We often feel like imposters because we’re chasing an idealized version of competence that exists only in our imagination. We construct this elaborate image of what a truly capable person would look like in our position, how they’d speak, what they’d know, how effortlessly they’d navigate challenges.
Now, artificial intelligence is rewriting this relationship. It’s altering how we perceive our own competence and deepening the imposter syndrome that many of us already carry.
A 2024 Korn Ferry survey of 10,000 workers found that 71% of US CEOs experience symptoms of imposter syndrome in their role, significantly more than early-stage professionals at 33%. What makes this striking is that 85% of these same CEOs express that they are totally competent in the role. There’s a painful disconnect between what leaders know they can do and how legitimate they feel doing it. The arrival of AI seems to have widened that gap.
A Different Kind of Doubt
The imposter syndrome that emerges in the AI era has a distinct character. It’s not the classic fear of being exposed as unqualified. Instead, it’s a creeping uncertainty about whether human judgment, creativity, and expertise matter when machines can approximate these qualities at scale. AI shifts intelligence from effort to access, creating doubt about authorship and originality.
What makes this particularly disorienting is that AI doesn’t replace us cleanly. A gynecologist diagnoses if you have PCOD or PCOS and gives you a cure plan, but the AI flagged you three possibilities first. A programmer still writes the code, but ChatGPT generated the initial framework. A teacher still designs the lesson, but the AI already created a version that’s “pretty good.” We’re left in this uncanny middle ground, neither fully replaced nor fully validated, questioning whether our contributions are essential or merely decorative.
Ethan Mollick, a Wharton School professor who studies AI’s impact on work, captured this strange moment: “The AI is better than a lot of people in a lot of jobs, but not at their whole job. And so whatever you’re best at, you’re almost certainly better than the AI is.” That “but not at their whole job” is where we’re all living now, in the uncomfortable space between what the machine can do and what it cannot.
The Productivity Paradox
Research from the Imagining the Digital Future Center found that nearly half of AI users believe their AI assistant is smarter than they are. Meanwhile, a significant percentage report concerns about dependence on AI for thinking and decision-making. More than half of workers are concerned that their role will be replaced by AI in the next three years.
There’s a particular irony that many people now use AI to help them work through their imposter syndrome by asking AI chat tools for affirmations or career advice, outsourcing even their self-doubt to the very technology that triggered it. Use AI and wonder if your work is authentically yours. Don’t use AI and wonder if you’re being willfully obsolete. Technology has managed to make us feel fraudulent coming and going.
The disconnect is striking. In Mollick’s research with the Boston Consulting Group, consultants who used GPT-4 showed a 40 percent improvement in quality, were 25 percent faster, and completed 12.5 percent more work without any optimization. People are objectively more effective and subjectively more doubtful. When work becomes easier, we paradoxically trust it less. We’ve internalized the belief that value requires suffering, that expertise should be hard-won. AI short-circuits this equation.
What Remains Human
But perhaps the most human response to AI-induced imposter syndrome comes from focusing on what remains irreducibly ours. The judgment about which AI output is actually useful. The wisdom to know when the efficient answer isn’t the right answer. The ability to sit with ambiguity, to change course, to care about something beyond optimization.
Mollick puts it simply: “You don’t know what AI is good for or bad for inside your job or your industry. Nobody knows. I think a lot of people think there’s a secret instruction manual out there. There is not.” That uncertainty and that need to figure things out yourself is precisely where human expertise lives now.
The value isn’t in what AI produces. It’s in what you do with what it produces. And knowing what to do with it requires exactly the kind of judgment that comes from years of making mistakes. That judgment is invisible in the final output. No one sees the ten AI-generated options you rejected, the subtle ways you refined the prompt, the domain knowledge that let you spot the plausible-sounding error. The technology makes expertise look effortless, which makes it feel worthless.
Living in the Uncertainty
Mollick describes two ways of working with AI: what he calls “centaurs and cyborgs.” Centaurs divide up their work: you do the coding, AI does the writing. The more effective way is cyborg, where you’re integrating your work with the AI interactively. In the cyborg model, you’re throwing ideas to the AI, working with it, and using your expertise to guide and validate the output.
The imposter syndrome of the AI era is real and pervasive. It’s making people question expertise they’ve legitimately earned. But it’s also forcing a necessary reckoning with what expertise actually is. Not the performance of difficulty. Not the hours logged. But the judgment, the care, the human-scale understanding that no amount of training data can replicate.
Adam Grant, organizational psychologist at Wharton, offers a reframe worth considering. “Imposter syndrome is not a clue that you’re unqualified,” he writes. “It’s a sign of hidden potential. When you think others are overestimating you, it’s more likely that you’re underestimating yourself.” In the context of AI, maybe our discomfort isn’t a sign we’re becoming obsolete. Maybe it’s evidence we’re adapting to something genuinely new and uncertain, which requires exactly the kind of growth that feels uncomfortable.
Arvind Narayanan, a Princeton computer science professor and author of AI Snake Oil, cuts through the hype from a different angle. He views AI as what he calls “normal technology” – not a separate species or existential force, but a tool like electricity or the internet. “To view AI as normal is not to understate its impact,” he and co-author Sayash Kapoor write, “but it is in contrast to both utopian and dystopian visions of the future of AI which have a common tendency to treat it akin to a separate species.” It’s never human versus AI, Narayanan argues. It’s always human-plus-AI versus AI – and the collaboration, not competition, is what shapes real outcomes.
Mollick’s advice seems practical in this case: “You need to push through, like there is a point of expertise with this, where you start to get what it does and what it doesn’t. Ten hours is my loose rule of thumb for how much time you have to spend using these systems to kind of get it.”
We’re not imposters. We’re just learning to value ourselves by different measures. That learning curve is uncomfortable. But discomfort, as it turns out, is one of the most human things there is. And in an age where AI can generate perfect-sounding answers instantly, maybe our willingness to sit with uncertainty, to question ourselves, to refine and iterate, maybe that’s not a weakness. Maybe it’s precisely what makes us irreplaceable.
Leave a Reply