From Expertise to Judgment: The Career Shift Adam Grant Sees Coming
For most of the twentieth century and well into the twenty-first, the path to professional success followed a recognizable script: master a body of knowledge, become indispensable through specialized expertise, and let that accumulated intellectual capital compound over a career. Lawyers memorized case law. Financial analysts internalized market mechanics. Engineers held proprietary technical knowledge in their heads. The expert was, by definition, the most valuable person in the room.
Adam Grant, organizational psychologist at the Wharton School of the University of Pennsylvania and one of the most-read management thinkers of his generation, is arguing that this script has been torn up. In a September 2026 interview with The Wall Street Journal, Grant made a deceptively simple but structurally significant claim: careers that once hinged on ability now depend on agility. Human judgment, he contends, is poised to supersede human expertise as the defining professional currency. “The currency of success is shifting under our feet,” he told reporter Chip Cutter.
This is not a minor refinement to existing career wisdom. It is a fundamental reordering of what organizations should reward, what professionals should cultivate, and what business leaders should prioritize as they build teams and shape cultures. Grant’s argument deserves serious scrutiny, not merely enthusiastic agreement, because its implications are both far-reaching and, in several respects, more complicated than the interview format allows.
From Expertise to Agility: The Central Claim
Grant’s core argument rests on a straightforward observation about information economics. Before AI saturated the professional environment, information itself was scarce. Possessing specialized knowledge created a genuine moat. The doctor who had read the literature, the attorney who had studied the precedents, the consultant who had analyzed comparable industries—each held something that was genuinely difficult to replicate quickly. That scarcity conferred value.
Artificial intelligence has largely destroyed that scarcity. Large language models can now synthesize regulatory frameworks, summarize scientific literature, draft legal briefs, generate financial models, and produce market analyses in seconds. What took a junior analyst a week now takes an AI tool minutes. In this environment, Grant argues, simply knowing things is no longer the differentiator it once was. The differentiating capabilities become pattern recognition, critical synthesis, credibility assessment, and the judgment to act wisely on imperfect information.
This analysis is largely correct, and it is supported by emerging workforce data. A 2023 study published in Science examining the impact of AI tools on professional productivity found that workers who used AI assistance saw productivity gains of roughly 14% on average, but the distribution was highly uneven. Workers with weaker baseline skills gained the most, while the highest performers gained comparatively less—suggesting that AI is compressing the knowledge gap between experts and novices in measurable ways (Brynjolfsson, Li, and Raymond, 2023). If AI equalizes knowledge access across skill levels, then knowledge possession alone becomes a weaker signal of professional value.
The World Economic Forum’s Future of Jobs Report 2025 reinforces this trajectory, identifying analytical thinking, creative thinking, and resilience as the top skills employers expect to grow in importance over the coming five years, while knowledge-intensive but routinizable skills face the greatest displacement pressure. Judgment, synthesis, and adaptability are not peripheral competencies in this framing. They are the new core.
The Threat Rigidity Problem
Grant introduces a concept from organizational psychology that is particularly worth sitting with: threat rigidity. Under conditions of perceived threat, individuals and organizations tend to narrow their behavioral repertoire, doubling down on what has worked before rather than experimenting with new approaches. It is a well-documented phenomenon in the psychology literature, first formally described by Staw, Sandelands, and Dutton in their landmark 1981 paper in Administrative Science Quarterly.
The relevance to the current moment is direct. As AI reshapes professional roles, many workers are responding by trying to identify the one skill set that AI cannot yet touch, then fortifying that position. Grant’s critique of this response is sharp and worth amplifying: the defensive crouch of threat rigidity is precisely the wrong posture when the landscape itself is shifting rapidly. What feels impenetrable today may be automated in six months. The search for a permanent AI-proof skill set is, in many respects, a search for a static solution to a moving problem.
But here is where Grant’s argument requires some pushback. The call to embrace agility—to run weekly behavioral experiments and rotate meeting leadership—is genuinely useful tactical advice. However, it risks underweighting the ongoing importance of deep domain expertise in certain contexts. Not all professional value is being commoditized at the same rate.
Consider surgery. AI diagnostic tools are now reading imaging scans with accuracy that rivals experienced radiologists in controlled settings. Yet the actual performance of complex surgical procedures still requires a depth of physical skill, situational judgment, and experiential learning that has not been replicated by machines. Consider also the work of elite litigators. AI tools can draft briefs and surface precedents, but the ability to read a judge’s temperament in real time, calibrate an argument to a specific jury’s likely concerns, and make split-second strategic decisions in oral argument remains stubbornly human. Deep expertise and judgment are not always substitutes for one another. In many of the highest-stakes professional contexts, they remain complements.
This is not a trivial distinction. Leaders who respond to Grant’s thesis by de-emphasizing the development of deep domain knowledge in favor of generalist agility may find themselves building organizations populated by people who are flexible but shallow—well-adapted but ultimately lacking the substantive grounding that genuine judgment requires. Judgment, after all, is not formed in a vacuum. It grows from experience, and experience is typically domain-specific.
Creativity Reconceived: Selection Over Generation
One of the more striking claims Grant makes in the interview concerns the nature of human creativity. He argues that creativity is shifting from idea generation toward idea selection and elaboration. AI can produce a hundred variations on a theme in the time it takes a human to draft one. The human advantage lies not in originating ideas but in recognizing which ideas are worth pursuing, developing them with emotional resonance, and synthesizing them into frameworks that stick.
This reframing has significant implications for how organizations structure creative work. Teams that spent the past decade rewarding brainstorming prolificacy (the person who generates the most ideas in a meeting) may need to start rewarding curatorial judgment: the person who most reliably identifies which of the many ideas generated by humans and machines alike is worth developing. These are genuinely different cognitive skills, and they tend to be distributed differently across individuals.
Research on creative cognition supports Grant’s intuition here. Psychologist Dean Keith Simonton’s work on creative productivity has long established that the quantity of creative output is a poor predictor of quality, and that what distinguishes highly creative individuals is often not raw ideation but a refined capacity to evaluate and develop promising work (Simonton, 2003). If AI handles the generation problem, human creative value concentrates in the evaluation and refinement stages. Organizations that recognize this early will have a structural advantage in deploying both human and machine creative capacity effectively.
Amazon’s product development process offers a partial real-world illustration. The company’s practice of writing the press release and FAQ document before beginning product development is essentially an exercise in curatorial judgment, forcing teams to evaluate whether an idea is worth building before investing resources in building it. As AI tools generate more product concepts, feature variations, and design options than teams can possibly evaluate, the discipline of curatorial judgment becomes more organizationally valuable, not less.
Behavioral Experiments as Professional Practice
Grant’s practical prescription—running weekly behavioral experiments rather than committing rigidly to established routines—is psychologically well-founded and more actionable than most career advice. The scientific thinking behind it draws on the same logic as iterative product development: form a hypothesis, test it at low cost, learn from the result, update your model. Applied to professional behavior, this means treating communication methods, meeting structures, collaboration patterns, and even relationship-building strategies as variables worth testing rather than fixed protocols.
Grant’s specific example of replacing email with spontaneous phone calls as a relationship-building experiment is both modest and instructive. It reflects a broader truth about professional relationships that research consistently confirms: richer communication channels build stronger relational bonds than leaner ones. A 2021 study by Amit Kumar and Nicholas Epley found that people systematically underestimate how much others appreciate unexpected phone calls compared to text-based communication, and that voice contact produces significantly warmer interactions than text equivalents (Kumar and Epley, 2021, Journal of Experimental Psychology). Grant’s instinct to rehabilitate the phone call is not nostalgia. It is evidence-based relationship strategy.
The weekly experiment framework also serves a psychological function beyond its direct practical benefits. It reframes professional development from a destination (acquiring a fixed skill set) into a continuous process. That reframing matters because it is more psychologically sustainable under conditions of rapid change. Professionals who define their competence by what they already know are vulnerable to anxiety and rigidity as that knowledge depreciates. Professionals who define their competence by their ability to learn and adapt are structurally better equipped for a period in which the half-life of specific skills is shortening.
The Organizational Imperative
Grant’s interview is framed primarily around individual career strategy, but its implications for organizational leadership are at least as significant. Leaders who accept the judgment-over-expertise thesis must ask themselves whether their organizations are actually structured to reward agility—or whether their promotion criteria, performance review systems, and talent acquisition processes still implicitly prioritize expertise accumulation.
The honest answer, in most large organizations, is that the structural incentives have not caught up with the emerging reality. Performance management systems in most enterprises still reward quantifiable expertise delivery—billable hours in professional services, lines of code in technology, cases closed in legal work. The harder-to-measure capacities Grant is pointing toward (quality of judgment, speed of learning, effectiveness of synthesis) rarely appear on a performance scorecard with the same clarity.
Microsoft’s experience restructuring its performance management system offers an instructive case. Under Satya Nadella, the company moved away from a stack-ranking system that had rewarded individual expertise and competition, and toward a model that explicitly valued learning, collaboration, and what Nadella called a growth mindset, drawing directly on Carol Dweck’s research on how beliefs about the malleability of ability shape behavior. The cultural and financial results of that transition have been extensively documented. Microsoft’s market capitalization grew from roughly $300 billion when Nadella took over in 2014 to over $3 trillion by the mid-2020s. While many factors contributed to that trajectory, the deliberate cultural emphasis on learning agility over static expertise was a foundational element of the transformation.
For leaders reading Grant’s argument, the question is not whether the shift from expertise to judgment is real. The evidence suggests it is. The question is whether the organizational systems they control are rewarding the capacities that will matter most going forward—or whether institutional inertia is preserving a reward structure designed for a professional environment that is already receding.
What Grant Gets Right and What Remains Incomplete
Grant’s thesis is timely, research-grounded, and directionally correct. The depreciation of stored knowledge as a primary professional asset is real and accelerating. The premium placed on judgment, synthesis, and learning agility is rising. The behavioral experiments he recommends are genuinely useful and appropriately humble in their scope.
What the interview leaves underexplored is the transition challenge for the millions of professionals who built careers precisely on the expertise model Grant describes as obsolescent. Telling a fifty-year-old subject matter expert that agility now matters more than expertise is accurate advice delivered into a context where agility is genuinely harder to develop than it is for a twenty-five-year-old who has spent less time building neural pathways around a specific domain. The organizational psychology literature on skill transfer suggests that deep expertise in one domain can actually impede flexible thinking in adjacent areas, a phenomenon sometimes called the curse of expertise.
Leaders, therefore, face a dual obligation. They must build organizational cultures that cultivate judgment and agility in their talent pipelines. They must also design transition pathways that allow experienced professionals to redirect their domain knowledge toward curatorial, mentoring, and synthesis roles where it remains genuinely valuable rather than obsolete. Grant is right that the currency of success is shifting. The leadership challenge is managing that currency exchange without writing off the accumulated human capital that organizations have spent decades developing.
The professionals and organizations that will navigate this transition most effectively are those that treat Grant’s provocation not as a reason to abandon expertise but as a signal to reposition it. Deep knowledge, filtered through sharp judgment and expressed with human resonance, is not obsolete. It is precisely what AI cannot yet replicate at the highest level. The task is not to stop knowing things. It is to ensure that knowing things remains the foundation for judgment rather than a substitute for it.