In the race to predict which jobs are “safe” from automation, there’s been a persistent myth: that blue-collar, hands-on jobs are somehow more protected than knowledge-based roles. But the truth is far more radical—and it’s hiding in plain sight.
At the heart of this shift is a simple but profound truth: jobs most aligned with codified systems of knowledge are the most vulnerable to automation. In contrast, roles requiring creativity, intuition, and judgment—those rooted in ambiguity and indeterminacy—are likely to thrive.
In short, the automatons will be automated, and the future belongs to thinkers who navigate complexity.
The Quiet Revolution of Automation
Automation isn’t arriving in the form of humanoid robots wielding wrenches. It’s already here, woven into integrated systems that replace labor-intensive roles with scalable, efficient technologies. Consider:
- Appliance Repair: Samsung’s AI diagnostics now resolve 85% of issues remotely, replacing repair technicians with modular replacement systems.
- Construction: Boxabl’s automated factories assemble 90% of a home indoors, slashing timelines from months to days.
- Healthcare: The IDx-DR system detects diabetic retinopathy with 87% accuracy—outperforming human specialists [FDA approval notice].
This isn’t just automation—it’s cognitive outsourcing. As Erik Brynjolfsson notes in The Second Machine Age, “The key issue is not whether machines can think but whether humans can adapt.”
AI vs. Automation: Why the Distinction Matters
Traditional automation (e.g., factory robots) follows pre-programmed rules. Modern AI, however, mimics human cognition:
- Multiple AI based tools now exist to perform tasks previously performed by lawyers.
- Midjourney generates ad concepts for designers at architecture firms.
Yet AI falters where ambiguity reigns. For example, ChatGPT might draft a corporate memo, but it can’t navigate a CEO’s ethical dilemma about layoffs. As Yuval Noah Harari argues in 21 Lessons for the 21st Century, “The real question is: What will humans do once AI handles routine cognitive tasks?”
The Jobs AI Can’t Take Over
What’s left for humans? Ironically, the jobs that survive—and thrive—are those requiring creativity, intuition, and decision-making under uncertainty. These are roles where ambiguity and human connection remain central.
- Entrepreneurs and CEOs: While AI can handle data-driven analysis, strategic decision-making and stakeholder management require human judgment. A 2023 survey by edX found that nearly half (49 percent) of CEOs believe most or all of their role could be automated or replaced by AI.
- Creative Roles: Musicians, artists, and writers still hold cultural value because people crave authentic human experiences. AI may compose hits, but it doesn’t replace the emotional connection of a live concert. Adobe’s Firefly AI aids designers but can’t replicate the cultural resonance of a human-curated campaign.
- Adventure and Human Connection: Adventure tourism, where clients seek human expertise and connection, is thriving. Even here, AI acts as a support tool rather than replacing humans entirely. Companies like Intrepid Travel report a surge in demand for human-led expeditions post-pandemic.
The future of work doesn’t belong to those who follow instructions—it belongs to those who think beyond them.
The Technicality-to-Indeterminacy Ratio: Why a 1970s Theory Explains Today’s AI Revolution
Continuing our theme of the importance of the sociology of the professions to help explain the impact of AI on the workforce, this article draws on the theory of the technicality to indeterminacy ratio, first proposed in 1970 by sociologists Jamous and Pelloise.
Their theory proposed that all professions balance two forces: technicality (codified, rule-based knowledge) and indeterminacy (judgment, creativity, and adaptability). For decades, society prioritised technicality—schools trained students to memorise formulae, workers followed rigid protocols, and success meant mastering standardised systems.
In healthcare, the move towards evidence based practice, standardised protocols and procedures, and competency based education consolidated the codification of practice even further (for example, see Traynor’s article on how medicine and nursing responded to the evidence based medicine movement). But AI has inverted this hierarchy. Machines now excel at technical tasks: diagnosing diseases, drafting legal contracts, and assembling homes with robotic precision.
The World Economic Forum warned that 85 million technicality-driven jobs (e.g., data entry, quality assurance) will vanish by 2025, while 97 million new roles emphasizing indeterminacy—critical thinking, ethical decision-making, and creativity—will emerge. For example, AI handles radiology scans, freeing doctors to focus on patient empathy and complex diagnoses. Educators no longer grade rote quizzes but mentor students in navigating ambiguity, like solving open-ended climate challenges.
This isn’t just a workforce shift—it’s a reckoning with what makes us human. AI lacks the intentionality to weigh ethical dilemmas or channel cultural nuance into art. The future belongs to those who embrace indeterminacy: leveraging AI as a tool for technical tasks while cultivating the irreplaceably human skills of judgment, curiosity, and creative problem-solving. Jamous and Pelloise’s framework isn’t just academic—it’s a survival guide for the age of AI.
Challenging traditional divisions of labour
The automation of professional work doesn’t always unfold the way we expect it. For example, in radiology and medical imaging, the traditional division of labour between radiologists, radiographers, and sonographers saw radiologists shedding the more technical, hands-on aspects of the job to focus on higher-order diagnostic reasoning. Ironically, it’s now this intellectual diagnostic work that AI is automating at pace, while the technical, patient-facing tasks remain stubbornly resistant to automation. This shift raises profound questions not just for radiology, but for the future of professional work more broadly. See our article expanding on AI and the division of labour between radiologists and medical imaging practitioners.
Rewriting the Rules of Education
To adapt to this shift our institutions must pivot away from rote learning of facts and instead focus on skills that build creativity, promote adaptation and interpretation:
- Teach Ambiguity: Stanford’s “d.school” assigns students open-ended challenges like “Redesign healthcare access”—no textbook answers.
- Value Nonlinear Thinkers: Greta Thunberg’s climate activism or Patagonia’s ethical supply chains show how indeterminacy drives progress.
A Call to Action
The future isn’t about competing with AI—it’s about amplifying what makes us human. To thrive:
- Upskill in Indeterminacy: Take courses on decision-making under uncertainty (e.g., MIT’s Sloan School).
- Audit Tools: Use AI for technical tasks (e.g., data analysis) but lead with human judgment.
- Demand Systemic Change: Advocate for education reforms that prioritize creativity over standardized testing.
Further Reading
- The Second Machine Age (Brynjolfsson & McAfee, 2016)
- 21 Lessons for the 21st Century (Harari, 2019)
- World Economic Forum’s Future of Jobs Report 2025
Verdict: The age of automatons is over. The age of radical thinkers has begun.
