One of the most practical things you can do about job uncertainty is to make your income depend on one employer a little less. Not by quitting. By running a small, cheap test.

Here's a 60-day version you can start planning tonight.

Step 1: Pick the skill, not the business idea.

Write down what colleagues already ask you for help with. Spreadsheet cleanup, editing, explaining a tool, reviewing a draft, setting up a process. Something people have already valued.

Step 2: Define one paid offer.

One sentence: "I do X for Y type of person, finished by Z." Keep it narrow and small enough to deliver in a few hours. Narrow beats clever.

Step 3: Set a cost ceiling before you start.

Decide the most you're willing to spend, in money and weekly hours, and write it down. A hedge that drains your savings or your sleep defeats its own purpose. If you're already short on runway, spend nothing but time.

Step 4: Find three people to ask, not a thousand to advertise to.

Message three people you already know who plausibly have the problem. Ask if they'd want it, or know someone who does. Skip mass posting.

Step 5: Check your employment terms.

Read your contract for outside-work and conflict-of-interest clauses, and keep the work off company time and equipment. If it's unclear, ask HR or check before you take money.

Step 6: Review at day 60, with three numbers.

Hours spent, money earned, and whether you'd want to do it again. That's it. The result may be "no, not for me," and that's a useful answer for the price of two months.

What this does for your head

A lot of career dread comes from one thing: all your income sitting behind one decision you don't control. Even a tiny second source, or the evidence that you could find one, shifts the picture from "I'm stuck" toward "I have options I've actually tested."

It also gives you a real skills log: what clients asked for, what you delivered, what you learned.

Keep it small, keep it boring, and protect your sleep. If worry is keeping you up, a calmer plan beats a bigger one. If it's heavier than that, US 988, UK Samaritans 116 123, elsewhere findahelpline.com.

More on building a buffer: https://straata.store/articles/financial-runway-job-loss-anxiety

Free 6-page PDF, The 3am Protocol: https://straata.store/The-3am-Protocol.pdf

Published by Nightshift Guides.

If you're job hunting right now, there's a question most candidates never ask: what is this company actually planning to do with AI, and where does my role sit in that plan?

You don't need inside information. You need a short list of questions and the discipline to listen to how they're answered. Here's one you can prepare tonight.

Six questions for the hiring manager or a future teammate

1. "Which tasks on this team use AI tools today, and which are off limits?" Specific answers (drafting, review, testing, summarising) are a good sign. Vague answers mean nobody has decided yet.

2. "Who decides what gets automated here, and who reviews the output?" You want to hear that people with judgement are part of the loop, not just a cost line.

3. "How did the team's work change in the last year because of these tools?" Listen for whether the answer is about quality and speed or about headcount.

4. "If a task in this role gets automated, what happens to the person who did it?" Retraining, moving to other work, or "we'd revisit the role" tell you very different things.

5. "What would success in this role look like in 18 months, given how fast the tools are changing?" A good manager has thought about this. A flustered one hasn't.

6. "Is this role new, or replacing someone? Why did the last person leave?" Not AI-specific, but it often surfaces the real story.

How to use the answers

Write down the answers within an hour of the conversation, in plain words, before your mood rewrites them. Then score each question simply: clear, partly clear, or evasive. Two or more evasive answers aren't a reason to walk away on their own, but they are a reason to ask a second person on the team before you sign.

The point isn't to find a company with no AI plans. That company may be the riskier one. The point is to choose a role where you can see how you'd stay useful as the work changes.

If you're deciding while stressed or short on sleep, give yourself one night before you answer an offer. Decisions made at 3am usually look different by daylight. If things feel heavier than a career decision, talk to someone: US 988, UK Samaritans 116 123, elsewhere findahelpline.com.

More on choosing your next move: https://straata.store/articles/career-adaptability-ai-disruption

Free 6-page PDF, The 3am Protocol: https://straata.store/The-3am-Protocol.pdf

Published by Nightshift Guides.

Once you start thinking about learning something new, the options multiply fast: bootcamps, certificates, degrees, online courses, a side project. Each one arrives with confident marketing. Before you spend money or months, it helps to put every option through the same three questions on one sheet of paper.

Draw three columns: Cost, Time, Proof.

Cost. Not just the fee. Add what you would give up: unpaid hours, income if you reduce your work, the price of tools and materials. Write the full figure for your own situation. If you cannot tell what something really costs, that is information too.

Time. Two numbers: how long until you finish, and how long until it could plausibly change what you are paid or hired for. Those are rarely the same. Ask people who completed the course how long the second part took them, and treat the provider's own claims with some caution.

Proof. This is the column most people skip. What would you have at the end that a stranger could look at? A portfolio piece, a working project, a credential an employer in your field recognises by name. If the honest answer is "a certificate nobody in my industry asks about", that tells you something.

Then add a fourth line for each option: "Smallest test." What could you do in one weekend, for free or close to it, to find out whether you like this work and have some aptitude for it? Many options can be tested before they are bought: follow a free tutorial, do a small real task, talk to someone who does the job.

Finally, check it against your tasks. If you did the weekly task audit, look at which of your tasks the new skill would extend. Skills that build on work you already do are usually a shorter road than a clean restart.

You are not trying to pick the perfect option. You are trying to make the decision with a clear head and your own numbers instead of someone else's urgency.

More on building flexibility into your career: https://straata.store/articles/career-adaptability-ai-disruption

Free 6-page PDF: https://straata.store/The-3am-Protocol.pdf

Published by Nightshift Guides.

When a company starts talking about AI tools, a lot of people go quiet about pay. The thinking is, "now is not the time to ask for anything." That is understandable, but it often leaves the person with the fewest facts on the table in the conversation that matters.

Here is a practical way to prepare, using one page and about thirty minutes. You do not need market statistics for it. You need your own record.

1. Write down what the role is now. List the three to five things you are actually responsible for today, not what your job description says. If your work has shifted (you review and edit more than you draft, you coordinate tools, you handle exceptions), say so plainly. A role that has grown in scope is a different conversation from a role that has stayed the same.

2. Write down outcomes, not activities. For each responsibility, one line: what happened because you did it. Time saved, an error caught, a client kept, a deadline met. If you keep a running record of this, today is a good day to start. If you have no record, reconstruct the last three months from your sent mail and calendar.

3. Name what you do that a tool cannot. Not in the abstract. Give a specific example of a judgement call, a relationship, or a context-heavy decision from the past quarter.

4. Decide your ask before the meeting. Pick a specific number or change (pay, title, scope, training budget) and a fallback. Vague asks get vague answers.

5. Ask for the conversation in neutral terms. Something like: "I'd like twenty minutes to talk about how my role has developed and what compensation looks like going forward." That is a normal request.

You may not get a yes. But you will leave with clearer information about where you stand, which is worth having whichever way the answer goes. If the answer is no, the page doubles as the start of a case file for your next move.

For the money side of this, the runway article is a useful companion: https://straata.store/articles/financial-runway-job-loss-anxiety

Free 6-page PDF: https://straata.store/The-3am-Protocol.pdf

Published by Nightshift Guides.

"Can I pick your brain about my career?" is one of the hardest messages to say yes to. It's open-ended, it sounds like an hour of someone's time, and the person can't tell what you need.

If you're unsure where your field is heading, someone a few years ahead of you can save you a lot of guesswork. The trick is to make the ask small and specific. Here's a template you can use tonight.

The structure: context, question, size

1. Context (one sentence): "I work in [field] and I'm trying to decide which skills to build as AI tools get used more in our work."

2. One specific question: "You moved from [X] to [Y]. What did you wish you'd learned first?" or "Which parts of your week changed most after your team adopted [tool]?"

3. Size of the ask: "Would 15 minutes by video or a reply by message work, whichever is easier for you?"

Why it works

• A narrow question can be answered from experience, without preparation.

• Offering a message reply lowers the cost of saying yes.

• Mentioning why them (a move they made, a project they led) shows you aren't mass-mailing.

Before the call

Write down what you'll do differently depending on their answer. That's the test of whether the question is useful. If no answer would change anything, sharpen the question.

After the call

Send a two-line thank-you naming one thing you'll act on. Weeks later, tell them what happened. That follow-up is what turns one conversation into a relationship.

Who to ask

Think former colleagues, people from your professional association, someone you've seen write or speak about your field. Start with people who already know your name; a warm contact is far likelier to reply than a stranger.

You don't need to have a plan to ask for help. You need one question.

If worry about work is affecting your sleep or mood, it's worth speaking to someone: US 988, UK Samaritans 116 123, elsewhere findahelpline.com.

Further reading: https://straata.store/articles/career-adaptability-ai-disruption

Free 6-page PDF: https://straata.store/The-3am-Protocol.pdf

Published by Nightshift Guides.

The email lands: "realigning around AI", "streamlining", "a new operating model". Your stomach drops, and within an hour you've written three versions of the story in your head.

Here's a practical way to slow that down. Before you decide anything, spend twenty minutes with a notepad and answer five questions using only what the announcement actually says.

1. What was actually announced, in one sentence?

Strip the adjectives. "The company is reorganising two departments and buying a new tool" is a fact. "They're getting rid of people like me" is a prediction. Write the fact only.

2. Which tasks does it touch, and which does it not?

Announcements talk about teams and titles. Your work is made of tasks. List the ten things you did last week and mark which ones the new tool or structure plausibly affects. Often it's fewer than the first-hour feeling suggests. Sometimes it's more, and then you know where to focus.

3. What's the stated timeline, and what's missing?

Note any dates given. Note what isn't said: who decides, when you'll hear. Missing information is a list of questions to ask your manager, not a gap to fill with worst cases.

4. What would I want to know if I were my own friend?

Write two or three questions you could put to your manager in your next 1:1: what will my role own after this, what skills will matter, what does success look like in six months.

5. What's one thing I do tonight that helps in every scenario?

Update your CV, work out your money runway in months, or add this week's wins to your evidence file. Pick one. Do it. Stop.

Then close the notepad. The point isn't to be calm about the news; it's to separate what is known from what is feared, so your next decision is made on the first list.

If this is keeping you awake or you feel hopeless, please talk to someone: US 988, UK Samaritans 116 123, elsewhere findahelpline.com.

Further reading: https://straata.store/articles/should-i-tell-my-boss-im-anxious-about-ai

Free 6-page PDF: https://straata.store/The-3am-Protocol.pdf

Published by Nightshift Guides.