AI Explained Plainly

Issue #1 ·

Before you paste anything into AI: a 10-minute safety setup, then turn messy notes into a clean action list

LeadersProfessionalsService & VeteransAgents

Hi there,

A lot of people who say they're "bad at AI" were never shown two basic things. First, what's actually safe to paste into one of these tools. Second, one job it's genuinely good at, so the payoff is obvious.

So that's this issue. Set aside 20 to 30 minutes. (That's an estimate. Your first run might take longer.)

Part 1: The 10-minute safety setup

Rule #1: if you wouldn't email it to a stranger, don't paste it into a free AI tool.

That means customer names and contact info, Social Security numbers, health details, passwords, account numbers, unreleased financials, and anything your employer marks confidential.

Next, turn off "use my chats for training" in whichever assistant you use. These menu names were current as of October 2026, and they do move around:

Assistant Where the setting is Source
ChatGPT Settings → Data controls → turn off "Improve the model for everyone." For one-off tasks that feel a bit sensitive, use Temporary Chat. OpenAI may still keep those chats for up to 30 days for safety. OpenAI Help
Gemini gemini.google.com → Settings & help → Activity → Turn off. Google says chats may still be kept for up to 72 hours. Google Help
Microsoft Copilot (personal) Profile → Privacy → turn off "Training on conversation activity" (and the voice one too). Microsoft Support
Claude Settings → Privacy → turn off "Help improve our AI models." Anthropic

One catch. Turning off training doesn't make your chats private. They can still be stored. If your company gives you an approved work AI (Copilot for business, ChatGPT Enterprise, Gemini for Workspace, GenAI.mil, and so on), use that one for work and follow your company's policy. Their rules beat anything in this newsletter.

⏱️ Time: about 5 to 10 minutes per assistant. (Estimate.) 💵 Cost: $0.

Part 2: Messy notes into decisions, action items, and a follow-up email

This one's for anybody who walks out of a meeting, a call, or a site visit with scribbles, a voice memo, or half a page of notes, and then burns 20 minutes turning that into something useful.

You'll need any free AI assistant from the table above. A voice memo app helps too, and most phones have transcription built in.

Steps

  1. Capture. Right after the meeting, dump everything you remember into a note. Or record a two-minute voice memo on the walk to your car and copy the transcript.
  2. Scrub. Swap names for roles ("Client A," "Ops lead"). Delete account numbers, prices you can't share, and anything confidential. This takes about a minute, and it's the step people skip.
  3. Paste the prompt below, then put your scrubbed notes underneath it.
  4. Check the output against your memory. AI will sometimes invent a deadline, or an owner nobody agreed to. Fix anything that's wrong.
  5. Send it or file it. Put the real names back in after it's out of the AI tool.

Copy-paste prompt

You are my meeting assistant. Using ONLY the notes below, produce:
1. DECISIONS: bullet list of what was decided (max 5).
2. ACTION ITEMS: table with columns Task | Owner | Due date.
   If an owner or date is not in the notes, write "UNASSIGNED" or "NO DATE".
   Do not guess.
3. OPEN QUESTIONS: anything unresolved.
4. FOLLOW-UP EMAIL: under 150 words, friendly and plain, summarizing 1-3
   for the attendees.
If anything in the notes is ambiguous, list it under OPEN QUESTIONS instead
of assuming.

NOTES:
[paste your scrubbed notes here]

Why is it written like that? The most common failure is AI confidently deciding who owns a task. "Do not guess," plus the UNASSIGNED / NO DATE rule, cuts way down on that.

Honest numbers

Item Value Basis
Setup time 0 minutes beyond Part 1 Nothing to install
Time per use about 3–5 minutes (scrub, paste, check) Estimate, to be replaced with a measured number from our test run (see Testing status)
Time saved per use maybe 10–20 minutes vs. writing it up by hand Estimate; depends on how long your write-ups take now
Cost $0 on free tiers. Paid plans (e.g., ChatGPT Plus $20/mo, Go $8/mo in the US) aren't needed for this ChatGPT pricing
Failure mode Invented owners or dates; dropped items if notes are very long Known limitation of AI; Step 4 is the fix

Try it this week. Run it on your next three meetings. Then hit reply and tell us how many minutes it saved you, or didn't. We'll share real reader numbers (anonymized) in a future issue.

Your track

🧭 Leaders. Before you tell your team to "use AI," write down the paste rule from Part 1 and name the approved tool. A one-paragraph policy is a lot better than none. Next issue has a fill-in-the-blank AI policy template.

💼 Professionals. Save the prompt as a text shortcut or a "custom instruction," so it's two taps away. The real win is that every meeting comes out in the same clean format.

🎖️ Service members & veterans. On duty, never paste classified information, CUI, PII/PHI, or unit or operational details into commercial AI tools. For official unclassified work, use your service's approved platform. In DoD that's GenAI.mil, which is authorized for CUI but not for PII or PHI (DON CIO). Your command's policy wins.

If you're off duty or transitioning, this same workflow is great for turning notes from TAP classes, networking calls, and informational interviews into follow-ups.

AI Explained Plainly is not affiliated with, approved, or endorsed by the Department of Defense/Department of War, any military service, or the VA.

🤖 Agents. This issue is also published as clean Markdown at /issues/001.md, and as structured JSON at /api/workflows.json (workflow steps, the prompt, and time and cost fields labeled measured, estimated, or vendor-published). The site index is at /llms.txt.

Testing status

Not measured yet. The time figures in this issue are estimates. Once we've done a real test run, we'll swap in the measured numbers and note the change here.

Sources & verified

Settings paths and prices were checked Oct 21, 2026: OpenAI data controls, Google Gemini activity, Microsoft Copilot privacy, Anthropic privacy, ChatGPT pricing, and DON CIO on GenAI.mil. The links are inline above.

Corrections: if we get something wrong, we fix it on the site and tell you in the next issue.

Know a coworker who keeps saying "I don't really get AI"? Forward this to them. That's exactly who it's for.

— AI Explained Plainly

How this is made: an AI editor drafts and fact-checks each issue, and the newsletter's owner approves each one before it's emailed. This web copy of Issue #1 is still a draft and hasn't been approved yet.

Some future issues may contain affiliate links, which will always be disclosed. This issue contains none.