The Enduring Currency of Curiosity: Preparing the Next Generation for an AI-Shaped Labor Market

Listen to this article:

Abstract: This article examines the evolving relationship between artificial intelligence and workforce dynamics, drawing on recent empirical evidence from large-scale usage data and labor market surveys. While AI capabilities are advancing rapidly, current deployment remains far below theoretical potential, creating a persistent gap between what AI can do and what it actually does in professional contexts. Analysis of occupation-level exposure measures reveals that workers in highly exposed roles—including programmers, customer service representatives, and financial analysts—have not experienced systematic increases in unemployment, though suggestive evidence points to slower hiring of younger workers in these fields. The article argues that adaptability, learning agility, and sustained curiosity represent durable human capital investments in an environment where specific skill requirements will continue to shift. Organizations and individuals alike benefit from focusing on these meta-competencies rather than attempting to predict which narrow technical skills will retain value. The findings support a human-centered approach to workforce development that emphasizes continuous learning, contextual judgment, and creative problem-solving—capabilities that remain complementary to AI systems even as those systems become more capable.

Every generation inherits a world mid-transformation. For Generation X, it was the advent of personal computing, mobile telephony, and the internet. For Millennials, social media and the smartphone redrew the boundaries of work and life. Today's emerging workforce faces a transformation of comparable—perhaps greater—magnitude: the rapid diffusion of generative and agentic artificial intelligence systems into professional work.

The pace of change can feel disorienting. Social media feeds overflow with predictions about which jobs will vanish, which skills will become obsolete, and whether today's educational pathways will prepare young people for the careers they will actually hold. Yet beneath the anxiety lies a more grounded reality. Recent empirical evidence suggests that while AI is reshaping certain tasks and occupations, the transformation is neither uniform nor inevitable. The gap between theoretical AI capability and actual deployment remains substantial, and the labor market consequences to date have been modest (Massenkoff & McCrory, 2026).

This article examines what the evidence reveals about AI's current and near-term impact on employment, drawing on usage data from millions of professional AI interactions and labor market surveys tracking unemployment and hiring patterns. It then considers what this means for individuals building careers and organizations developing talent in an era of ongoing technological change. The core argument is straightforward: in a world where specific job requirements shift rapidly, the most durable investment is in adaptability itself—curiosity, learning agility, and the capacity to integrate new tools into meaningful work.

The AI and Work Landscape

Defining AI Exposure in Professional Contexts

Measuring AI's impact on work requires clarity about what "exposure" means. Researchers typically define exposure at the task level: certain tasks within a job can be accelerated, automated, or augmented by AI, while others remain beyond current technical capabilities (Eloundou et al., 2023). For example, an AI system might grade homework quickly and accurately, but it cannot manage a classroom's social dynamics or respond to a student's unspoken distress. A teacher's role thus has partial exposure—some tasks are AI-amenable, others are not.

Recent work by Massenkoff and McCrory (2026) introduces a refined exposure measure that combines theoretical capability with observed usage. Theoretical capability asks: Could an AI system, in principle, double the speed of this task? Observed usage asks: Are people actually using AI for this task in professional settings? The distinction matters. Many tasks that could theoretically be accelerated by AI see little real-world deployment due to regulatory constraints, workflow integration challenges, quality-control requirements, or simply the inertia of established practices.

The researchers analyzed millions of interactions with Claude, Anthropic's AI assistant, mapping those interactions onto occupational task databases. They found that 68% of observed usage involved tasks rated as fully feasible for an LLM alone, while only 3% involved tasks deemed not feasible. This alignment suggests that users are indeed directing AI toward tasks where it can plausibly help—but it also reveals that actual usage covers only a fraction of theoretically exposed tasks (Massenkoff & McCrory, 2026).

State of Practice: The Gap Between Potential and Deployment

The gap between what AI can do and what it currently does is enormous. Consider the Computer & Math occupational category, where theoretical task coverage reaches 94%. Yet observed coverage—based on actual professional usage—stands at just 33%. Similarly, Office & Admin roles show 90% theoretical coverage but far lower observed deployment. This pattern repeats across occupational categories: potential outstrips practice by a wide margin (Massenkoff & McCrory, 2026).

Why does this gap persist? Several factors contribute. First, many tasks require integration with proprietary systems, domain-specific data, or regulatory approval processes that slow adoption. Second, organizations often need time to redesign workflows around new tools; simply introducing an AI capability does not automatically transform how work gets done (Brynjolfsson et al., 2025). Third, quality and reliability thresholds matter—an AI that performs a task adequately 80% of the time may not be deployable if the remaining 20% creates unacceptable risk. Finally, human oversight and judgment remain essential in many contexts, limiting the extent to which tasks can be fully automated even when AI could technically speed them up.

This gap is not static. As capabilities advance, as organizations learn to integrate AI into workflows, and as users become more sophisticated, observed coverage will rise. But the lag between potential and practice means that predictions based solely on theoretical capability overstate near-term disruption. The workforce adjustments we observe today reflect not what AI could do in principle, but what it is actually doing in practice—and that is still relatively limited.

Organizational and Individual Consequences of AI Deployment

Organizational Performance Impacts

From an organizational perspective, AI deployment offers the potential for productivity gains, cost reductions, and quality improvements. However, realizing these benefits requires more than installing software. It requires rethinking processes, retraining workers, and managing the cultural and operational transitions that accompany technological change.

Early evidence suggests that productivity gains from AI vary widely by context. Research using Claude conversation data found that users report saving time and improving output quality, but these gains are concentrated in specific task types—particularly those involving information synthesis, routine writing, and code generation (Tamkin & McCrory, 2025). In customer service, AI-driven tools have enabled faster response times and reduced reliance on human agents for routine inquiries, though complex or emotionally charged interactions still require human judgment (Handa et al., 2025).

Organizations that successfully integrate AI often do so by identifying high-volume, well-defined tasks where AI can provide consistent support, while preserving human involvement for tasks requiring creativity, empathy, or contextual judgment. This hybrid model—AI handling routine components, humans managing exceptions and strategic decisions—appears more common than wholesale automation (Brynjolfsson et al., 2025).

Individual Wellbeing and Career Impacts

For individual workers, AI's impact depends heavily on occupation, age, and role. Massenkoff and McCrory (2026) found that workers in the most AI-exposed occupations are disproportionately female, older, more educated, and higher-paid compared to those in less exposed roles. Computer programmers, customer service representatives, data entry keyers, and financial analysts rank among the most exposed occupations based on observed usage patterns.

Despite this exposure, aggregate unemployment rates for highly exposed workers have not increased systematically since late 2022, when ChatGPT was released. Analysis using Current Population Survey data shows that the unemployment rate gap between the most exposed quartile and the least exposed group remained essentially flat through 2025. This suggests that, to date, AI has not triggered widespread job displacement in the occupations where it is most actively used (Massenkoff & McCrory, 2026).

However, there are early signals of more subtle shifts. Evidence points to slower hiring of younger workers (ages 22–25) in highly exposed occupations. The job-finding rate for young workers entering these roles declined by approximately 14% in the post-ChatGPT period, while remaining stable for less exposed occupations. This pattern suggests that firms may be adjusting hiring practices—perhaps filling fewer entry-level roles or redesigning onboarding processes—even if they are not laying off existing employees (Massenkoff & McCrory, 2026). Similar findings emerge from payroll data analyzed by Brynjolfsson et al. (2025), who report a 6–16% decline in employment among young workers in exposed occupations, driven primarily by reduced hiring rather than increased separations.

These early impacts on young workers merit attention. If entry-level hiring slows persistently, it could narrow pathways into certain careers, even if incumbent workers remain largely unaffected. At the same time, it is possible that young workers are adapting by pursuing different roles, returning to school, or entering adjacent occupations where AI exposure is lower. The long-term implications remain uncertain, but the pattern underscores the importance of monitoring labor market dynamics at a granular level rather than relying solely on aggregate unemployment statistics.

Table 1: AI Labor Market Exposure and Organizational Impact

Occupation or Industry	Reported Employment or Hiring Impact	Key AI-Amenable Tasks	Human-Centered Tasks (Low Exposure)	Organizational Strategy or Case Study
Computer & Math	Reduced hiring of entry-level workers (ages 22–25) by 6–16% despite stable aggregate unemployment.	Code generation, information synthesis, and routine technical writing.	Complex contextual problem-solving and strategic architecture.	Amazon: Invested billions in upskilling programs like 'Career Choice' for cloud and data analytics.
Office & Administrative Support	Identified as highly exposed with potential for slower entry-level hiring.	Data entry, routine document drafting, and information retrieval.	Managing social dynamics and navigating complex organizational exceptions.	JPMorgan Chase: Utilized COiN to automate document review, redeploying staff to high-value client advisory.
Financial Analysts	Identified as a highly exposed occupation via usage patterns; current employment remains stable.	Data synthesis and routine report generation.	Strategic risk assessment and ethical decision-making.	JPMorgan Chase: Shifted operational focus from manual data extraction to strategic risk management.
Customer Service	High exposure ranking; shifting toward human-in-the-loop models for complex queries.	Responding to routine inquiries and frequent information requests.	Handling emotionally charged interactions and nuanced human judgment.	Salesforce: Established an 'AI Council' to guide ethical deployment and manage workforce transitions.
Education / Teaching	Partial exposure; low displacement risk due to high social requirements.	Grading assignments and providing routine feedback.	Classroom social dynamics and responding to student distress.	Not in source

Evidence-Based Organizational Responses

Organizations that navigate AI-driven change successfully tend to adopt a multi-pronged approach, balancing technological adoption with investment in human capital, transparent communication, and adaptive governance. Below are several evidence-based strategies that have proven effective across industries.

Transparent Communication and Expectation Management

Workers adapt more effectively to technological change when they understand what is happening and why. Organizations that communicate clearly about AI deployment—explaining which tasks will shift, what new skills will be needed, and how roles may evolve—reduce anxiety and build trust (Gimbel et al., 2025).

Effective approaches include:

• Town halls and open forums where leadership addresses AI strategy and takes questions

• Regular updates on pilot projects, lessons learned, and rollout timelines

• Clear documentation of how AI tools will be used and what protections are in place for workers

• Feedback mechanisms that allow employees to share concerns and suggest improvements

Salesforce, for example, established an "AI Council" comprising technical experts, ethicists, and employee representatives to guide deployment decisions and ensure that workforce impacts are considered alongside business objectives. This governance structure provides a formal channel for surfacing concerns and adjusting plans in response to employee input.

Skill Development and Reskilling Programs

As AI assumes responsibility for certain tasks, workers need opportunities to develop new capabilities. Effective reskilling programs focus not only on technical skills—such as learning to use AI tools—but also on higher-order skills like critical thinking, problem framing, and cross-functional collaboration.

Key elements of successful reskilling initiatives:

• Modular, on-demand training that fits into workflows rather than requiring extended time away from the job

• Peer learning and mentorship that leverages internal expertise

• Credentialing and recognition for new skills, providing tangible career benefits

• Experimentation zones where employees can test AI tools in low-stakes environments before deploying them in critical work

Amazon has invested billions in upskilling programs, offering employees access to training in cloud computing, data analytics, and machine learning. The company's "Career Choice" program pre-pays tuition for courses in high-demand fields, even if those skills are not directly relevant to the employee's current role. This approach recognizes that long-term career resilience may require transitioning to adjacent roles rather than simply adapting within a current position.

Hybrid Work Models and Task Redesign

Rather than automating entire jobs, many organizations are redesigning roles to take advantage of AI's strengths while preserving human judgment where it matters most. This often involves breaking jobs into component tasks, identifying which tasks AI can handle reliably, and restructuring workflows accordingly.

Approaches to task redesign:

• Task audits that identify high-volume, routine activities suitable for automation

• Pilot programs that test AI in controlled settings before broader rollout

• Human-in-the-loop systems where AI generates options or drafts, and humans review, refine, and approve

• Continuous feedback loops that capture user experiences and surface quality issues

JPMorgan Chase deployed an AI system called COIN (Contract Intelligence) to review legal documents and extract key data points—a task that previously required thousands of hours of lawyer and loan officer time annually. Rather than eliminating roles, the bank redeployed affected employees to higher-value activities such as client advisory and strategic risk assessment. This shift required training and role redesign, but it preserved employment while improving service quality.

Career Pathing and Internal Mobility

When AI reduces demand for certain roles, organizations can mitigate displacement by facilitating internal mobility. This requires proactive career pathing, transparent job posting systems, and support for employees making lateral or upward moves.

Best practices for internal mobility:

• Skills inventories that map employees' current capabilities and identify transferable skills

• Internal job boards with priority access for current employees

• Transition support including coaching, interview preparation, and mentorship

• Financial incentives such as retention bonuses or tuition reimbursement for employees pursuing new credentials

AT&T launched a multi-year workforce transformation initiative in response to technological change in telecommunications. The company created an internal "career marketplace" where employees could explore new roles, identify skill gaps, and access training resources. By investing in internal mobility infrastructure, AT&T reduced reliance on layoffs and external hiring while building capabilities in emerging areas like cybersecurity and software development.

Financial and Benefits Support During Transitions

Even with the best planning, some workers will face disruption. Organizations that provide financial cushions—severance, extended benefits, or transition assistance—ease the personal impact of workforce changes and maintain goodwill.

Supportive measures include:

• Generous severance packages that provide income continuity during job searches

• Extended health and retirement benefits beyond the final employment date

• Outplacement services offering resume assistance, interview coaching, and job search support

• Tuition assistance for workers pursuing education or retraining

Microsoft has committed to providing advance notice and transition support when AI-related role changes occur. The company's approach includes extended severance, career counseling, and partnerships with community colleges to facilitate reskilling. This investment recognizes that workforce transitions carry real human costs, and that organizations have a responsibility to mitigate those costs where possible.

Building Long-Term Workforce Resilience

Beyond immediate responses to AI deployment, organizations and individuals alike benefit from cultivating capabilities that remain valuable across technological shifts. These capabilities—often termed "meta-skills" or "durable skills"—transcend specific tools or domains and support adaptability over the long term.

Curiosity and Continuous Learning as Core Competencies

In a world where job requirements shift rapidly, the ability to learn quickly becomes a foundational skill. Workers who approach new tools, processes, and domains with curiosity—and who have developed effective learning strategies—can adapt more readily than those who rely on static knowledge bases.

Organizations can foster a culture of continuous learning by:

• Normalizing experimentation: Create space for employees to try new tools, fail safely, and share what they learn.

• Recognizing learning behaviors: Reward curiosity, initiative, and skill development, not just performance on existing metrics.

• Providing learning infrastructure: Offer access to online courses, conferences, books, and mentorship programs.

• Modeling learning from the top: Leaders who visibly invest in their own development signal that learning is valued at all levels.

Research on learning organizations suggests that cultures emphasizing curiosity and knowledge-sharing outperform those focused narrowly on efficiency and standardization, particularly in volatile environments (Brynjolfsson et al., 2025). As AI continues to evolve, organizations that cultivate learning mindsets will adapt more fluidly than those that treat skills as fixed assets.

Judgment, Creativity, and Contextual Intelligence

AI excels at pattern recognition, data synthesis, and routine decision-making. It struggles with ambiguity, novelty, and contexts where human values and social dynamics matter. This means that roles requiring judgment—especially in situations with incomplete information, competing priorities, or ethical complexity—remain strongly human-centered.

Judgment-intensive tasks include:

• Strategic planning where goals are contested or emergent

• Creative problem-solving requiring lateral thinking or analogical reasoning

• Stakeholder management involving negotiation, empathy, and trust-building

• Ethical decision-making where competing values must be weighed and justified

Organizations can strengthen judgment capabilities by:

• Case-based learning: Use real scenarios to practice decision-making under uncertainty.

• Diverse teams: Bring together people with different backgrounds and perspectives to challenge assumptions.

• Reflection practices: Build time for debriefing, learning from mistakes, and refining mental models.

• Ethical training: Equip employees to recognize and navigate moral dilemmas in their work.

AI can support judgment by surfacing relevant information, generating options, or highlighting risks—but the ultimate decision remains human. Roles that emphasize judgment will likely prove more durable than those focused on routine execution.

Collaboration and Relationship-Building in Distributed Work

As AI handles more transactional and analytical tasks, the relative importance of interpersonal skills rises. Building trust, navigating conflict, inspiring teams, and forging partnerships require emotional intelligence and social awareness that AI systems do not possess.

Collaborative capabilities include:

• Active listening and empathetic communication

• Conflict resolution that balances interests and preserves relationships

• Team coordination across functions, geographies, and cultures

• Influencing and persuasion in contexts where authority is limited

Organizations can invest in collaboration by:

• Facilitating cross-functional projects that require diverse expertise and shared accountability.

• Providing communication training focused on feedback, active listening, and difficult conversations.

• Creating social infrastructure such as team rituals, informal gatherings, and mentorship programs.

• Rewarding collaborative behaviors in performance reviews and promotion decisions.

In an AI-augmented workplace, humans will increasingly focus on the "glue work" that holds teams and projects together—coordinating, aligning, and building shared understanding. These capabilities are difficult to automate and become more valuable as technical work becomes more distributed and interdependent.

Conclusion

The labor market our children will enter will differ from the one we know. That has always been true, and it will remain true. What matters is not predicting specific job titles or technologies, but cultivating the mindsets and capabilities that support adaptation across change.

The evidence to date suggests that AI's impact on employment is real but nuanced. Theoretical exposure far exceeds actual deployment, and even in highly exposed occupations, aggregate unemployment has not risen systematically. However, early signals—particularly slower hiring of young workers—warrant continued monitoring. Organizations and policymakers should prepare for the possibility that AI's labor market effects will manifest gradually and unevenly, with certain cohorts and roles experiencing disruption while others see little change.

For individuals, the path forward centers on curiosity, learning agility, and the cultivation of distinctly human capabilities—judgment, creativity, empathy, and collaboration. These are not fixed traits but skills that can be developed through deliberate practice and supportive environments. Organizations that invest in these capabilities—through transparent communication, reskilling programs, task redesign, internal mobility, and a culture of continuous learning—will build more resilient workforces and more adaptable institutions.

The future belongs not to those who can predict which skills will remain valuable, but to those who can learn, adapt, and integrate new tools into meaningful work. That capacity—curiosity sustained over a lifetime—is the most durable investment we can make, for ourselves and for the generations that follow.

Research Infographic

The Curiosity Advantage Slide Deck

References

1. Brynjolfsson, E., Chandar, B., & Chen, R. (2025). Canaries in the coal mine? Six facts about the recent employment effects of artificial intelligence. Digital Economy.

2. Eloundou, T., Manning, S., Mishkin, P., & Rock, D. (2023). GPTs are GPTs: An early look at the labor market impact potential of large language models. arXiv preprint arXiv:2303.10130.

3. Gimbel, M., Kinder, M., Kendall, J., & Lee, M. (2025). Evaluating the impact of AI on the labor market: Current state of affairs. Research Report, The Budget Lab at Yale.

4. Handa, K., Tamkin, A., McCain, M., Huang, S., Durmus, E., Heck, S., Mueller, J., Hong, J., Ritchie, S., Belonax, T., Troy, K. K., Amodei, D., Kaplan, J., Clark, J., & Ganguli, D. (2025). Which economic tasks are performed with AI? Evidence from millions of Claude conversations.

5. Massenkoff, M., & McCrory, P. (2026). Labor market impacts of AI: A new measure and early evidence. Anthropic.

6. Tamkin, A., & McCrory, P. (2025). Estimating AI productivity gains from Claude conversations.

---

Jonathan H. Westover, PhD is Chief Research Officer (Nexus Institute for Work and AI); Associate Dean and Director of HR Academic Programs (WGU); Professor, Organizational Leadership (UVU); OD/HR/Leadership Consultant (Human Capital Innovations). Read Jonathan Westover's executive profile here.

---

Suggested Citation: Westover, J. H. (2026). The Enduring Currency of Curiosity: Preparing the Next Generation for an AI-Shaped Labor Market. Human Capital Leadership Review, 31(4). doi.org/10.70175/hclreview.2020.31.4.2