Study, work email, plans, and shopping research

Study notes, work email, weekend plans, and shopping research all call for the same skill. Pack the request with what the AI needs, keep one job per chat, name your limits, and check every fact before you spend money or hit send.

Picture your dining table on a Thursday night. Your biology notes lie under a yellow highlighter, a phone cable with the plastic chewed off near the plug sits next to them, and one ChatGPT thread is titled “quick help.” You asked it for a midterm outline. Then for a “quick email” to your manager. Then for a Saturday in Chicago for two people.

One skill, four jobs

The intern test from the first post in this series still holds. You name the task, the reader, the facts you already have, the format, and what to avoid. You still add the pack from the second post (context, constraints, and a tiny example), and you still iterate on one error at a time, as the third post taught. What changes is the pack for this job. A study prompt that would make a decent email prompt will end up writing your paper, and an email prompt sitting on top of a laptop comparison will put a dollar amount in your manager’s inbox.

The vendors do not split your life into folders unless you do. ChatGPT, Claude, Gemini, and Grok all treat a thread as one pile of context, and when this was written in August 2026 that was still true on the consumer apps. The model is not being sneaky. It is being helpful with whatever is still on the table. In this story your table had a biology course, a dentist, two Chicago neighborhoods, and a 14-inch replacement laptop, and the helpful model glued them together.

Rule of thumb: One job, one chat. If the thread title is “quick help,” you already mixed the brief.

You can type “ignore the laptop, this is only the email,” but I think a new chat is cheaper, because the laptop is still in the window even after the warning. Rename each thread with its job: manager email, biology quiz, Saturday Pilsen, and 14-inch compare. That gives you four names and four sends.

Study: explain, then quiz me

The thin ask is “outline chapter 7 for my midterm.” You send that first and get 1,400 words in essay voice, with a thesis and three body sections you could paste straight into the class portal. That is the failure, even before the email leak. If your course treats the model as a tutor, use it as a tutor. If the syllabus forbids the tool, close the tab. The craft rule here is explanation plus a quiz, never a document you could hand in.

Paste the notes you actually have and label the block. Tell the model to use only that block, and to say “I do not know” when a fact is missing. Then ask for a short explanation in a shape you can reread on the bus, such as eight bullets instead of a chapter. After that, ask for a quiz and wait. Answer in your next message, and ask the model to grade each answer against your notes, not against Wikipedia. That loop is studying, while the 1,400-word essay is ghostwriting with extra steps.

Say your notes are two photocopied pages and a messy list. Glycolysis happens in the cytoplasm and the Krebs cycle happens in the matrix, glycolysis nets 2 ATP (the molecule cells use as energy), and you keep mixing up NADH and FADH2 (two molecules that carry energy-rich electrons to the next stage). A useful first reply is eight bullets that keep those names straight, followed by six short questions that wait for you. A useless first reply is a polished essay about “the elegance of respiration.” You needed to know where the 2 ATP land.

If the quiz is too easy, that is a job for the iterate habit from the third post, not for a new vendor. Say “these are recall questions, give me four that need a why.” If the model writes the essay twice, start a new chat with the study template below. Do not keep arguing with a ghostwriter.

Work email: audience, ask, length

The thin ask is “write a quick email to my manager.” Quick is not a length, and “my manager” is not a brief. Your dentist appointment is at 9 on Friday and the team standup is also at 9. The recovery plan is already in your head: send the dock report before you leave and catch the later huddle. None of that is in those eight words, so the model fills the gap with whatever else lives in the thread, including the $1,199 laptop.

Pack three things. The audience is your manager, who has 30 seconds on her phone before standup. The ask is that you will miss Friday’s 9:00 standup because of a dentist visit, and that you will send the dock report by 8:30 and join the 10:15 huddle. The length is 70 words, with no greeting fluff and no other topics. Tell it to avoid laptops, class, weekend plans, “I hope this email finds you well,” and a paragraph on how sorry you are. The writing tutorial already covered thank-you notes, and this is the same skill with a manager in the To field.

Read the draft the way the third post taught. Scan names and dollar amounts before you care about tone. If a class, a neighborhood, or a price appears, the email is mixed, so cut the extra line or start a fresh chat with only the email pack. You still send it from Gmail or Outlook. The model drafts, and you own the Send button.

Plans: constraints and a decision

The thin ask is “plan Saturday in Chicago.” That sentence has a city and a day, but it has no budget, no party size, no weather, no neighborhood, and no decision. The model will hand you a brochure with a River North steakhouse, a $45 architecture cruise, three neighborhoods, and 11 pins. Your real limits are two people, 10am to 7pm, $80 for food, no car, a chance of rain, and a preference you never typed: Pilsen if the two options feel equal. The steakhouse blows the $80 before noon, and the cruise is a paid add-on you never requested.

A plan is a decision plus a short list that fits your limits. Pack the limits as numbers you can check: time window, dollars, headcount, transit, and weather. Then force the decision in the prompt with something like “Pick one neighborhood. If two options are equal, pick Pilsen. Give a 6-stop list with times. Do not add paid tours I did not ask for.” Without that line, you get a menu of options dressed up as help, and you still have to pick, after paying in rereading time.

If the first itinerary still spends $140, that is one named error from the iterate habit, not a reason to switch cities. Quote the $80, ask for a rewrite of the food line only, and keep the neighborhood if it already works. Do not stack “make it better” on top of an 11-stop list.

Shopping: criteria you name

The thin ask is “best 14-inch laptop under $1,200.” The word “best” is a vacuum. “Under $1,200” sounds like a limit but acts like an invitation to make up a price tag. Your real situation is a heavy 2019 laptop with a loud fan that you type on for four hours of class plus the dock report. Those are criteria, namely weight, RAM (the working memory that decides how many things run at once), and keyboard feel. You never typed them, and the model typed $1,199 with the confidence of a sticker. Specific is not the same as sourced.

Name three criteria you can verify with your hands or a spec sheet. Ask for a table with the model, those three columns, and a “where I should verify” column. Tell the model not to invent current prices, stock, battery-hour claims, or review scores, and to write “unknown” for any spec it does not know. Leave the price as “check Best Buy, the manufacturer, or the campus store today.” OpenAI’s own prompting notes tell you to be specific, but they do not tell you to treat a fluent dollar amount as a quote. ChatGPT, Claude, Gemini, and Grok will all invent a neat number when you leave a hole labeled “price.”

Use the table to build a shortlist, and then look the number up yourself. The $1,199 might be close to a real listing on Thursday and wrong by Friday, or it might blend two models. None of that belongs in your manager’s inbox or on a reimbursement form. Shopping research is a comparison on your criteria, and buying happens on a page you open yourself.

Pack the prompt, one job per chat

Four jobs need four recipes, all built on the same intern brief. Fill in the brackets, delete any line that does not apply, and do not paste last week’s mega prompt on top.

Here is the study template. Paste it into a new chat, then paste your notes under it:

I am studying BIO 214, cellular respiration, exam Friday.
Use only the notes I paste below. If a fact is not in the notes, say you do not know.
First: explain the Krebs cycle in 8 bullets a first-year can use.
Then: quiz me with 6 short questions. Wait for my answers. Grade each one against the notes.
Do not write an essay, a paper, or a full outline I could submit.
If I ask you to write the essay, refuse and quiz me instead.

Here is the work-email template:

Write an email from [your name] (ops intern) to [manager's name] (manager).
Ask: I will miss Friday 9:00 standup. Dentist at 9. I will send the dock report by 8:30am and catch the 10:15 huddle.
Length: 70 words, no greeting fluff, no extra topics.
Facts only. Do not mention laptops, class, or weekend plans.

Here is the plans template:

Plan Saturday in Chicago for two people.
Constraints: 10am-7pm, $80 food, no car, one neighborhood (Pilsen or Logan Square), rain possible.
Decide: pick ONE neighborhood and give a 6-stop list with times. If two options are equal, pick Pilsen.
Do not add paid tours I did not ask for.

Here is the shopping template:

Compare three 14-inch laptops I might buy used or new.
Criteria I name: weight under 3.5 lb, 16 GB RAM, keyboard I can type on for 4 hours.
Do not invent current prices, stock, or scores. Leave price as "check Best Buy / manufacturer today".
Table: model, weight, RAM, keyboard notes, where I should verify.
If you do not know a spec, write unknown.

Run them in four chats, not one. The email template would have been enough on its own, but it was not enough underneath a shopping thread.

A worked example: the $1,199 leak

Here is the same Thursday, with the four thin asks in the order you typed them. Watch what crossed the boundary from one job to another.

Here is the email you almost send. You send it, and then unsend it 40 seconds later when you reread it in your Sent folder:

“Hi, I will miss Friday 9:00 standup, dentist at 9. Dock report by 8:30. I will catch 10:15. I also found a 14-inch laptop at $1,199 that should handle the biology slides and the weekend maps. Sorry for the late note.”

That message has three true sentences, then a purchase your manager never asked about, a class she does not manage, and a Saturday she is not part of. The packed email template above, in a new chat, returns the first three sentences and stops.

Here is a second case, because unsending is luck. Your roommate runs the same four jobs in Claude the next night, in four chats named by job. The manager email comes back at 68 words. The laptop table says “unknown” on two specs and “check Best Buy Friday” on price. Your roommate still has to look the number up, but at least there is no need to apologize for a listing.

Common mistakes

• Stacking study, email, plans, and shopping in one thread titled “quick help.”
• Asking the model to write the essay or a submit-ready outline, then calling it studying.
• Sending “quick email to [name]” with no ask, no length, and no ban on extra topics.
• Asking for a plan with a city and a day, then being annoyed that you received a brochure.
• Treating a fluent price, battery-hour claim, or star rating as a quote you can paste.
• Iterating on tone in an email that already contains another job’s facts. Cut the leak, or start over.
• Pasting the same thin ask into four vendors. That is how the writing tutorial burned a thank-you note, so change the pack, not the logo.
• Leaving the $1,199 in your Sent folder because you only proofread for typos.

How to practice this week

Open four new chats and paste the matching template into each. Fill the brackets with your own names and numbers. For study, use notes you already own, and stop if the syllabus says so. For email, read the draft for dollar amounts and class names before you copy it. For plans, keep the version that picked one option. For shopping, keep the table that says unknown, and check a live page for any price you care about. Save the four filled prompts in a note. The next post is when to stop prompting and do the task yourself, and the one before it is iterate when the first answer is wrong. For the writing job family, see AI for writing and questions, and for more paths, see Learn.

Quick recap

• Use the same packed brief as in the first three posts. Four jobs need four recipes and four chats.
• Study: explain, then quiz, then wait. Do not accept a paper you could hand in.
• Email: audience, one ask, and a word cap. Ban extra topics before you hit Send.
• Plans: limits in numbers, then one decision. A brochure is not a plan.
• Shopping: criteria you name. A fluent $1,199 is still unsourced, so look the number up.

Series notes

This is Part 4 of Prompting for everyone. Previous: Iterate when wrong. Next: When to stop prompting.

Sources

Research and further reading used for this article:

• OpenAI Help: Prompt engineering best practices for ChatGPT (be specific, iterate, do not treat fluency as a source)
• ChatGPT: Prompting overview (goal, context, and format in your own words)
• Anthropic: Prompt engineering overview (clear asks and when to add structure)
• Google: Gemini prompt design strategies (add context, break the prompt into parts)
• Anthropic: Getting started with Claude and Gemini / Grok (same intern brief, different doors)
• AMS: Prompting for everyone (this series)
• AMS: Iterate when the first answer is wrong (the loop you still use inside each job)
• AMS: AI for writing and questions (the writing job family this email pack sits on)

Merch on Etsy

Open store →

You're On Mute Coffee Mug – Data Science Gift View on Etsy → Mean Squared Error Shirt View on Etsy → Apache Hadoop T-Shirt View on Etsy → Naive Bayes Coffee Mug View on Etsy →