5 AI Prompts for Career Design You Can Use This Week

Most career anxiety comes from not knowing what you are actually building. You have a current role, some skills you use, some skills you ignore, a side interest or two — but no map of how those pieces relate or where they are headed. The problem is not a lack of ambition; it is a lack of structure for thinking about career as a portfolio that can be actively managed rather than a single track you are either on or off. AI is well-suited to career design conversations because it does not have a stake in what you decide, will surface assumptions you have not examined, and can run a pre-mortem on a move before you make it. The five prompts below are structured for that kind of thinking — not motivational, not generic, but analytical. They work in sequence or independently, depending on where you are.

Prompt 1: The Career Thread Map

Use when: You want a structured map of what you are actually building professionally — not what your job title suggests, but the real threads of skill and audience you are maintaining, whether or not you have named them.

I want to map my current career as a portfolio of professional threads. Here is everything I do professionally:

Current role: [title, company type, industry]
Main tasks I do weekly: [specific list]
Skills I use regularly: [list]
Skills I have but rarely use at work: [list]
Work I do outside my main role: [consulting, advising, writing, teaching, side projects, etc.]
Any past roles or work that still feels relevant: [brief description]

Based on this, identify 3–5 distinct career threads. For each thread, name it, describe the core skill and audience, and label it as Primary (most energy), Secondary (real competence, less investment), or Exploratory (low-commitment, high-optionality).

Then identify the one thread that appears most underdeveloped relative to its underlying potential.

What good output looks like: Three to five named threads with distinct labels (not just “tech work” and “writing”), each classified as Primary, Secondary, or Exploratory, and one thread flagged as underdeveloped with a specific reason. A failed response groups too many things under a single thread because the input was not specific enough. If the map looks like one thread with sub-items, add more detail to the skills you have but rarely use and the work outside your main role.

Prompt 2: The Automation Exposure Audit

Use when: You want an honest, task-by-task assessment of where AI tools are likely to compress or eliminate the routine portions of your work — and what that implies for where to invest your skill development time.

I work as a [role] in [industry]. My main weekly tasks are: [list].

For each task, assess: (1) How automatable is this with current or near-term AI tools — low, medium, or high? (2) What is the skill that remains valuable even when the routine portions are automated? (3) Should I be developing more of that remaining skill, or finding tasks that are less automation-exposed?

Be direct. I'm not looking for reassurance — I'm making investment decisions.

What good output looks like: For each task you listed, a clear automation risk level (low/medium/high) with reasoning tied to what current AI tools can actually do, a named residual skill that remains valuable even after automation, and a development recommendation. A failed response rates everything “medium” with generic hedging, or provides reassurance that your work is safe without engaging with the specific tasks. If you get that, push back: “What specifically about task X makes you rate it medium rather than high?”

Prompt 3: The Thread Strength Assessment

Use when: You are considering investing significantly more time and energy in a specific career thread and want a realistic assessment of its market strength, skill requirements, and risk profile before committing.

I'm evaluating the strength of a career thread I'm considering investing in more heavily. The thread is: [description — skill domain + audience + delivery format].

Please assess: (1) How rare and hard to replicate is deep competence in this domain? What would it take someone starting from scratch to reach a competitive level? (2) What concrete market evidence exists that this work is valued — job postings, consulting rates, audience size benchmarks, etc.? (3) What does a credible practitioner at each level (junior, competent, expert) actually produce or deliver? (4) What is this thread's biggest risk over the next 3–5 years?

Flag areas where you have low confidence in the market data.

What good output looks like: Distinct assessments at each of the four dimensions — rarity, market signal, practitioner output at each level, and risk — with an explicit flag on any area where the model lacks reliable data. A response that skips the low-confidence flag should be treated skeptically: market data for niche domains is genuinely sparse, and a model that expresses equal confidence about all four dimensions is probably filling gaps with plausible-sounding generalities.

Prompt 4: The Career Pre-Mortem

Use when: You are close to committing to a significant career move — a new role, a transition to a different thread, launching an independent practice — and want to surface the most likely failure modes before you are inside them.

I'm considering [describe the career move — new role, transition to a different thread, launching a side practice, etc.].

Run a pre-mortem: assume it is 24 months from now and this move has failed. I'm earning less, doing work I don't enjoy, or in a worse position than I started.

What are the three most likely reasons that happened? For each reason: (1) How probable is it given my specific situation? (2) Is there an action in the first 90 days that would prevent it or give early warning that it's occurring?

Don't be encouraging. Be accurate.

What good output looks like: Three named failure modes that are specific to your described move — not generic reasons career changes fail — each with a probability assessment tied to your situation and a specific 90-day action that would either prevent the failure or give you early warning it is happening. A response that identifies “mismatch with company culture” as a risk without explaining why that applies to your specific move has not done the work this prompt asks for.

Prompt 5: The 90-Day Career Experiment Design

Use when: You want to test a career direction without making an irreversible commitment — and you want the experiment to produce real evidence about whether to invest further, not just additional experience.

I want to explore [career direction or thread] through a 90-day experiment. I have approximately [X hours per week] available alongside my current obligations.

Design the experiment specifically: (1) What would I build, create, test, or do during these 90 days? (2) What is the concrete output or outcome at the end of the experiment that tells me whether to invest further? (3) What would a successful outcome look like? What would a not-promising outcome look like? (4) What is the smallest possible version of the first action I could complete in the next 7 days?

The experiment should generate real evidence, not just experience.

What good output looks like: A specific description of what you will build, write, test, or deliver in the 90 days — not “explore the space” — a named output or outcome at the end that functions as a decision criterion, and a first action small enough to complete in the next seven days. A failed response gives you a learning plan instead of an evidence-generating plan. The distinction matters: learning accumulates without a clear decision point; evidence forces one.

Common Pitfalls

Pitfall: Using Prompt 1 with a job title instead of actual tasks. Inputting “I’m a product manager at a tech company” will produce a generic career map. Fix: list your actual weekly tasks, the skills you use versus the ones you have but rarely deploy, and any work you do outside your primary role. The thread map is only as specific as the raw material you give it.

Pitfall: Running Prompt 4 after you’ve already made the decision mentally. The pre-mortem is most useful when you are genuinely open to being told no. If you run it after you have emotionally committed, you will unconsciously push back on every risk the AI surfaces. Fix: run it before you tell anyone else about the move.

Pitfall: Designing a 90-day experiment that is really just 90 days of learning. Courses, reading lists, and informational interviews are not experiments — they produce familiarity, not evidence. Fix: insist on a concrete output at the end of the 90 days that an external observer could evaluate. A piece of client work, a published analysis, a demonstrated consulting engagement. If there is no artifact, there is no evidence.

Pitfall: Accepting the automation risk assessment without checking it. AI models can be overconfident about automation exposure — both overstating and understating it. Fix: after getting the assessment for Prompt 2, ask: “What is the strongest counter-argument to your automation risk ratings?” The friction surfaces where the model is guessing.

Run Prompt 1 today. The thread map it produces is the foundation for everything else — and it takes less than 30 minutes to generate a first draft you can work from. The full framework behind these prompts — including how to rotate between threads over time and when to promote an Exploratory thread to Secondary — is in the Complete Guide to AI for Career Design.

Related: The Complete Guide to AI for Career Design · How to Design a Career with AI · Beyond Time Career Design Walkthrough

Tags: AI prompts for career design, career design prompts, AI career planning, career portfolio prompts, career design AI