a guide." "The app boasts a dashboard." Write "is" or "has" instead. 7. AI uses Title Case in Headers. Humans use sentence case. 8. AI writes every sentence the same length. Real writing breathes. Mix short sentences with longer ones. Drop in a fragment. Find the full guide here → https://lnkd.in/gb-6qd57 ♻️ Repost this so your network can stop sounding like AI.

You don't need better prompts to fix AI-sounding output. You need one file in your Claude folder. Here's the 5-step setup: 1. Download the anti-ai-writing-style file from the full guide below. 2. Upload it to your Claude Cowork folder under the ABOUT ME subfolder. 3. Open Global Instructions in Claude Cowork settings. Tell Claude to read the file before any task. 4. Test it on a fresh chat. Ask Claude to write anything. It should read the file before answering. 5. To audit text Claude already wrote, paste the output back and run: "Audit this text against the anti-ai-writing-style .md file from my folder." If the rewrite still sounds robotic, reply: "Edit harder. Find the tells you missed." Once it's installed, every chat starts cleaner.  No prompt engineering. No begging Claude to "sound human." Find the full setup → https://lnkd.in/gb-6qd57 ♻️ Repost this so your network can stop sounding like AI.

Stop telling Claude what to do every chat. Set it once in Cowork's Global Instructions. Here's how to set it up in 5 steps: Find the full workflow here → https://lnkd.in/gb-6qd57 1. Open Claude Cowork settings. 2. Find Global Instructions. 3. Paste this block: "I usually start my Cowork session by pointing you to my Cowork folder. Before any and every single task, you must read every file in ABOUT ME/: - about-me: who I am, what I love and hate - anti-ai-writing-style: audit any text you produce against this file - my-company: where I work, my role Never read the folders OUTPUTS/ or TEMPLATES/ unless I specifically point you to a file. Save all deliverables in OUTPUTS/ under a subfolder named after the project. If the brief is unclear, use AskUserQuestion. Don't over-explain. Deliver the work." 1. Save the instructions. 2. Test it. Ask Claude any task. It should read your folder before answering. Now every chat starts with full context. Your style, your role, your file system. ♻️ Repost this so your network can stop repeating themselves to Claude.

Writing emails from scratch in 2026 is embarrassing Here's what I do instead (only takes 2 minutes): Step 1. Install Granola. ai. It's free. → It silently transcribes every Zoom/Teams/Meet call you join. No bot creeping in your meeting. → People talk normally & you get the real conversation. Always open Granola when in calls. (To gather the context of all your meetings.) Step 2. Open the Claude desktop app. → Settings > Connector > Connect Gmail + Granola → Claude can now read your inbox AND meeting transcripts at the same time. → Most people connect just Gmail. It's not enough. → Without Granola, Claude drafts emails with zero memory of what was actually said in the meetings. Step 3. Open the email you need to reply to. → Note the sender + the meeting it ties to. → This step is what makes the prompt 10x sharper. → You're telling Claude exactly where to look in Granola so it doesn't guess. Step 4. Prompt Claude (with context): "Pull [sender]'s email about [topic]. Search Granola for our meetings on [topic] in the last [timeframe]. Draft a reply that [confirms / answers / declines]. Tone: warm + direct. Length: short." Step 5. Ask Claude to "save as draft." → It lands in your Gmail. 0 copy and paste. You just open Gmail, review it, and hit send. → The whole loop happens without leaving Claude. Pro tip: Don't grant write access on day one. Read-only first. Watch how Claude uses it for a week, then upgrade. Most accidents happen when you enable Claude with full access & asks to "clean up my inbox." To download all of my other Claude infographics: Step 1. Go to how-to-ai.guide. Step 2. Subscribe for free. Don't pay anything. Step 3. Open my welcome email (most skip this). Step 4. Hit the automatic reply button inside. Step 5. Download my infographics from my Notion. Bonus. Enjoy my best copy-paste prompts, too.

You can now turn any meeting into a deck with AI. Here's the exact 10-minute workflow: Most people spend 2 hours writing the recap, formatting the slides, and chasing action items. What you actually need: → Granola for the meeting notes → Claude for the outline of the deck → Gamma to build the deck Here's how to run it: 1. Open Granola before your meeting starts.  It runs in the background and transcribes automatically across Zoom, Google Meet, and Teams. No bot joining the call. 2. After the meeting, go to claude .ai.  Click "+". Find the Granola connector. Switch it on and complete the 1-click authentication. 3. Send this prompt: "Pull my latest Granola meeting notes. Extract all key decisions, action items, and owners. Identify any open questions or blockers. Then build a presentation outline structured for a follow-up deck." 4. Claude reads your actual meeting notes and returns a structured outline.  Review it. Ask it to adjust anything before moving to the next step. 5. In the same chat, click "+" again.  Find the Gamma connector. Switch it on. 6. Send this prompt: "Use that outline to generate a full presentation in Gamma. Keep it clean, skimmable, and structured for a professional audience." 7. Gamma builds the deck.  Click Present to preview the live version. Export to PPTX or share the Gamma link directly with your team. More (free) guides at how-to-ai.store