Do professionals near the end of their careers also need to change?

Yes, at a compatible scale. Not to become an AI developer — but to understand conceptually, converse fluently, decide well. A 60-year-old board member who can't differentiate an agent from a chatbot is in Structural Illiteracy. In 2026, that burns credibility.

Which careers are most at risk?

High-volume analysis/production (junior financial analyst, technical writer, repetitive-task programmer, template designer), administrative processing, Tier 1 customer service, transcription. In 5 years, a large part of these roles will be fully under agents.

What about human careers like therapy and teaching?

More protected in the short term (5-10 years), but also impacted. Therapists use AI for analysis and research. Teachers use AI to personalize content. Human careers that do NOT incorporate AI also fall behind within their own field.

I'm 45 years old, is it too late to change?

No. But it's urgent. Minimum plan: 4 hours/week of deliberate AI study for 12 months. In 1 year, you'll be far more prepared than 95% of professionals in your age range. That's the leap that becomes an opportunity.

Career in the AI Era: the important thing is not to execute tasks, it's...

Q: What advice would you give a young person starting their career in 2026?

Paulo Castello's direct answer: 'dive headfirst into learning Artificial Intelligence. For at least the next 10 years, our role as humans will be to direct AI.' Careers built on executing tasks will suffer. Careers built on directing AI will grow.

Career in the AI Era: the important thing is not to execute tasks, it's to direct with intelligence

For the first time in decades, the classic advice "specialize in doing X very well" has become a trap. In 2026, professionals who keep specializing in executing tasks will be replaced. Those who learn to direct AI become the protagonists of the next decade.

"I've been saying that it's not easy to guide new professionals. I had difficulty advising a nephew who works in financial markets. If he were at Fhinck, in a year his role wouldn't exist anymore."

— Paulo Castello, February 2026

The Shift That Makes Traditional Career Advice Obsolete

For decades, the classic career advice was: "pick a field, specialize deeply in executing it well."

Want to be an accountant? Learn to analyze financial statements with excellence.
Want to be a designer? Learn to create illustrations with excellence.
Want to be a software engineer? Learn to code with excellence.
Want to be a financial analyst? Learn to model spreadsheets with excellence.

That advice worked because, at the time, executing well was a differentiator. In 2026, executing well has become a commodity. AI executes very well, with more consistency, without tiring, without vacations, without turnover.

The differentiator has shifted. Those still chasing "being good at execution" are chasing the commoditization of their own careers.

The Question That Defines Careers in 2026

"In 5 years, will well-configured AI agents be able to do what you do today, with equal or superior quality?"

Honestly.

If the answer is "yes, largely" — your career will face a forced transition in the next 36 months. Those who act now, redirect. Those who deny it, get redirected.

If the answer is "no, because what I do requires X that AI can't do" — find out exactly what that X is, and build your career around it.

What AI Still Doesn't Do Well (and Where to Build a Career)

In 2026, the 5 areas where humans have a clear advantage over agents:

1. Strategic Direction Under Radical Ambiguity

When the problem is not yet defined and needs to be formulated. AI executes well within clear scope — but someone needs to set the scope.

2. Deep Human Relationships

Complex enterprise sales, therapy, political negotiation, team leadership. No agent replaces presence, social reading, built trust.

3. Genuine Conceptual Creativity

Not "generating 50 logo variations" (AI does that). But "deciding which problem is worth solving" (humans do that). Long-term vision, conceptual disruption, non-trivial intuition.

4. Ethical Judgment in Edge Cases

When a clear rule doesn't apply and a moral dilemma appears, agents stall or fail. Humans with solid ethical training decide. This layer of human judgment will become increasingly valued.

5. Agent Orchestration

The role being born: specialist in making agents work together to solve compound problems. It combines prompt engineering, business process understanding, and flow modeling. By 2030, it will be one of the highest-paid roles in the market.

Practical Advice by Career Stage

Young Starters (20-30 years old)

Forget specializations in executing tasks (analyst of X, programmer of something specific, writer of something).
Study AI technically (not just at surface level — understand LLMs, agents, MCP, complex prompts)
Build transversal skills: communication, negotiation, systems thinking.
Maintain optionality — try several areas before locking in a career for decades (decade-long career in one specialization is a 20th-century model).

Mid-Career (30-45 years old)

Identify the component of your role that will disappear in the next 5 years. Plan: migrate to the component that will grow.
Invest in becoming an orchestrator in your field (the person who directs AI + a hybrid human-agent team to deliver results).
Build public authority around your intersection (you + AI + your specialty).
Accept that you will change — those who cling to their current job description lose.

Senior (45-60 years old)

Position your experience as a unique asset — AI has technical capability but not 20 years of market scar tissue.
Learn conceptually about AI to converse fluently in councils, boards, executive committees.
Become a board advisor — the "company advisor" role will grow dramatically in the next 10 years, especially for those who combine deep sector experience + AI understanding.
Mentor younger professionals — your experience + their intuition = a rare combination.

C-Level (any age)

Study AI personally (don't delegate). As Satya Nadella, Pichai, and Altman do.
Demonstrate publicly what you do (LinkedIn, events) — signals capability, attracts talent.
Build a culture of continuous learning in the company (Sharpening the Axe)
Don't fear role elimination — if your role disappears, show that you built the next one.

Sam Altman's Take on 2026 (and What to Do With It)

"Sam Altman stated that superintelligence is near and that not even CEOs will be spared."

This excerpt went viral in January 2026. Common reaction: panic or denial.

The productive response is: assume it's coming, and build a career that survives and thrives in the new order.

Career in executing tasks won't survive. Career in directing AI survives — and thrives. Career in deep human relationships survives. Career in judgment under radical ambiguity survives.

You can redirect yours. A 12-month plan is enough — as long as you start today, not tomorrow.

Conclusion

Paulo Castello's advice to his nephew applies to any professional in 2026:

"Dive headfirst into learning Artificial Intelligence. For at least the next 10 years, our role as humans will be to direct AI."

Those who absorb this become protagonists. Those who resist become statistics.

Fhinck trains C-levels and board members on exactly this kind of transition. Schedule a conversation if you want to think through next steps.