The Research
88% of organizations use AI.
81% see no real return.
This is the gap. It is not a technology problem. It is a thinking problem — and it has a specific, researchable cause.
of organizations now use AI in at least one function
report no significant bottom-line impact from their AI investments
have redesigned workflows around AI — 2.8× more likely to see real EBIT impact
Why the gap exists
Organizations add AI to existing work.
The 21% redesign the work itself.
The factor most correlated with meaningful AI returns is not budget, technology choice, or sector. It is workflow redesign: examining how work actually happens and rebuilding it around what AI can now do reliably.
The other 79% add AI to existing processes and measure the result in disappointment. The system stays the same. The system produces the same results, slightly faster.
The investment ratio that changes everything
For every dollar spent on AI technology, winning organizations invest five in people: judgment, strategic thinking, and the capacity to use AI well.
Automation Exposure by Skill Type
Not all skills are equally at risk.
McKinsey mapped 800+ occupations and 2,100+ activities. The pattern is consistent: physical and cognitive execution is most exposed; social, creative, and judgment-based work is least.
How Work Evolves
Same role. Different work.
Professional work migrates upward — from execution to interpretation, from preparation to judgment. AI handles the preparation; you handle the meaning.
The Value Migration
Your job title stays.
Your actual work transforms.
Professional work migrates upward. AI handles the preparation; you handle the judgment.
Hours of data analysis
Minutes of analysis → hours of interpretation & decision-making
AI compiles and patterns; you contextualize and decide
Research compilation & document prep
AI first drafts → your question-framing & quality control
The bottleneck shifts from finding information to asking the right questions
Demand forecasting from spreadsheets
AI-generated forecasts → scenario planning & strategic bets
Prediction is automatable; wisdom about which prediction matters is not
Routine client updates
AI handles updates → you focus on relationship & trust-building
One-third of work relies on social-emotional capacities AI can't replicate
Learning new tools every quarter
Building transferable judgment that works across any tool
"Transferable skills matter more than specialization when circumstances shift rapidly" — MGI
800+ Occupations Mapped
Where does your work sit?
McKinsey segments all occupations by which partner — people, agents, or robots — handles the majority of tasks. This determines your disruption exposure and your opportunity.
Minimal Disruption
Nurses, therapists, coaches, social workers, firefighters
Evolving Roles
Sales, HR, educators, consultants, project managers
Significant Shift
Accountants, lawyers, developers, analysts, writers
Physical Evolution
Electricians, plumbers, tradespeople, technicians
High Automation
Logistics workers, welders, assembly line, warehouse
All-Three Roles
Medical assistants, receptionists, operations coordinators
The Partnership Model
People. Agents. Robots.
McKinsey maps 800+ occupations into a three-way partnership. The shift isn't replacement — it's redistribution of what each partner does best.
McKinsey maps every occupation into a primary partnership model — the mix of humans, AI agents, and physical robots that handles the majority of its tasks. The $2.9 trillion figure represents the combined economic value unlocked by 2030 when all three partners operate at full capacity.
Social interaction, ethical judgment, creative direction, leadership. The work that requires genuine human presence and accountability.
Knowledge synthesis, research, drafting, analysis, coordination. Software agents handling cognitive tasks at scale.
Physical logistics, manufacturing, warehouse operations. Robotics still lack dexterity for complex physical environments.
43%
People
Social interaction, ethical judgment, creative direction, leadership. The work that requires genuine human presence and accountability.
44%
AI Agents
Knowledge synthesis, research, drafting, analysis, coordination. Software agents handling cognitive tasks at scale.
13%
Robots
Physical logistics, manufacturing, warehouse operations. Robotics still lack dexterity for complex physical environments.
— Over 70% of skills employers seek today apply to both automatable and non-automatable work — they evolve, they don't disappear. MGI, "Agents, Robots, and Us," 2026
The Fastest-Growing Skill
7× in two years.
AI fluency demand in global job postings grew from roughly 1 million to 7 million in just two years — faster than any professional skill on record. But fluency is not prompting. It is judgment.
AI fluency is not
Knowing every prompt trick. Chasing new tools. Replacing your judgment with AI output. Speed without direction.
AI fluency is
Knowing when to use AI and when not to. Framing the right questions. Interpreting results with domain expertise. Deliberate judgment.
The Netherlands, 2030
More jobs. Fewer prepared people.
The Netherlands isn't losing jobs — it's reshuffling them. A net gain of 250,000 roles by 2030. But only for those who make the shift intentionally.
McKinsey Global Institute, "How AI Reshapes Work and Skills in Europe," 2026
The question isn't whether this shift happens. It's whether you're positioned for the 400,000 or the 150,000.
AI exposure by sector
% FTE hours automatable — EU comparison
Cross the gap
The gap is not fixed. It is a choice.
Whether you are an individual professional or an organization, the path from the 79% to the 21% is the same: examine how work actually happens, and be willing to change it.