By Jim Shimabukuro (assisted by Claude)
Editor
David Brooks, in “The People Who Will Thrive in the AI Age” (Atlantic, 28 June 2026), cited two pieces of research that together tell a story both bracing and clarifying (1). The first, from ActivTrak’s Productivity Lab — which analyzed more than 443 million hours of digital workplace activity across some 163,000 workers — found that AI adoption has made work faster, denser, and more demanding, not easier (2,3). Collaboration surged 34 percent, multitasking rose 12 percent, and weekend work climbed more than 40 percent. The second, from UC Berkeley’s Haas School of Business, found that workers who used AI did not use their freed-up time for rest; they used it to take on tasks they had previously outsourced or deferred (4). Xingqi Maggie Ye, a Haas doctoral student, spent eight months observing 200 employees at a technology company and discovered that AI expanded what workers felt capable of and willing to tackle. Scope grew; boundaries between work and personal time dissolved.
Together, these studies imply a sharp distinction: AI advantages some workers more than others, and that distinction is not primarily technical. It is dispositional. That is the territory this article explores. Drawing on research from sources including the World Economic Forum, Microsoft, Boston Consulting Group, Accenture, and peer-reviewed journals, the following ten traits profile the worker who gains ground when AI enters the room — and, by contrast, identify the habits of mind that leave others behind.
1. A Genuine Appetite for Hard Thinking
Psychologists call it “need for cognition”: the tendency to seek out and enjoy intellectual challenges rather than settle for the first plausible answer. Brooks pointed to this quality as a dividing line in the AI age, and research bears him out (1). A 2025 study presented at the ACM CHI Conference — the leading academic venue for research on human-computer interaction — found that individuals with high task-specific self-confidence engage in more critical thinking when using generative AI, even though they perceive the effort involved as significant (12). Those with low confidence in their own domain, or high uncritical confidence in the AI, tend to skip the hard work of verification.
The implications for talent identification are direct: someone who has always gravitated toward difficult puzzles, dense texts, and complicated arguments will be more likely to wrestle productively with AI — pushing back on outputs, probing for gaps, and integrating results with their own judgment. The cognitive miser, who finds hard thinking unpleasant and takes any opportunity to avoid it, will tend to accept AI outputs at face value. In a world where AI outputs can be plausible and wrong simultaneously, that is a serious professional liability.
2. Openness to Experience
Of the five major personality dimensions identified in psychological research — openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism — openness is the most consistently robust predictor of how well a person adopts and innovates with technology. A 2025 peer-reviewed study published in Behavioral Sciences found that AI usage is positively associated with employee innovation behavior, and that openness to experience significantly moderates this relationship (11). People high in openness show heightened curiosity, divergent thinking, and receptiveness to unconventional problem-solving approaches.
Practically, this means that standard personality assessments already used in many recruiting pipelines can serve double duty as AI-readiness screens. The study’s authors suggest that organizations “introduce personality assessment tools during recruitment to screen candidates who are receptive to new things and have high cognitive flexibility” and that high-openness employees be positioned to lead teams in exploring creative uses of AI (11). Openness is not the same as enthusiasm for technology; it is the underlying disposition that allows enthusiasm to translate into actual performance.
3. Domain Self-Confidence
There is a common misconception that AI-productive workers are those most impressed by AI’s capabilities. The research suggests the opposite. The CHI 2025 study by Lee and colleagues at Microsoft Research found that high confidence in generative AI — untempered by confidence in one’s own expertise — is associated with less critical thinking, not more (12). The worker who assumes the AI knows better than she does is less likely to catch the errors, hallucinations, and logical gaps that make AI outputs genuinely risky in high-stakes contexts.
Domain self-confidence — knowing enough about your field to recognize when an AI output is off — is therefore a foundational asset. An attorney reviewing a brief, a data analyst checking a model’s assumptions, or a journalist vetting a summary all need enough independent expertise to audit what the AI has produced. The CHI study found that workers with this kind of grounded self-confidence are more likely to verify AI claims against external sources and their own knowledge, to refine prompts based on what they already know, and to maintain what the researchers called “task stewardship” — a sense of ownership over the final product (12).
4. Verification Discipline
Related to domain self-confidence, but worth distinguishing: verification discipline is a work habit rather than a personality trait. It is the practice of treating AI outputs as drafts that require fact-checking, cross-referencing, and independent judgment before use. Sylvie Sturm, a reporter and producer at the San Francisco Public Press, illustrates this habit precisely. Writing for Muck Rack’s 2026 State of Journalism survey, Sturm described using AI transcription and synthesis tools to process complex documents — public budgets, court cases, public health datasets — but insisted on listening to original recordings before citing any quotes and cross-referencing all calculations against official reports, data, and expert sources (14).
That combination — use AI heavily for speed, but own the verification — is the signature of this trait. Workers who lack it tend to introduce errors at a rate that eventually exceeds the productivity gain AI provided. For HR administrators and hiring managers, verification discipline is worth probing directly in interviews: does the candidate have a principled method for checking AI-assisted work, or do they treat the output as finished?
5. Willingness to Experiment First, Systematize Later
Ethan Mollick, associate professor at the Wharton School and author of Co-Intelligence: Living and Working with AI, has argued consistently that the workers who gain the most from AI are those who approach it empirically — who try things, notice what works, and share findings — rather than waiting for official guidance or a proven workflow to be handed to them (13). Mollick himself is an example: his curiosity about ChatGPT when it appeared in late 2022 drove months of hands-on experimentation that eventually produced one of the more widely read books on the subject.
This trait is measurable at the organizational level. Microsoft’s 2025 Work Trend Index, drawn from a survey of 31,000 knowledge workers across 31 countries, found that workers at “Frontier Firms” — those with the most sophisticated AI integration — were significantly more likely than their peers to say they brainstorm AI opportunities as a team, share tips, learnings, and mistakes, and discuss quality standards for AI-assisted work (6). Eighty percent of Frontier Professionals reported producing work they could not have produced a year earlier, compared to the broader population of AI users (6). The experimentation habit, in other words, compounds.
6. Voluntary Scope Expansion
The Haas finding that workers broadened their definitions of their own jobs — rather than narrowing their effort — when AI arrived points to a trait that is easy to undervalue: the willingness to take on more when capability increases (4). This is not simply ambition; it is a particular professional orientation in which one’s sense of what belongs in one’s domain expands in proportion to what one can realistically accomplish.
Ye and Ranganathan’s study identified a pattern in which workers, finding that AI could handle tasks they had previously outsourced — simple coding, first-draft writing, data cleaning — brought those tasks back in-house and executed them alongside their primary work (4). The result was more intense work, not less. For HR administrators, the trait to watch for is the person who, when handed a new tool, asks “what else can I do now?” rather than “what can I stop doing?” The latter orientation produces short-term efficiency; the former produces long-term professional growth.
7. Adaptability as a Daily Practice
The Educational Testing Service’s 2026 Human Progress Report, which surveyed 32,558 adults across 18 countries, found that 77 percent of workers now believe job security requires continuous evolution, and 61 percent have shifted their focus from stability to staying relevant (10). But the report also identified what it called an “adaptability paradox”: 71 percent of workers cannot envision the future jobs they are preparing themselves for, even as they actively build new skills.
This suggests that adaptability cannot be treated as a one-time response to disruption; it is a disposition toward ongoing learning that precedes any particular crisis. The World Economic Forum’s Future of Jobs Report 2025 ranked curiosity and lifelong learning among the top core skills demanded by employers through 2030, alongside analytical thinking and creative thinking (5). INSEAD’s research on what it calls “resilience thinking” frames the same quality as the capacity to “bounce forward” rather than merely recover — learning from disruption rather than simply surviving it (16). Workers who do this instinctively are better positioned for AI-assisted environments than those who learn only in response to crises.
8. Collaborative AI Literacy
The data from Microsoft’s 2025 Work Trend Index includes a finding that is easy to overlook: at the organizations where AI integration is most effective, individuals share what they know (6). Frontier Professionals are twice as likely as their peers to say that experimenting with AI is rewarded regardless of outcome, and significantly more likely to document agent workflows and quality standards at the team level (6). The learning is collective, not merely individual.
Boston Consulting Group’s AI at Work 2025 report, which surveyed more than 10,600 employees across 11 nations, found that 79 percent of those who received more than five hours of AI training were regular AI users, compared to 67 percent of those with less training — and that frontline workers are most underserved in this respect (9). Accenture’s Talent Reinventors research found that the roughly 18 percent of organizations it classified as exemplary grew revenue 1.8 percentage points faster and profits 1.4 points faster than peers while also strengthening workforce adaptability (15). What set those organizations apart was not merely AI deployment but a culture of co-learning — dynamic, continuous collaboration between people and AI that became embedded in daily workflow (15).
The implication for individual assessment: a worker who actively teaches colleagues what they are learning about AI, who writes up their workflows, who asks questions and shares failures, is not merely being collegial. They are accelerating their own learning and the organization’s.
9. Social and Emotional Intelligence
It might seem paradoxical to include social intelligence in a profile of AI-productive workers, since the workers who struggle with AI are not typically defined by social strengths. But McKinsey’s research makes the case plainly: as AI absorbs more routine cognitive tasks, the premium on interpersonal judgment rises in proportion (8). McKinsey’s “Agents, Robots, and Us” report projected that demand for social and emotional skills could rise 11 percent in Europe and 14 percent in the United States through 2030, even as AI automates a growing share of structured work (7).
The reason is partly substitution and partly complementarity. AI is already capable of drafting, summarizing, coding, and analyzing; it remains limited in conflict resolution, empathy-grounded negotiation, team coaching, and reading a room. Workers who are genuinely good at those things become more valuable as AI handles more of the analytical load around them. McKinsey’s 2026 human–AI workforce analysis found that the hybrid roles projected to grow most sharply are those in which humans manage relationships and judgment while AI agents handle structured tasks (7). Social and emotional intelligence is not a consolation prize in the AI age; it is an amplifier.
10. Autonomous Initiative
Much of the organizational conversation about AI adoption has focused on top-down deployment: which tools to buy, which workflows to redesign, which training programs to build. But the ActivTrak data tells a different story at the individual level: the workers who benefited most from AI were those who adopted it without waiting to be told (2,3). Their weekends grew busier, their collaboration rates climbed, their multitasking increased — not because their managers pushed them, but because they saw possibility and moved toward it.
BCG’s survey found that just 25 percent of frontline workers say their leaders provide sufficient guidance on AI (9). Workers who wait for that guidance are in a long queue. The ones who open a tool, run an experiment, produce something useful, and then share what they learned are not only ahead professionally — they are the informal faculty members who end up driving adoption for everyone around them. For HR administrators, autonomous initiative is not the same as impulsiveness; it is the combination of self-direction and judgment that characterizes the learner who does not need permission to grow.
What This Means for HR Administrators
These ten traits divide roughly into three categories. The first — need for cognition, openness to experience, and domain self-confidence — are primarily dispositional, meaning they are relatively stable characteristics a person brings to a job. Standard behavioral interviews, personality assessments, and work-sample tests that probe for intellectual curiosity and domain expertise can surface them reliably.
The second category — verification discipline, voluntary scope expansion, and collaborative AI literacy — are behavioral habits that can be observed in work samples, reference checks, and structured interviews. Asking a candidate how they handle an AI-generated answer they are not sure about, or what they last taught a colleague about a tool they use, will reveal whether these habits exist.
The third — willingness to experiment, adaptability as practice, social intelligence, and autonomous initiative — sit between disposition and habit, and are best assessed through past behavior. The most productive questions here are retrospective: describe a time you adopted a new tool before your organization had a policy for it; describe a situation where interpersonal skills determined an outcome that technical knowledge alone could not have produced.
Not every AI-productive worker will score high on all ten. But an organization that systematically identifies and recruits people high on most of them is building a workforce that will do more with AI, more reliably, than one that hires for technical credentials and hopes the right dispositions follow. The studies from ActivTrak and Berkeley Haas suggest that AI is already making work more intense and more demanding — and that the workers who rise to meet that are not necessarily those with the most technical training, but those with the right orientation toward challenge, collaboration, and growth.
References
1. David Brooks, “The People Who Will Thrive in the AI Age,” The Atlantic, June 28, 2026.
2. ActivTrak Productivity Lab, “2026 State of the Workplace.”
5. World Economic Forum, Future of Jobs Report 2025, January 2025.
6. Microsoft, “2025 Work Trend Index Annual Report: The Year the Frontier Firm Is Born,” April 2025.
7. McKinsey Global Institute, “The Rise of the Human–AI Workforce,” 2026.
8. McKinsey.org, “The Human Skills You’ll Need to Thrive in 2026’s AI-Driven Workplace.”
9. Boston Consulting Group, “AI at Work 2025: Momentum Builds, but Gaps Remain,” June 2025.
13. Ethan Mollick, Co-Intelligence: Living and Working with AI (New York: Penguin Random House, 2024).
14. Muck Rack, “3 Writers Weigh In on AI and Journalism in 2026.”
15. Accenture, “Talent Reinventors: Delivering Value with and for People in the Age of AI,” 2025.
16. INSEAD Knowledge, “How Talent Can Thrive in an AI-Driven World.”
###
Filed under: Uncategorized |
Educational Technology and Change Journal 
Leave a comment