You take it. The first 18 months are exactly what you needed.

Two-way doors look like one-way doors when you're scared. Most jobs are reversible — you can leave. The reversibility test (Bezos): 'if this turns out wrong in 18 months, can I walk it back?' If yes, the cost of a wrong yes is much lower than the cost of a wrong no. The career risk in 2026 isn't usually 'took the wrong job.' It's 'froze and took none.' You unfroze. First year: comp + skill + network as planned. Year 2: the role shifts. You're either ready for the shift or you're not — depending on whether you treated this as a destination or a stepping stone. What protects you: starting the next-move work at month 12, not month 24.

You take it — and you reshape it.

The decision isn't always 'this offer as written.' It's often 'this offer with the right modifications.' People underestimate how much can be negotiated *post-offer, pre-acceptance* — title, scope, growth path, review timeline, even the team you're attached to. The AI-displacement question isn't 'is this role safe?' It's 'can I shape this role toward the part that's harder to automate?' Most of the time the answer is yes if you ask early. Tonight: harder ask than you'd planned, lower regret either way. Months 1-6: doing work that AI augments rather than replaces. Long-term: you've practiced negotiating the shape of work — which is the meta-skill of a 2026 career.

You decline. You hold out.

Declining is the right move *only if* the no is in service of something — a clearer target, a real alternative, a runway you've actually counted. Declining as avoidance dressed up as principle is the most expensive version of this decision. The diagnostic: can you describe what you're holding out for in one sentence? If yes, the no is real. If you find yourself saying 'something better,' you're avoiding. Tonight: relief that turns into doubt by morning. Months 1-3: the test is whether you're actually moving toward the alternative. Long-term: the no is right if it was *for* something. Wrong if it was *from* fear.

You don't decide tonight. The 48 hours change what you can see.

Most career decisions made under same-day pressure are decisions about pressure, not about the role. Asking for 48 hours costs almost nothing reputationally — companies expect it for senior offers. The 48-hour delay isn't procrastination; it's getting your judgment back. The version of you that's tired and feed-poisoned at 11pm is not the version that should make a 2-year career commitment. Tonight: small itch of 'what if they pull the offer' — they won't. By Friday: the right answer is more obvious than it was Wednesday. As a habit: any career decision that has to be made same-day, isn't.

A sentence to sit with

Whatever you decided, name what it cost — and what it bought. Career decisions get made looking forward but only get judged looking back. The version of you in 18 months will appreciate having something to read.

Should I take this job that AI might eliminate in 3 years? - Decision Tree

Q: Should I take this job that AI might eliminate in 3 years?

The offer is good. The role is the kind of thing AI might absorb by 2029. You don't know how seriously to weight that. The honest answer is: nobody does. Here's how to decide anyway.

Loading Psychology Insights...

Preparing your mental wellness content

Should I take this job that AI might eliminate in 3 years? - Decision Tree | PsychStreet