Career Decision Framework

Learning Objectives

• Apply a structured framework to evaluate how AI affects your specific role
• Use the T-shaped professional concept to identify where to focus skill development
• Assess your field's AI adoption trajectory and what it means for your career strategy

Why Generic Advice Falls Short

Most advice about careers and AI is either too optimistic ("AI will augment everything, don't worry") or too pessimistic ("AI will take your job, retrain immediately"). Both framings miss what actually matters: the specific intersection of your role, your industry, your experience level, and your adaptability.

This section gives you a structured way to think about your own situation — not a universal answer, but a set of questions that produce a clearer picture than generic advice can.

The Four Questions

Start with four diagnostic questions about your current role or target role:

Question 1: How automatable is the core task?

The core task is the thing you spend most of your professional time doing. Not peripheral tasks — the central activity that defines the role.

• Is the core task text-based? (Writing, summarizing, translating, reviewing, classifying)
• Is it highly repetitive with consistent inputs and outputs?
• Is there a clear definition of "correct" that AI can optimize for?
• Could you describe what "good output" looks like to someone who has never done the job?

If you answered yes to most of these: the core task has significant automation potential.

If the core task involves: novel judgment in ambiguous situations, trusted human relationships, physical manipulation in variable environments, ethical accountability — it has lower automation potential.

Question 2: Does the role leverage AI or compete with AI?

This is the key distinction that people miss.

• Leverages AI: The role uses AI tools to amplify output. The human directs, reviews, and makes judgment calls. AI does the volume work. Examples: software engineer using Cursor, lawyer using Harvey AI, analyst using AI for data summarization. A Harvard/BCG study found that consultants using GPT-4 completed tasks 25% faster with 40% higher quality — a clear augmentation win.
• Competes with AI: The role's primary output can be replicated directly by AI without the human in the loop. Examples: writing generic SEO articles, basic code generation for well-defined specifications, transcribing audio to text.

Most roles can move from "competing" to "leveraging" with skill development. The question is whether you're making that move proactively.

Question 3: What is my human value-add that AI cannot replicate?

This question forces specificity. It is not enough to say "judgment" or "creativity" — every job involves some degree of both. The useful question is: what specific aspect of your judgment or creativity, applied in what specific context, is irreplaceable?

For a corporate lawyer, it might be: understanding this particular client's risk appetite, their regulatory environment, and the specific judge in the jurisdiction where they operate — and synthesizing these into advice that no AI could give without that context.

For a product manager, it might be: the ability to walk into a room with six cross-functional stakeholders with conflicting interests and build consensus around a decision — a mix of social reading, negotiation, and organizational knowledge that is deeply human.

The clearer and more specific your answer, the more robust your position.

Question 4: Is my industry AI-adopting or AI-resistant?

Not all industries adopt AI at the same rate. Some structural factors determine adoption speed:

Being in a high-adoption industry means AI fluency amplifies your effectiveness more, and AI avoidance costs you more. Being in a slow-adoption industry gives you more time — but not indefinite time.

The 2×2 Automation Risk Matrix

Plot your role on two dimensions:

                    HIGH AUTOMATION POTENTIAL
                            |
                     QUADRANT 2     |    QUADRANT 1
                   (Augment or pivot)|  (Most urgent action)
                                    |
LOW AI ADOPTION -------------------+------------------- HIGH AI ADOPTION
                                    |
                     QUADRANT 3     |    QUADRANT 4
                   (Relative safety)|  (Best position)
                            |
                    LOW AUTOMATION POTENTIAL

Quadrant 1 — High automation potential + High AI adoption (Most urgent action needed) Your role's core tasks are automatable, and your industry is actively deploying AI. This is the highest-risk position. The time to act is now — either build AI fluency to move the role toward "leveraging AI," or develop adjacent skills that move you toward a lower-risk quadrant.

Quadrant 2 — High automation potential + Low AI adoption (Time-limited safety) Your role is automatable but your industry is slow to adopt. You have more runway than Quadrant 1, but the threat is real. Use this time to develop skills and fluency before the adoption wave hits your sector.

Quadrant 3 — Low automation potential + Low AI adoption (Relative safety) Your role involves tasks that are difficult to automate, and your industry is not aggressively adopting AI. You are relatively insulated — but "relatively" is the key word. No role is fully immune, and active engagement with AI tools remains valuable.

Quadrant 4 — Low automation potential + High AI adoption (Best position) Your role requires human judgment and your industry is adopting AI rapidly — meaning AI amplifies your effectiveness without threatening your position. Doctors using AI diagnostics, architects using AI-assisted design, engineers using AI coding tools all benefit from being in this quadrant.

The T-Shaped Professional

The most resilient career architecture for the AI era is T-shaped:

Broad AI literacy across tools, concepts, and applications
←────────────────────────────────────────────────────────────→
                        │
                        │ Deep domain
                        │ expertise in
                        │ your field
                        │
                        ↓

The horizontal bar: Broad AI fluency — understanding what AI can and cannot do, knowing which tools exist and how to use them, keeping current with developments in the field. This does not require engineering skills. It requires genuine engagement with AI tools regularly.

The vertical bar: Deep specialization in a specific domain where you have experience, context, and credibility. This is your moat. AI + deep domain expertise is far more valuable than AI alone.

The T-shaped professional is more resilient than the specialist who ignores AI (shrinking vertical bar with no horizontal bar) or the AI generalist with no deep expertise (horizontal bar with no vertical bar).

Some thought leaders are extending this model further: the "Pi-shaped" professional develops two or more deep verticals rather than just one, with AI literacy as a required horizontal bar across all of them. McKinsey has pushed the concept of "skills-based organizations" where dynamic portfolios of skills matter more than any fixed shape. But the core principle remains: broad AI fluency combined with deep domain expertise is the most resilient architecture.

Applying the Framework to Real Roles

Example: Marketing Manager at a Mid-Size Company

1. Core task automatable? Content generation and campaign reporting are high-automation. Strategic planning, customer empathy, and brand voice judgment are lower-automation.
2. Leverage or compete? Currently competing in content generation. Could shift to leveraging: let AI generate draft copy and spend human time on strategy, voice, and judgment.
3. Human value-add? Deep knowledge of this specific customer base, brand history, and competitive position. Judgment about what resonates authentically with the audience.
4. Industry adoption? Marketing has high AI adoption — McKinsey reports it as one of the top three AI-adopting functions, and AI content tools are table stakes.

Diagnosis: Quadrant 1 risk for the content-production portion of the role. Action: immediately build fluency with AI content tools and reposition toward strategy and voice judgment.

Example: Pediatric Nurse Practitioner

1. Core task automatable? Clinical assessment involves pattern recognition (AI-assistable) but primarily requires physical examination, patient relationship, judgment in ambiguous presentations, and the ability to interpret what a child or parent is not saying.
2. Leverage or compete? Can leverage AI for documentation, reference lookups, and administrative tasks — freeing more time for the human work.
3. Human value-add? The trusting relationship with the child and family, physical presence, the clinical judgment that integrates verbal and non-verbal signals.
4. Industry adoption? Healthcare is adopting AI for documentation, imaging analysis, and triage, but the clinical relationship remains deeply human.

Diagnosis: Quadrant 4. AI adoption in healthcare amplifies NP effectiveness without threatening the core role. Focus: adopt AI documentation tools and stay current with AI diagnostics.

Key Takeaways

• The four diagnostic questions — automatable core task, leveraging vs. competing with AI, specific human value-add, industry adoption — reveal your actual risk profile better than industry generalizations
• The 2×2 matrix locates your role: Quadrant 1 (high automation + high adoption) requires the most urgent response; Quadrant 4 (low automation + high adoption) is the most advantageous position
• The T-shaped professional — broad AI fluency plus deep domain expertise — is the most resilient career architecture for the AI era
• Moving from "competing with AI" to "leveraging AI" is a choice available to most roles, but it requires deliberate action, not passive adjustment