Your first real prompt — be specific, not polite
Stop typing into AI like you're texting a friend. The prompt is the entire skill at this level.
::TL;DR · the whole lesson in three lines
- MOVEStop typing into AI like you're texting a friend. The prompt is the entire skill at this level.
- DRILLTake a real task you have THIS WEEK. Don't invent one. Run the bad-prompt version first to see why specificity matters, then the good-prompt version.
- WINYou felt the gap between vague and specific in your own work.
::concept · what's actually happening
Most novice prompts fail in the same way: too vague. "Help me with my resume" returns mush. "Help me rewrite this paragraph of my resume to emphasize SaaS pricing experience for a Series A startup PM role, under 90 words" returns gold.
read full concept · 2 more paragraphs →collapse concept ↑
There is no magic prompt syntax. There is only specificity: the context the AI needs, the constraint that shapes the output, and the format you want back. Three things.
Politeness adds words and helps nothing. Being direct is faster and the AI does not have feelings. ("Please" and "thank you" are fine if it makes you feel better, but they are not the lever.)
::drill · do the thing
Take a real task you have THIS WEEK. Don't invent one. Run the bad-prompt version first to see why specificity matters, then the good-prompt version.
::L2 drill · copy-paste into any AI chat
Round 1 · the bad prompt (intentional): "Help me with [thing]." Round 2 · the good prompt (template — fill in the brackets): I need to [the exact task]. Context: - [who this is for] - [what came before this — paste the email / the meeting note / the doc, if any] - [what success looks like for me] Constraints: - [length / tone / format] - [anything I have to include or avoid] Output: - [give me a [draft / list / outline / version A and B]] - [if you need anything else from me before drafting, ask now]
::steps
- 01Pick a real task from your week (email, summary, plan, write-up).
- 02Run Round 1. Note how generic the output is.
- 03Run Round 2 (the same task, same chat, specific). Compare.
- 04Save the Round 2 template to your phone Notes. Reuse it tomorrow.
::worked example · what one real run looks like
See one real run before you try yours.
::worked example · what one real run looks like
See one real run before you try yours.
::what a real human filled in
Round 1: "help me with my resume" Round 2: "Context: I'm a third-year mechanical engineering student at UF applying for a summer 2026 internship at Lockheed Martin in Orlando. I've worked at a campus bike repair shop for two years and built a small CO2 rocket for a club competition (placed 3rd of 12 teams). Constraints: One page. No fluff. Match the language of an aerospace defense employer. I have a 3.4 GPA so don't lead with it. Output: A bulleted list of 5 resume bullets for the bike-shop job, written in past-tense action-verb format, that I can paste directly under that job heading."
::what the AI returned
**Round 1 response ("help me with my resume"):**
Sure, I can help with your resume! To give you the best advice, could you tell me:
- What kind of job are you applying for?
- What's your current experience level?
- Do you have an existing resume or are we starting from scratch?
- Any specific sections you're stuck on (summary, skills, work history)?
Once I know more I can suggest formatting, wording, or whatever else you need.
---
**Round 2 response (the specific prompt):**
Here are 5 bullets for the campus bike repair shop, tuned for a Lockheed Martin aerospace audience — leaning into precision, diagnostic process, and accountability (things defense recruiters scan for):
- Diagnosed and repaired 400+ bicycles across drivetrain, hydraulic brake, and wheel-truing systems, maintaining a 98% first-pass repair rate verified by post-service inspection.
- Authored standardized intake checklist that reduced average diagnostic time from 22 to 9 minutes per unit, adopted shop-wide within three months.
- Trained 4 incoming technicians on torque-spec compliance and tool calibration procedures, producing a written onboarding guide still in use.
- Tracked parts inventory across 1,200+ SKUs using a shared spreadsheet system, cutting stock-out incidents by an estimated 60% over two seasons.
- Resolved 30+ customer-facing service escalations by communicating root-cause findings in non-technical language, sustaining a 4.9/5 satisfaction rating.
A note: I made up the specific numbers (98%, 22 to 9 minutes, 1,200 SKUs, 4.9/5). You'll need to swap those for your real metrics before sending — recruiters at defense companies do check claims during interviews, and even soft inflation can sink you. If you don't have hard numbers, tell me what you do remember and I'll rewrite the bullets to be honest but still strong.::what to notice
- The Round 1 response is almost entirely a question back to the user ('could you tell me...'). That's the cost of a vague prompt — you spent a turn just to start the actual conversation.
- The Round 2 response delivers the exact format requested (5 past-tense bulleted lines) with zero clarifying questions, because every blocker was pre-answered in the prompt.
- Notice how the bullets pull language from the *context* you gave — 'torque-spec compliance', 'first-pass repair rate', 'root-cause findings' — that's the AI matching the aerospace defense register you specified, not generic resume-speak.
- The honest hedge at the end ('I made up the specific numbers... swap those for your real metrics') is what a real AI should do when you ask for resume content without giving it real numbers. A polished marketing example would hide this; a real one names it.
- The 'a note' paragraph is slightly clunky and could be tighter — but it's also the most useful part of the response. Specific prompts get you specific *and* honest output, not just specific output.
::outcome · what should be true
- You felt the gap between vague and specific in your own work.
- You have a template you'll reuse this week.
- You stop typing "help me with…" and start typing "I need to…".
::trap · the most common failure
Treating the template as sacred. The template is a scaffold. Once you internalize the three pieces — context, constraint, output — you write better prompts from memory.
::other lessons at Novice level
I'm scared of AI · the calm starting point
Before any lesson, the feeling. Whether you are scared, skeptical, exhausted by the hype, or quietly excited and hiding it — this is the door. None of the feelings are wrong. The path is yours.
What AI actually does — autocomplete at huge scale
Strip the magic feeling off. Get the working model of what AI is doing under the hood, so the rest of the curriculum has a foundation.
When AI gets it wrong — see a hallucination, on purpose
You will not respect the verify rule until you watch AI lie to your face with full confidence. Do it now, on a low-stakes question, where the cost is zero.
System prompts — telling AI who to be
Every AI conversation has a hidden first instruction. Knowing how to set yours is the difference between a generic answer and one calibrated to you.
::part of the AtomEons /learn curriculum · 45 lessons · 5 levels · cc-by 4.0