Technology, Jobs, and the Future of Work: How AI, Automation, Robotics, and Digital Platforms Can Create Prosperity, Displace Workers, Reshape Wages, and Redefine Economic Opportunity

1. How much productivity does technology create?

2. How quickly does demand expand in response?

3. Which new jobs and industries emerge?

4. Can workers transition into those opportunities?

5. Who captures the resulting income and wealth?

The strongest policy and business response is not to stop technological adoption. It is to increase society’s capacity to convert technological progress into broadly distributed prosperity.

That requires:

• Better education and lifelong learning
• Employer investment in training
• Skills-based hiring
• Stronger worker-transition systems
• Portable benefits
• Digital infrastructure
• Labor-market information
• Entrepreneurship and business creation
• Competition policy
• Responsible AI governance
• Policies that support wage growth and economic mobility

The central question is not:

Will technology destroy jobs or create jobs?

It is:

Under what conditions will technology create enough productive activity, new demand, better work, and accessible opportunity to outweigh the disruption it causes?

1. Technology Does Not Have One Effect on Employment

Public debate often frames technology as either:

• A job destroyer
• A job creator

Both descriptions are incomplete. A technological innovation can eliminate one activity while creating demand elsewhere.

Online banking reduced the need for some in-person transactions while increasing demand for:

• Software engineers
• Cybersecurity professionals
• Data specialists
• Digital-product managers
• Compliance experts
• Customer-support systems

E-commerce reduced demand for some physical retail activities while increasing demand for:

• Warehousing
• Logistics
• Delivery
• Software
• Digital marketing
• Payment technology

Artificial intelligence may reduce the labor needed for routine document production while increasing demand for:

• AI integration
• Evaluation
• Data engineering
• Security
• Model governance
• Workflow redesign

The complete economic effect depends on several interacting mechanisms.

2. The Five Main Employment Effects of Technology

2.1 Substitution

Technology performs work previously completed by people.

Examples include:

• Automated checkout
• Industrial robots
• Document classification
• Routine bookkeeping
• Basic customer-service responses

Substitution can reduce demand for a particular activity or occupation.

2.2 Augmentation

Technology helps people produce more or better work.

Examples include:

• AI-assisted medical analysis
• Engineering simulation
• Coding assistants
• Automated financial analysis
• Navigation and routing tools

Augmentation may increase the productivity and value of the worker.

2.3 Demand expansion

Technology can lower the cost of a product or service. Lower prices may increase consumption. For example, cheaper computing expanded demand for software, digital services, and data-intensive products. In some cases, increased demand creates more employment than automation removes.

2.4 New task and industry creation

Technology creates work that previously did not exist.

Examples include:

• Mobile-app development
• Cloud engineering
• Social-media management
• Cybersecurity operations
• AI governance
• Creator-economy services

The future labor market will include roles that cannot yet be predicted precisely.

2.5 Reallocation

Technology changes which companies, regions, occupations, and workers receive income. An economy may grow while specific communities or professions decline. This is why national employment growth does not guarantee that every worker benefits.

3. Technical Automation Potential Is Not the Same as Job Loss

A common mistake is to interpret technical automation estimates as employment forecasts. If technology can theoretically perform a task, that does not mean the task will be automated immediately.

Adoption depends on:

• Cost
• Reliability
• Regulation
• Customer acceptance
• Infrastructure
• Skills
• Organizational readiness
• Investment cycles

McKinsey’s 2025 research estimated that current technologies could theoretically automate activities occupying about 57 percent of US work hours, including approximately 44 percent through software agents and 13 percent through robots. It explicitly warns that this does not imply that the same percentage of jobs will disappear. Jobs are combinations of activities. A profession may remain economically important even after a substantial share of routine work is automated.

4. Why Jobs Usually Change Before They Disappear

Consider a financial analyst.

The role may include:

• Gathering data
• Cleaning spreadsheets
• Creating charts
• Explaining performance
• Challenging assumptions
• Advising executives

AI may reduce the time spent gathering, cleaning, and formatting information.

The analyst may then spend more time on:

• Interpretation
• Scenario analysis
• Risk
• Communication
• Decision support

The occupation remains, but the valuable portion of the job changes.

This pattern can occur in:

• Law
• Accounting
• Engineering
• Medicine
• Education
• Marketing
• Software development
• Management

Organizations should therefore redesign roles rather than assume that existing job descriptions will remain stable.

5. Automation Can Increase the Value of Complementary Skills

When technology becomes better at one activity, the economic value of complementary human skills may rise.

If AI generates code rapidly, then the ability to:

• Define the product
• Design architecture
• Verify quality
• Secure systems
• Understand customers

may become more important.

If AI drafts documents, then judgment about:

• Accuracy
• Relevance
• persuasion
• ethics
• legal consequences

may become more valuable. McKinsey’s recent research indicates that more than 70 percent of skills currently sought by employers are used in both automatable and non-automatable activities. This suggests that many skills will be adapted rather than simply made obsolete.

6. Productivity Is the Main Economic Opportunity

Productivity measures how much value an economy produces from its available labor, capital, technology, and resources.

Sustained productivity growth can support:

• Higher wages
• Lower prices
• Higher profits
• More investment
• Better public services
• Shorter working time
• Higher living standards

AI and automation could raise productivity by reducing the cost and time required for many activities. McKinsey estimates that people, agents, and robots could unlock substantial economic value if organizations redesign entire workflows and develop the skills required to use the technology effectively. However, productivity potential does not become real merely because tools exist.

Organizations must change:

• Processes
• Data
• Roles
• Management
• Incentives
• Architecture
• Customer behavior

7. The Productivity Paradox

New technologies often take time to appear in economic productivity statistics.

Companies may initially spend more because they must:

• Purchase systems
• Integrate data
• Train workers
• Redesign processes
• Manage transition
• Operate old and new systems simultaneously

Productivity gains may emerge only after complementary changes are completed. An organization that adds AI to an inefficient process may create faster inefficiency. The deeper value comes from redesigning the entire workflow.

8. Productivity Growth Does Not Guarantee Wage Growth

An economy can become more productive without distributing the gains evenly.

Productivity gains may flow toward:

• Business owners
• Investors
• Highly skilled workers
• Technology platforms
• Intellectual-property owners

Workers whose tasks are easily substituted may experience:

• Lower bargaining power
• Wage pressure
• Fewer hours
• Reduced employment

The relationship between productivity and wages depends on institutions such as:

• Labor-market competition
• Worker bargaining power
• Tax policy
• Corporate governance
• Education
• Ownership
• Competition policy

Technology creates a surplus. Society determines how that surplus is distributed.

9. Technology Can Increase Inequality Through Several Channels

9.1 Skill-biased change

Technology may raise demand for highly educated or specialized workers while reducing demand for routine work.

9.2 Capital ownership

Those who own technology, intellectual property, and platforms may receive a disproportionate share of gains.

9.3 Superstar firms

Digital markets may allow a small number of firms to serve enormous markets, concentrating profits and talent.

9.4 Geographic concentration

High-value technology industries may cluster in selected regions.

9.5 Unequal access

Workers and companies with better infrastructure, education, capital, and data may adopt technology faster. The original McKinsey briefing noted that digital gains were distributed unevenly across countries and that infrastructure and skills strongly affected whether individuals and businesses could participate.

10. AI Exposure Is Not Limited to Low-Wage Work

Earlier automation often affected routine physical and clerical tasks.

Generative AI has expanded exposure into occupations involving:

• Language
• Analysis
• Coding
• Research
• Professional services

This does not necessarily mean that highly educated workers face the greatest risk of unemployment. They may be better positioned to use AI as an augmentation tool.

The effect depends on whether the technology:

• Replaces the worker
• Increases the worker’s productivity
• Changes the worker’s responsibilities
• Expands demand for the service

OECD research shows that AI exposure is affecting skill demand within occupations, with business and management capabilities remaining important in many highly exposed professions.

11. The Labor Market Can Grow While Transitions Become More Difficult

The US labor market is projected to grow overall through 2034. That does not eliminate transition challenges.

New jobs may differ from declining jobs in:

• Skills
• Location
• wages
• education
• schedule
• working conditions

McKinsey estimated that approximately 12 million additional US occupational transitions could occur by 2030, with a large share concentrated in office support, food service, customer service, sales, and production work. A worker displaced from one occupation cannot necessarily move immediately into a growing profession.

The transition may require:

• Training
• Credentialing
• income support
• relocation
• childcare
• employer acceptance

12. Net Job Creation Can Hide Personal Loss

Suppose an economy loses one million administrative jobs and creates 1.2 million healthcare, technology, and skilled-trade jobs. At the national level, employment rises.

For the displaced administrative worker, the outcome depends on whether the new job is:

• Accessible
• nearby
• adequately paid
• compatible with family responsibilities
• available without years of education

The distribution and accessibility of opportunity are therefore as important as the aggregate number of jobs.

13. Worker Transitions Need an Economic Infrastructure

Successful transitions require more than motivational advice.

A complete transition system may include:

• Career guidance
• Skills assessment
• Training
• recognized credentials
• apprenticeships
• paid work experience
• income support
• childcare
• transportation
• relocation assistance
• employer participation

The original McKinsey briefing called for stronger transition support, redesigned safety nets, better labor-market matching, and greater employer involvement in training. These recommendations are even more important when technologies evolve quickly.

14. Employers Must Become Education Institutions

Companies frequently complain about skills shortages while expecting schools, universities, or workers to carry the entire cost of preparation. This model is increasingly unrealistic.

Employers possess valuable information about:

• Emerging skills
• real workflows
• tools
• performance requirements

They should contribute through:

• Apprenticeships
• internal academies
• paid training
• rotations
• mentoring
• entry-level pathways
• partnerships with colleges

McKinsey’s original recommendations explicitly called for the private sector to take a more active role in education and capability building.

15. Hiring for Skills Can Expand Opportunity

Rigid degree and experience requirements can exclude capable candidates.

Skills-based hiring can recognize:

• Demonstrated ability
• Work samples
• apprenticeships
• certifications
• military experience
• career transitions
• independent projects

This is particularly important when labor markets need millions of people to move into new occupational categories. McKinsey recommends broader hiring approaches that focus more heavily on skills and competencies rather than credentials alone.

16. Education Must Prepare People for Repeated Change

The future labor market will require more than one period of education followed by several decades of stable employment. Workers may need to learn repeatedly.

Education systems should strengthen:

• Basic literacy and numeracy
• digital literacy
• analytical thinking
• creativity
• systems thinking
• communication
• AI literacy
• career adaptability

The original McKinsey briefing emphasized lifelong learning, STEM capability, creativity, and critical thinking.

17. Digital Access Is Economic Infrastructure

Participation in the modern economy increasingly depends on:

• Broadband
• devices
• digital identity
• payments
• cloud access
• digital skills

Without this infrastructure, workers and small businesses may be excluded from:

• Remote work
• online education
• digital commerce
• financial services
• AI tools

The original briefing treated digital infrastructure as a prerequisite for capturing the benefits of technological adoption. The principle remains valid even though global connectivity has improved since 2017.

18. Digital Platforms Improve Matching but Create New Risks

Technology platforms can connect:

• Workers with jobs
• freelancers with clients
• students with courses
• employers with skills
• entrepreneurs with customers

Better matching can reduce unemployment and help people find opportunities outside their immediate geography.

However, platform-mediated work may also involve:

• Income volatility
• opaque algorithms
• weak benefits
• unilateral rule changes
• limited bargaining power

Technology-enabled labor markets need transparency and appropriate worker protections.

19. Independent Work Can Expand Economic Participation

Digital tools make it easier for individuals to:

• Find clients
• market services
• collaborate remotely
• receive payments
• operate small businesses

AI can further reduce the administrative burden of entrepreneurship through:

• Marketing assistance
• bookkeeping
• customer support
• research
• content creation
• workflow automation

This could make one-person and small-team businesses more capable.

The economic benefit will be greater when workers also have access to:

• Health protection
• retirement systems
• training
• credit
• legal support

20. Job Creation Matters as Much as Training

Training workers for jobs that do not exist is not a complete strategy.

Governments and businesses must also support demand through:

• Investment
• infrastructure
• entrepreneurship
• business formation
• research
• market competition
• regional development

The original McKinsey briefing included job creation and digitally enabled entrepreneurship among its central solution areas. The labor market needs both employable workers and employers willing to hire them.

21. Competition Policy Affects Labor Outcomes

If a small number of companies control important digital markets, they may gain substantial power over:

• Suppliers
• workers
• customers
• innovation

Limited competition can weaken the relationship between productivity and broadly shared prosperity.

Healthy competition may support:

• New business formation
• wage competition
• innovation
• customer choice
• regional opportunity

The future of work is therefore partly a question of market structure.

22. Technology Can Improve Job Quality

Automation can remove work that is:

• Dangerous
• physically exhausting
• repetitive
• unpleasant
• error-prone

AI can help workers:

• Find information
• reduce administration
• make better decisions
• serve customers
• learn faster

The OECD notes that AI can improve productivity, job quality, and workplace safety, while also creating risks involving automation, privacy, bias, transparency, and worker agency. The outcome depends on how technology is introduced.

23. Technology Can Also Degrade Job Quality

Poorly governed systems may increase:

• Surveillance
• work intensity
• algorithmic control
• unpredictability
• loss of autonomy
• error-correction burdens

A worker may become responsible for correcting AI mistakes while being evaluated against faster output targets. Job quality should therefore be measured alongside productivity.

Useful indicators include:

• Autonomy
• schedule stability
• safety
• learning
• workload
• voice
• income
• career progression

24. The Human-Machine Division of Work Must Be Designed

Organizations should not assume that technology vendors will determine the ideal allocation of tasks.

For each workflow, leaders should decide:

• Which activities machines perform
• Which require human approval
• Which decisions remain human-led
• Who is accountable
• How errors are corrected
• How workers develop

McKinsey’s original article called for innovation in how people and machines work together, recognizing that greater interaction would require new interfaces, skills, investments, and potentially new wage models.

25. AI Agents Add a New Organizational Layer

Software agents can potentially:

• Research
• generate documents
• update systems
• monitor operations
• coordinate tasks
• communicate with other agents

Organizations must decide whether agents function as:

• Tools
• supervised digital workers
• workflow components
• autonomous service providers

The classification matters because it affects:

• Access
• governance
• performance measurement
• liability
• security

Agentic systems do not remove human accountability. They make accountability design more important.

26. Robotics Will Affect Physical Work Unevenly

Robots perform best in environments that are:

• Structured
• predictable
• repeatable
• physically accessible

Warehouses and factories may automate more rapidly than homes, construction sites, or complex care environments. McKinsey’s recent research finds that physical automation potential remains constrained by fine motor skills, dexterity, mobility, and situational awareness in unstructured environments. This means many physical occupations may be augmented before they are fully automated.

27. Demographics May Increase the Need for Automation

Aging populations can reduce labor-force growth while increasing demand for:

• Healthcare
• elder care
• public services
• retirement support

Automation can help compensate for selected labor shortages.

However, many care activities depend on:

• Empathy
• physical presence
• trust
• judgment

The future economy may simultaneously become more automated and more dependent on human care work.

28. Migration Will Remain Part of the Workforce Solution

Labor shortages cannot always be solved quickly through automation or domestic training.

Migration can support:

• Healthcare
• technology
• agriculture
• construction
• research
• entrepreneurship

Migration policy affects both economic growth and the ability of regions to fill critical roles.

It also requires attention to:

• housing
• integration
• public services
• recognition of credentials

29. Policies for Income Stability May Need Updating

When technology creates faster transitions, workers may experience periods of:

• unemployment
• lower wages
• retraining
• reduced hours

Possible policy tools include:

• Unemployment insurance
• wage insurance
• portable benefits
• training accounts
• earned-income supplements
• conditional cash assistance

The original McKinsey article also suggested testing broader approaches to income support if automation created significant employment or wage pressure. The correct design depends on national institutions and labor-market conditions.

30. Universal Basic Income Is One Possible Tool, Not a Complete Strategy

Universal basic income is frequently proposed as a response to automation.

It could provide:

• Income stability
• bargaining power
• support during transitions
• freedom to learn or start businesses

It would not directly solve:

• Skill shortages
• job quality
• regional decline
• healthcare access
• social isolation
• loss of professional purpose

Income policy should therefore be considered alongside job creation, training, infrastructure, and labor-market reform.

31. Human Capital Should Be Treated as an Investment

Companies typically treat spending on:

• Machinery
• software
• buildings

as investment. Training may be treated as a discretionary expense. This creates a mismatch when workforce capability is essential to realizing technology value. The original McKinsey briefing recommended incentives that encourage businesses to treat human-capital development more like other forms of investment.

Possible mechanisms include:

• Tax incentives
• training credits
• apprenticeship support
• public-private programs
• co-funded education

32. Measuring the Future of Work Requires Better Data

Traditional labor-market statistics may not capture rapidly changing:

• Skills
• tasks
• freelance work
• AI usage
• hybrid roles
• agent-supported work

Governments and researchers need better information about:

• Real-time skill demand
• wage changes
• occupational transitions
• geographic mismatches
• technology adoption
• job quality

Official projections remain essential, but they should be complemented by more frequent and granular data.

33. What Companies Should Do

Map work, not only jobs Break roles into activities. Evaluate automation realistically Consider technical feasibility, economics, risk, and adoption. Redesign complete workflows Do not add AI to broken processes. Invest in workers Create learning and mobility pathways. Protect entry-level development Replace routine learning tasks with apprenticeships, simulations, and supervised human-AI work. Measure outcomes Track productivity, quality, employee experience, wages, and customer value.

Share gains visibly Use productivity improvements to support combinations of growth, wages, learning, and better work.

34. What Governments Should Do

Strengthen education Support foundational, technical, and adaptive skills. Improve transition systems Connect training with actual jobs. Invest in infrastructure Expand broadband, transportation, energy, and regional development. Modernize social protection Support people across different forms of employment. Promote competition and entrepreneurship Ensure that technological markets remain open enough for new firms and ideas. Govern high-impact AI Protect workers and citizens from opaque, discriminatory, or unsafe systems.

35. What Workers Should Do

Workers cannot control macroeconomic change, but they can improve resilience by:

• Building transferable skills
• Developing domain expertise
• Learning to use AI
• Creating evidence of practical ability
• Maintaining professional networks
• Following labor-market trends
• Preparing for repeated learning

The objective is not to chase every new tool. It is to combine durable human capabilities with changing technical tools. A Framework for Evaluating Any Technology’s Employment Impact Question 1: What tasks can it perform? Avoid beginning with job titles. Question 2: Is automation economically attractive? Technical ability is not enough. Question 3: What complementary tasks grow? New technology often increases demand for adjacent work. Question 4: Does lower cost expand demand? Cheaper services may create more total employment. Question 5: Which new products become possible?

Innovation can create industries. Question 6: Which workers and regions bear the transition cost? Aggregate gains may hide concentrated losses. Question 7: Who owns the technology? Ownership affects income distribution. Question 8: Can workers access new opportunities? Skills, geography, credentials, and family responsibilities matter. Question 9: How is job quality affected? Productivity is not the only outcome. Question 10: Which policies and institutions shape distribution? Technology does not determine distribution by itself.

Key Takeaways

1. Technology simultaneously substitutes, augments, creates, and reallocates work.

2. Technical automation potential should never be confused with a forecast of job losses.

3. Jobs usually change at the task level before entire occupations disappear.

4. Complementary human skills may become more valuable as machines improve.

5. Productivity growth is the main economic opportunity created by AI and automation.

6. Technology purchases do not automatically generate productivity.

7. Productivity gains do not automatically produce wage growth or shared prosperity.

8. Technology can increase inequality through skill differences, capital ownership, market concentration, and unequal access.

9. An economy can create jobs overall while millions of workers face difficult occupational transitions.

10. Reskilling must connect learning with practical experience and real employment.

11. Employers must become more active providers of training and career pathways.

12. Skills-based hiring can expand access to growing occupations.

13. Digital infrastructure is now essential economic infrastructure.

14. Labor platforms improve matching but can also create instability and algorithmic control.

15. Entrepreneurship and job creation are necessary complements to worker training.

16. Technology can improve or degrade job quality depending on implementation.

17. Human-machine work allocation must be designed explicitly.

18. Robotics will affect structured physical environments faster than highly variable ones.

19. Income protection, portable benefits, and transition support may need modernization.

20. The future of work is shaped by economic institutions and political choices, not technology alone.

Frequently Asked Questions

Does technology create or destroy jobs?

It does both. It replaces some activities while creating new products, industries, tasks, and sources of demand.

Will AI eliminate most employment?

Current research does not support treating technical automation potential as a direct job-loss forecast. Many occupations will change as selected tasks are automated or augmented.

What percentage of work can AI automate?

McKinsey’s 2025 analysis estimated that currently demonstrated agents and robots could theoretically perform activities accounting for approximately 57 percent of US work hours. This is technical potential, not a prediction of actual adoption or unemployment.

Why does automation not always reduce employment?

Automation may reduce prices, increase demand, create complementary work, and support new products and industries.

Which jobs are most exposed?

Tasks that are repetitive, digitally represented, rules-based, and easy to evaluate are generally more exposed.

Are professional jobs exposed?

Yes. Generative AI can affect writing, analysis, coding, legal research, financial work, and other professional activities. Exposure does not necessarily mean replacement.

Which human skills remain valuable?

Important capabilities include:

• Judgment
• Communication
• leadership
• empathy
• negotiation
• domain expertise
• ethical reasoning
• problem definition

Is productivity the same as working harder?

No. Productivity means producing more value from available resources. It can come from better technology, processes, skills, management, and capital.

Will higher productivity increase wages?

It can, but the outcome depends on labor-market institutions, ownership, competition, bargaining power, and company policies.

Can AI increase inequality?

Yes. The gains may concentrate among technology owners and highly skilled workers unless access, competition, education, and distribution are addressed.

Will the United States have fewer jobs in 2034?

BLS projects overall US employment growth of approximately 5.2 million jobs from 2024 to 2034, although some occupations will grow while others decline.

What is an occupational transition?

It is a move from one occupational category to another, usually involving a meaningful change in tasks or skills.

How many US occupational transitions may occur?

McKinsey estimated that approximately 12 million additional occupational transitions may be needed by 2030.

Can online courses solve displacement?

Courses can help, but successful transitions usually also require practical experience, employer recognition, income support, and access to available jobs.

Should companies pay for worker training?

Companies have strong economic reasons to invest in training because workforce capability is necessary to realize the value of technology.

What is skills-based hiring?

It is hiring based more heavily on demonstrated competencies and practical ability than on degrees, titles, or employer prestige alone.

Will independent work grow?

Digital platforms and AI can make independent work easier, although issues involving income stability, benefits, and bargaining power remain.

Can universal basic income solve technological unemployment?

It may provide income stability, but it does not by itself create jobs, skills, community, healthcare, or meaningful work.

What are portable benefits?

They are benefits that remain with the worker across jobs, contracts, and self-employment.

How can governments help?

Governments can invest in:

• Education
• infrastructure
• apprenticeships
• transition support
• portable benefits
• regional development
• competition
• responsible AI rules

How can workers prepare?

Workers can strengthen transferable skills, domain expertise, AI literacy, professional networks, and evidence of practical ability.

Conclusion

Technology does not move through the economy like a single wave that simply removes jobs in its path. It changes the cost of production. It alters which skills are valuable. It creates new businesses. It eliminates some activities. It increases the scale of others. It shifts income among workers, companies, investors, and regions. The result can be greater prosperity. It can also be greater inequality and insecurity. Both outcomes are economically possible. The original McKinsey briefing was correct to frame the future of work as a challenge requiring education, employer training, infrastructure, transition support, job creation, technology-enabled matching, and new forms of human-machine collaboration. The arrival of generative and agentic AI has made that framework more urgent.

AI can now affect tasks once considered safely inside the domain of educated knowledge workers. Robotics is moving into more physical environments. Digital platforms are making labor markets more global and more flexible. The speed of change may outpace institutions designed for slower industrial transitions. The answer is not to prevent technological progress. Without productivity growth, aging economies may struggle to raise living standards, fund public services, or respond to labor shortages. The answer is to build a system capable of absorbing progress. That system must help workers move. It must help employers develop people. It must encourage new companies and industries. It must protect individuals during transitions. It must make digital infrastructure broadly accessible.

It must ensure that technological markets remain competitive. It must connect productivity growth with wage growth, better services, and wider economic opportunity. The future of work will not be judged only by how intelligent machines become. It will be judged by whether human institutions become intelligent enough to use them well.

The defining question is not:

How many jobs will technology eliminate?

It is:

Can society convert technological capability into enough productivity, investment, new demand, worker mobility, and broadly shared income to make the transition economically and socially worthwhile?

Relevant Articles and Resources

1. Technology, Jobs, and the Future of Work

McKinsey Global Institute’s foundational briefing on automation, digital adoption, education, worker transitions, job creation, infrastructure, and productivity.

2. Agents, Robots, and Us: Skill Partnerships in the Age of AI

McKinsey’s 2025 analysis of how people, software agents, and robots may divide activities and reshape skill demand.

3. Generative AI and the Future of Work in America

Research on US occupational transitions, automation, infrastructure investment, demographic trends, and changing labor demand through 2030.

4. Employment Projections, 2024 - 2034

The US Bureau of Labor Statistics’ official outlook for industry and occupational employment growth.

5. Future of Work

OECD research covering artificial intelligence, job quality, telework, skills, globalization, demographic change, and labor-market policy.

6. How AI Is Changing Work and Skill Demand

OECD analysis of how AI exposure is influencing activities and skills within occupations.

7. A New Future of Work: The Race to Deploy AI and Raise Skills

McKinsey research on automation, occupational transitions, productivity, and skill development in Europe and the United States.

8. AI in Work, Innovation, Productivity, and Skills

The OECD program examining how AI affects training, labor markets, productivity, job quality, and human-centered policy.