💪 **Workout Preview — Thursday** There's a person on every team who walks in already understanding the thing nobody else had time to read up on. The new regulation. The competitor's product. The acronym in the email everyone's nodding along to. Today's Pro Workout is a 30-minute hands-on build to become that person — on a real topic from your actual job. Here's how it starts. **Step 1 — Pick the topic (5 minutes)** Don't pick something hypothetical. Pick the thing that's actually been sitting in the back of your mind at work. Ask yourself: *What do I keep nodding along to without fully understanding? What would I look smart for knowing next week?* Good real topics: • A regulation or policy hitting your industry • A competitor, tool, or product everyone keeps mentioning • A concept that keeps coming up in meetings • A market or trend your boss cares about Write your topic on a sticky note in one line. That's Step 1 done. Then the next step in the Workout takes that topic straight into your AI tool, with a prompt that gets you the whole landscape in under two minutes — tailored to your actual role, not a generic overview. Pro members got the full 6-step session today — a complete research dossier, ready to share with your team, built in 30 minutes including the step that makes the AI show you its own weak spots.

You don't have to read the whole thing to know if it matters. That's the move this week. You're going to paste any wall of text into an AI tool and get a plain-English summary you can actually use — in under a minute. Reports, articles, long emails, policy docs, meeting notes. If you can copy it, you can summarize it. **What you need** An AI assistant. Two good free options: • ChatGPT — OpenAI's AI assistant, free at chat.openai.com. Open it, click "New chat." • Claude — Anthropic's AI assistant, free at claude.ai. Open it, click "Start new conversation." Either works for today's drill. **Do it — 4 steps** 1. Find something you've been meaning to read but haven't. An article, a PDF page, a long email — anything. Copy the text. (Select it all, then Ctrl+C on Windows or Cmd+C on Mac.) 2. Open ChatGPT or Claude. Click into the message box at the bottom of the screen. 3. Type this prompt exactly — then paste your text right after it: ``` Summarize this in plain English. Give me: (1) what it's about in one sentence, (2) the 3 most important points, and (3) whether I need to do anything as a result. Here's the text: [paste your text here] ``` 4. Press Enter (or click the send arrow). Read the output. **The win** In about 60 seconds you'll have a clear answer to the only question that matters: *does this need my full attention, or not?* That's a filter you can use every single day — on anything. **Share your result** Run it on something real from your week. Then drop your output (or just the topic you summarized) in this week's Drill Thread. You'll earn likes from the community, and likes turn into points that move you up the Skool Levels board. This is 1 of 3 drills Pro members got this week. The other two: Explain It Like I'm 12 (fastest path through a confusing topic) and Get the Counter-Argument (asking AI to argue against its own answer, so you never walk into a meeting with only one side of the story).

⚡ **Daily Rep — Tuesday** One sentence to memorize: *reach for AI when you want thinking done, reach for Google when you need something current or verifiable.* AI is a thinking partner — it synthesizes, explains, and reasons through nuance. Google is an index — it finds live pages, real URLs, and today's information. The trap: AI will sometimes give you a phone number, citation, or URL that *looks* exactly right but doesn't exist. It's not lying — it's pattern-matching without a fact-check. This is called hallucination. The fix is simple: any specific fact, number, or source that matters gets a 30-second Google check before you use it. That one habit is the sorting logic most people never learn. Tuesday's Pro teardown goes deeper — it covers three worked examples of the decision in action, a full map of when to use Perplexity (an AI tool with live web search built in), and exactly how to turn on web-search mode in ChatGPT and Claude so you're not missing live information when you need it.

Four things that happened in AI last week, and what they mean for people who have actual jobs to do. **1. Google's research tool now browses the web so you don't have to** NotebookLM (a free Google tool — think of it as a smart notebook that reads documents for you) added a "Deep Research" mode. You type a topic, and it searches up to 50 websites, pulls the best sources, and writes you a summary while you get on with something else. *Why it matters:* Staying current on a topic used to mean an hour of browser tabs. One typed question can now hand you a briefing most people would take a morning to put together. *What to do:* Go to notebooklm.google, sign in with your Google account, click "New Notebook," and look for the Deep Research option. Type a topic you've been meaning to catch up on. **2. Half of workers say AI is making them less sharp — and they're not wrong to worry** A survey of 2,500 professionals found roughly half worry that heavy AI use is dulling their own thinking. Nearly 4 in 10 said they feel less intelligent because of it. *Why it matters:* The professionals who stay valuable aren't the ones who use AI most — they're the ones who use it to sharpen their own conclusions, not replace them. *What to do:* Next time AI gives you an answer you plan to use, take 60 seconds and ask yourself: does this actually hold up, or am I just accepting it because it sounds right? **3. Microsoft quietly put a research assistant inside Office** Copilot (the AI built into Microsoft 365 — Word, Teams, Outlook) now includes a "Researcher" feature that pulls information together and organizes it into a shareable page your whole team can contribute to. *Why it matters:* If your organization uses Microsoft 365 (common in corporate and mid-size business environments), you may already have a research tool inside the apps you open every day. *What to do:* In Microsoft 365 Copilot Chat (look for the Copilot icon in your Microsoft apps), type: "Research [topic] and create a summary page." Not sure if you have access? Ask your IT team.

This week in AI Express, we made you sharper on the page. Here's what landed: 📰 **Monday — News in 60:** The research is in — authentic voice is now scarce, and that makes it more valuable. Three tools to know: Wispr Flow's Transforms feature, Microsoft Word's Coaching with Copilot, and the AI-writing plateau that's good news for people who write like humans. ⚡ **Tuesday — Daily Rep:** Speaking is faster than typing. The voice memo → transcript → AI cleanup pipeline is the move. 60 seconds of talking, clean email out the other side. ⭐ **Wednesday — Drill of the Week:** One prompt, two sharper versions of any message. If you haven't run the clarity drill yet — it's still in the feed. Takes two minutes. Worth it. 💪 **Thursday — Workout Preview:** Step 1 of training your AI on your voice: gather 3–5 real samples of how you write. That's the move. Your voice is the asset — AI just needs to learn it. 🔁 **Friday — Daily Rep:** Long doc you haven't read? Paste it, ask for one paragraph, then ask what got left out. That second question is the trust move. 📋 **Saturday — Steal This Preview:** Three paste-ready document prompts — one for reports, one for summaries, one for decks. Ready in the feed, copy and go. **Pro members got more this week.** The full voice-to-text pipeline (5 steps, free tools, exact cleanup prompt). Two more drills — tone-matching across four named registers and a verified doc-cutting technique. The full 25-minute session to build your Voice Card and save it as a Custom GPT or Claude Project. Ten paste-ready document prompts across reports, summaries, and decks. A career-angle piece on how to ship more visible work without triggering the "lazy AI" flag managers already know how to spot. And a 5-day challenge: rewrite one real thing you have to write each day this week. Next week — *AI as Your Researcher*: learn anything 10x faster, and know when AI is wrong. See you Monday.