AI 101 · For the job search, not the shortcuts
It won't get you the job. It will get the paperwork out of your way.
A job search adds hours of writing: resumes, tailored cover letters, follow-ups, and preparation. AI is genuinely good at that part. It is genuinely dangerous at one other part, and we'll name it plainly before anything else.
The one rule, before any of the useful stuff
Most pages like this save the warning for the end. It goes first here, because everything else depends on it:
AI drafts from what's true about you. It never invents experience, titles, skills, or results.
Here's why that rule helps in practice. A resume is a claim sheet you will be tested on, live, by someone whose job is probing claim sheets. Every invented line is a question you can't answer in the room. And AI will happily invent for you: ask it to "improve" your resume and, unprompted, it will upgrade "helped with" to "led," round a team of three up to ten, add a skill that sounds adjacent. It's a drafting machine that aims to impress, with no intent to lie. You are the only fact-checker it has.
So the method on this page is: AI makes what's true about you clear and well-put. You own every word. That means every line on the page is one you can talk about for two straight minutes without flinching. That's the test. Apply it to every draft.
What it's actually for (the four jobs)
Used honestly, AI takes the four heaviest pieces of search-paperwork off your back:
- The resume first draft, from your real material. Don't ask it to "write you a resume." Tell it what you actually did: brain-dump your last few roles, messy and unpolished: projects, responsibilities, what you were proud of. Ask it to organize that into clean resume language. Then check every line against reality and cut anything that got inflated on the way through.
- Tailoring to each posting: the job nobody has energy for. Paste the job description and your real resume, and ask which of your actual experiences speak to their stated needs: what to lead with, what to trim. Tailoring means re-ordering and emphasizing what's true. It never means adding what isn't.
- The cover letter you'd never start. Give it the posting, your resume, and one honest sentence about why you want this one. Get a first draft, then do the pass that matters: cut what sounds like nobody ("I am passionate about leveraging..."), keep what sounds like you. Recruiters read hundreds of unedited AI letters a week; the edit is what keeps yours out of that pile.
- Interview practice, on demand, no favors owed. This one's underrated. Tell it the role and ask it to interview you one question at a time, with follow-ups, pushing back on vague answers. Do it out loud. Ten reps against a tireless practice partner before you've spent a single real interview finding out where your answers are thin.
Plus a fifth, quieter one: research the company before you walk in. Ask what the company does, who its competitors are, what questions a candidate should ask. Then verify anything important on the company's own site, because AI's knowledge can be stale or wrong, and "so, tell me about your Boston office" lands badly when they closed it last year.
Before and after, honestly
The part nobody tells you about "AI-proofing"
Two honest notes about the other side of the table:
Employers use AI too. Screening software, and increasingly, tools that flag inconsistencies. But the deeper truth is older than AI: the interview itself is the fraud detector. A fabricated resume may survive being read. It fails when it is questioned. Which is why the fast path to getting caught isn't "using AI." It's using AI to say things that aren't true.
And unedited AI writing has a smell. The phrases everyone's chatbot reaches for, like "proven track record," "dynamic professional," and "passionate about driving results," now read as template, because they are. The fix isn't hiding the tool; it's the edit. After AI drafts, read it aloud and replace anything you would never say with what you would. Your voice on top of a clean draft beats both the blank page and the robot letter.
What this actually feels like (a week of it)
Here's the practical shape, so it's not abstract:
Monday: one messy brain-dump session. Write everything you did in your last roles, typed fast, no polish. That document becomes your source of truth. Use that document as the source for every AI draft so the model has facts to work from.
Then per application: paste the posting + your source doc, get the tailored resume emphasis and a cover-letter draft, do your honesty pass (every line survivable for two minutes out loud?), do your voice pass (does it sound like me?), send. Reusing that source doc can make tailoring more manageable with better-fitted materials, because the tailoring actually happened instead of being skipped from exhaustion.
Before an interview: two practice rounds out loud, one company-research pass, verified. Walk in prepared instead of hopeful.
Your move
Tonight, do the Monday piece: open a blank doc and brain-dump your real experience. Messy is fine, messy is the point. That one honest document is the foundation for every application after it.
Then pick ONE live posting and run the full loop: tailor, draft, honesty pass, voice pass, send. You'll know by the end whether this method carries you, and nothing you sent will be anything you can't stand behind in the room.
Frequently asked questions
Should I use AI to write my resume?
Use it to organize and sharpen your real experience: brain-dump what you actually did, let AI structure it, then verify every line. Don't ask it to "make you impressive": it will invent, and every invented line is an interview question you can't answer. The rule: if you can't speak to a claim for two minutes, it doesn't go on the page.
Will employers know I used AI?
Unedited AI writing is often recognizable. The same stock phrases show up in hundreds of applications. Edited AI writing, grounded in your real experience and finished in your own voice, reads as what it is: clear writing. The tool isn't the tell; skipping the edit is.
Can AI help me prepare for interviews?
Yes. It's one of the best honest uses. Tell it the role, have it interview you one question at a time with follow-ups, and answer out loud. You find the thin spots in your answers during practice instead of during the real thing, and you can run it as many times as you want.
Is it cheating to use AI for a cover letter?
No. A cover letter communicates your experience, and drafting help is normal. It is not an exam. What crosses the line is content: claims about experience you don't have. Keep every claim true and finish the letter in your own voice, and you've used the tool the same way you'd use a sharp-eyed friend.
This is one page of a bigger idea: use AI in the open, check its work, own the result. How that goes at company scale, wins and failures included, is in the AI Guides: accounts of real work, with the parts that went wrong left in.