It's the last Friday of May. Quick check-in.For the past few weeks I've been asking you to lock in, share wins, and treat this place like the room you'd want in your corner. Reflect on May for a second. - What worked in your search this month? - What's one thing that moved, even small? - What didn't work? - What's one thing you're going to stop doing in June? Mine: Even though I'm not searching right now, I'm always open to opportunities. I'm not a believer in shutting doors. This month I interviewed with a competitor. I was qualified, I met all the requirements, and the interview went well. But the vibes from the recruiter were off. I had that inkling feeling going into it as I reviewed the company's website, and I chose to ignore it because I thought my skill would outshine whatever was going on underneath. It didn't. They had an agenda, I was aware of it, and I walked in anyway. - What worked: Staying open and taking the meeting. You never know what a conversation will teach you. - What moved: My ability to read a room in real time, even when I want to be wrong about it. - What didn't work: Ignoring my gut because the role looked good on paper. - What I'm stopping in June: Overriding what I already know. What's for me is for me. Drop your May reflection in the comments. Win, lesson, or both. One sentence is enough.June starts Monday. Let's keep going.

Most people use Claude for a single chat, paste, response, close. There are three features that change what's possible if you're using it for a real job search. 1. Projects I covered the basics in week 1. Worth repeating: Project Knowledge holds your resume, LinkedIn, target role notes, and any drafts. Every chat in that Project starts with full context. That’s a game changer for a long search. 2. File uploads Drop in a JD, a company’s annual report, a recruiter’s LinkedIn profile, or your full work history doc directly into Claude. You can ask it to compare two JDs side by side, summarize a 40-page company report into the five things you should know before an interview, or pull patterns across 10 rejection emails. Most people are still pasting text. Just upload the file. 3. Voice mode (in the mobile app) If you’re prepping for an interview, talk to Claude like you’re talking to the interviewer. Out loud. It’s the cheapest mock interview you’ll ever do. You hear your filler words. You catch where you ramble. You realize you don’t actually have a clear answer to “tell me about yourself.” Way more useful than reading your prep notes silently. Pick one of these you haven’t used and try it this week. Which one are you trying first? A 4th would be Skills, let me know if you want a quick lesson.

The market is hard. AI doesn’t fix that. But here’s what it actually does. I want to be clear about something, because I see a lot of AI hype around job searching that I don’t agree with. AI is not going to land you a job. The market is hard. Hiring is slow. Roles are more competitive, not less. There are fewer postings in some industries. Recruiters are buried. Hiring managers are picky. None of that changes because you opened ChatGPT. What AI does do is compress the time between effort and output. The thing that used to take 4 hours — researching a company, customizing a resume, drafting a cover letter, prepping interview answers — can take 45 minutes if you’re using AI well. That doesn’t mean you do less work. It means you do the same work faster, which means you can do more of it, at higher quality, with more energy left for the parts that actually matter: - real conversations - real follow-ups - real preparation AI is a force multiplier, not a magic wand. It compounds the effort you’re already willing to put in. If you’re not willing to put in the effort, no tool fixes that. If you are, AI can absolutely change what’s possible in 30 days. Where are you on this? Are you over-relying on AI, under-using it, or somewhere in between? 👀

Three things I'd change if I started over today. 1, I'd talk to humans before I touched the application portal. I spent the first 4 days in spreadsheet mode, building lists, researching companies, organizing my pipeline. I should have spent those days messaging 10 people in roles I was curious about. The intel I got later from real conversations was 10x what I got from job descriptions. 2, I'd batch my tasks instead of jumping between modes all day. The first two weeks I was researching, applying, networking, and prepping in the same hour. By Friday I was exhausted and couldn't tell you what I'd actually moved forward. Two hour blocks for each type of work, one type at a time, would have doubled my output and cut the burnout in half. 3, I'd have given myself a daily stop time. The search will eat your entire day if you let it. I had days where I worked until 11pm, applied to 20 to 30 roles in a single day, and was still trying to prep for the next morning's interview. A clean stop time at 6pm would have made the whole month sustainable. What would you do differently in your search if you started over tomorrow?

Most people skip context when they prompt AI. Then they wonder why the output is generic. Here’s a 4-line primer you can paste at the top of any new chat (if you’re not using a Project or Custom GPT yet). Fill in your specifics, then ask whatever you need. I’m a [your level] [your function] with [years] years of experience in [industry]. I’m currently searching for [target role] roles at [company stage/size]. My non-negotiables are [3 things] and my deal breakers are [2 things]. Be direct, push back on weak ideas, and don’t pad your responses. That’s it. Four lines. Drop that in before you ask Claude, or your preferred GPT, to review your resume, draft a cover letter, prep you for an interview, or evaluate a JD. The output will get sharper immediately because the AI now knows who’s asking and what good looks like for you. The reason this works: AI defaults to safe, generic, middle-of-the-road advice when it doesn’t know who you are. Once it knows, it can actually be useful. This isn’t a guide prompt, it’s a context primer. Use it on top of any prompt you’re already running. Try it this week and tell me what shifted in the output.