The 10 most AI-resistant careers

Read this on victoriaprew.com​

Read time: 3 minutes

Hey! Happy Monday.

I want to tell you about a tweet I haven't been able to stop thinking about:

"Not enough people are emotionally prepared for if it's not a bubble."

Here's the thing: calling AI a bubble feels good. We saw the dot-com crash. Or you've heard the stories. Pets.com. Webvan. $5 trillion evaporating overnight.

But here's what actually happened after.

The internet didn't die. It ate everything. Google went public in 2004. Facebook launched. Amazon survived and became one of the most valuable companies in the world. Netflix killed Blockbuster. The people who dismissed the internet after 2001 missed the biggest wealth creation in human history.

So I want to ask you something that might be uncomfortable: what does your career look like if it's not a bubble?

[the system]

The asymmetry nobody talks about

Here's why I think investing your time into AI is a complete win-win. Think about it:

If AI is a bubble and you go all-in learning it:

• You lose some hours and maybe a few hundred pounds on subscriptions
• You learn to think more clearly, communicate with more precision
• Those skills don't vanish because a market crashes
• Worst case, you were faster at your job for a year

If AI is not a bubble and you waited:

• The tools evolved while you watched
• The people who started early compounded their advantage month after month (if you're not using it everyday, you're already behind)
• A gap compounds. You can't get that time back.

For me personally, AI has changed 60% of how I work in the last year. My small team now handles 10x what we could do alone three years ago. And I'm not special. I don't code. I'm learning in-real-time too. The only real difference between me and someone who closed the ChatGPT tab in 2024 is that I kept the tab open. And kept playing, testing and learning.

Stop asking "is AI a bubble?" Start asking: what am I actually capable of with AI?

[the framework]

The 10 most AI-resistant careers

If you're worried AI is coming for your job, or you're advising the next generation on their career, here's where the evidence actually points. I've linked every source so you can go deeper.

1. Therapists and Coaches: Building emotional trust and guiding personal transformation. Oxford University's 2024 reappraisal found that long-standing relationships built on in-person interaction remain firmly in the realms of humans. The more transactional a relationship becomes, the more it can be automated. Therapy is the opposite of transactional.

2. Healthcare Providers: Surgeons, Nurses, Occupational Therapists Delivering empathetic care and split-second judgment under pressure. The WEF Future of Jobs Report 2025 identifies healthcare and care economy roles as high-growth precisely because of ageing populations and the irreplaceable need for human presence. Healthcare consistently scores among the lowest AI applicability roles in Microsoft's research.

3. Teachers and Mentors: Fostering curiosity, ethics, and human potential in dynamic group settings. Oxford's research makes clear that the key bottleneck to automation is social tasks, and teaching is one of the most socially complex jobs in existence. AI can give a student information. It cannot make them want to learn.

4. Skilled Tradespeople: Plumbers, Electricians Hands-on problem-solving in unpredictable physical environments. This is perhaps the most surprising career bet on this list, but it's backed hard by data. Microsoft's 2025 occupational research analysed 200,000 real Copilot conversations and found that trades like plumbing and electrical work have among the lowest AI applicability of any occupation. Every job is in a different building, behind a different wall. Nvidia CEO Jensen Huang made the same point at Davos in January 2026: "We're going to have plumbers and electricians and construction workers... this is the largest infrastructure build-out in human history."

5. Personal Trainers and Fitness Coaches: Personal connection, real-time form correction, motivation, adaptation. AI can generate a workout plan in seconds. It cannot read your energy at 6am or push you when you're about to quit. The in-person element is the entire product.

6. Strategic Leaders and Executives Navigating ambiguity, inspiring teams, making high-stakes calls. The WEF Future of Jobs Report 2025 names leadership and social influence as one of the skills growing fastest in importance through 2030. Vision-setting, culture-building, conflict resolution. These are not tasks you can hand to a model.

7. Brand Strategists and Storytellers: Crafting original narratives, cultures, and movements that actually resonate. Oxford's research found that while LLMs are good at generating new combinations of existing ideas, they rarely make conceptual leaps. Deciding what a brand should stand for, and why people should care, is still a deeply human call. Storytelling is becoming one of the most valuable skills in business, not one of the least.

8. Cybersecurity Professionals: Anticipating threats, responding to real-time attacks, understanding the human behaviour behind the code. The US Bureau of Labor Statistics projects 33% job growth for information security analysts from 2023 to 2033, one of the fastest growth rates of any profession. You're not fighting machines. You're fighting humans using machines. That distinction matters enormously.

9. Complex Systems Designers: Urban Planners, Sustainability Experts Integrating technology, humans, and ecosystems holistically. The WEF Future of Jobs Report 2025 lists environmental and renewable energy engineers in the top 15 fastest-growing professions through 2030. These roles require weighing competing human values and designing for actual people, not running an algorithm.

10. Experiential Creators: Event Curators, Community Builders Designing irreplaceable live human connections. The Oxford team's conclusion is unambiguous: in-person interaction is the final moat. Moments and communities that only exist with human energy and intention in the room cannot be replicated by any model, however capable.

[the lesson]

The through-line across all 10?

Every single career is defined by what AI structurally cannot do: hold space, build trust, read a room, make a values-based judgement call, or be physically present. These aren't soft skills. They are the hardest skills.

If you want to stop theorising and actually experience what this means for your specific work, here's how to do it properly. Open Claude. Find the prompt below that matches your role. The key is to give it real context, not a hypothetical. Paste your actual work, your actual numbers, your actual team dynamics. That's when it gets genuinely useful.

If you lead a business or team: Paste your last board or management update, your current OKRs, and your three biggest open decisions. Then ask: "Act as a world-class operator who has scaled multiple businesses to this stage. Based on everything I've shared, identify the one constraint I'm treating as permanent that is actually a choice, and tell me what it would look like to eliminate it in the next 90 days."

If you write or build content: Paste three pieces of your best recent work. Then ask: "Study my voice, the rhythm, the specific language, the structure, the way I open and close. Now write a first draft on [your topic] that a reader would believe I wrote. Don't write AI content. Write mine."

If you manage a team: Paste your last round of performance conversations or reviews. Then ask: "What patterns am I too close to see clearly? Surface the one uncomfortable dynamic I'm probably avoiding, and give me the specific question to ask each person in my next one-to-one to open it up."

If you're working on a pitch or fundraise: Paste your current deck narrative and the last three pieces of feedback you've received. Then ask: "You are a sceptical investor who has seen 500 decks at this stage. Tear this apart. What's the single weakest assumption I'm asking you to accept, and how would you stress-test it?"

Fifteen minutes. Not tomorrow. Tonight.

The people emotionally prepared for a world where AI is not a bubble are already in it.

Until next week,

Victoria