Imposter Syndrome 2.0: The Camera Didn’t Kill the Painter

By Dr. Matt McKenna — The Doctor of Digital Transformation

Something quiet is happening in your organization right now. It doesn’t show up on an engagement survey, which is precisely why nobody is doing anything about it. It looks like a senior analyst deleting a deck she spent four hours building, because a model produced something 85% as good in about ninety seconds. It looks like a mid-career leader privately wondering whether the thing he was known for is still a thing.

That feeling has a name, Impostor Syndrome 2.0.

The feeling is rational — that’s what makes it different

Classic imposter syndrome is a psychological misfire. You doubt yourself despite the evidence that you belong. Imposter Syndrome 2.0 is different, and this is the part most leaders miss: it is a correct reading of a real change.

The data backs that up. An Eastern Washington University survey of 1,000 employees found that daily AI users were the most likely to report regularly experiencing imposter syndrome, at 30% — higher than occasional users, higher than non-users Eastern Washington University. A Korn Ferry survey of more than 15,000 professionals found that 43% of senior executives struggle with imposter syndrome — not the juniors. The veterans Forbes.

As Nirit Cohen puts it in Forbes, these professionals “are not only asking whether they are good at their jobs. They are questioning whether the way they have always worked still counts” Forbes.

I want to state my position plainly, because it runs against most of what lands in your news-feed: this discomfort is not a warning sign. It is a growth signal. And let’s be honest — we’ve all spent the last few years side-eyeing our laptops, waiting for them to finally ask us to hand over our security badges with the constant corporate layoffs.

The machine is faster than you, and so was every machine before it.

Let’s look at the facts:

1839. The French painter Paul Delaroche saw a photograph and reportedly declared, “From today, painting is dead!” He was wrong about the death and completely right about the disruption. The camera destroyed the commercial market for exacting likenesses, and painters responded by moving toward what the lens structurally could not do. As the Museum of Arts & Sciences notes, “with the demand for exacting likenesses in portraits, landscapes, and other scenes diminished, painters turned to new expressions that seemed to revel in something unique to their art — the brushstroke” MOAS. Impressionism. Then abstraction. Then a century of work no camera could have produced, because it was never in front of the camera. Turns out you can’t photograph a feeling. Several very wealthy patrons found this out the expensive way.

1450. Gutenberg didn’t erase the scribe. Book prices fell roughly 90% by the end of the 16th century, and hand-copying — commercially dead as a commodity — was reclassified as a craft collectors actively preferred History of Printing. So the scribes didn’t lose their jobs; they lost their market segment and gained a snobbery premium. Honestly, a decent career pivot.

1950s. At NASA’s Jet Propulsion Laboratory, the people called “computers” were human beings — mostly women with mathematics degrees — calculating spacecraft trajectories by hand. Then machines showed up that could do arithmetic faster. “Over the course of time,” NASA records, “these women not only performed hundreds of thousands of mathematical calculations crucial to the U.S. space program, but also eventually became some of the first computer programmers at NASA” NASA JPL. Note the sequence. The machine didn’t fire them. It got promoted, and they got promoted with it.

1967-1969 The ATM. Everyone confidently predicted the end of the teller. Tellers per urban branch dropped from 20 to 13 between 1988 and 2004 — and because branches became cheaper to run, urban branches increased by 43%. Teller employment since 2000 has actually grown slightly faster than the labor force IMF. The job shifted from counting cash to building relationships, and the pay went up. So the machine freed the teller from the most tedious part of her day and gave her the interesting part instead. That is genuinely the best deal in this entire article.

1979. VisiCalc made bookkeeping arithmetic instant. The US ended up with 400,000 fewer accounting clerks — and 600,000 more accountants. As the BBC summarized, the repetitive, routine parts of accountancy disappeared and what remained required more judgment and more human skill BBC. Nobody has ever asked me to explain compound interest at a dinner party. Somebody has definitely asked me to explain it in a meeting, which is the same job with worse lighting.

1997. Deep Blue beat Kasparov and human chess was supposedly finished forever. Instead, active professional players nearly doubled between 2009 and 2014 and open tournaments grew 37% worldwide Quartz. Kasparov invented Advanced Chess — human plus engine — and found the pairing beat the engine alone Advanced chess. The machines got unbeatable at chess and somehow more people signed up to play. You cannot buy marketing like that.

Every one of these people was told they were obsolete. Every one of them was standing in the doorway of a bigger room — and all they had to do was stop guarding the door.

The camera records light. It cannot record a vision.

Here’s where I want to be precise about the technology, because the precision is the entire point — and it’s the most optimistic message I can hand your workforce.

A camera records light. Photons that physically arrived — bounced off a face, bent through a lens — strike a sensor and leave a trace. That’s the whole mechanism, and it carries a hard physical limit: what never stood in front of the lens cannot appear in the frame.

A large language model is a next-token predictor. Trained on a vast corpus of human text, it forecasts how a sequence continues — fluently, convincingly, and always from the corpus, inheriting both the insight and the bias of what we already produced Large language model. It has read everything and experienced nothing. It is the world’s most well-read intern.

That isn’t an insult. It’s a specification. And it maps onto the camera almost exactly.

When AI runs your research, your analysis, your structuring and your first draft, it hands you back the industry’s composite. That’s genuinely useful — the composite is where you start, not where you finish. Your job was never to produce the composite. Your job has always been to produce the specific — the thing that isn’t in the training data yet, because it didn’t exist until you thought of it.

Where the AET Framework lands on this

This is the part I care most about, because the entire premise of the AI-Enabled FSI Transformation (AET) Framework is that AI is not a technology project. It is an enterprise-wide operating model shift AET Framework.

The framework exists because of a brutal ratio: 88% of organizations are wasting billions on AI hype, while the 12% “Vanguard” are achieving real revenue growth and cost reduction. Most institutions are stuck in pilot purgatory — isolated labs and departmental tools disconnected from enterprise strategy. “Pilot purgatory” is my clinical term for the place where good ideas go to be politely ignored forever.

Four of the ten dimensions speak directly to Impostor Syndrome 2.0.

Dimension 3 — Culture: from change tolerance to algorithmic trust. This is the 14% problem. Only 14% of workers use generative AI daily, even as 73% of consumers worldwide say they trust content created by generative AI AET Culture. So we trust AI to write things at work roughly five times more than we trust ourselves to use it — which tells you the problem was never the technology. High-performing AET organizations are 3x more likely to redesign their workflows around human–AI teaming rather than handing out tools and hoping. The proof cases are concrete:

• Morgan Stanley reached 98% adoption among Wealth Management Advisors against a 14% industry average — earned with practical, productivity-boosting use cases, not mandates. Nobody was ordered to like it. They just noticed the ones using it were home by six.
• DBS Bank built the P-U-R-E framework — Purpose-driven, Unbiased, Responsible, Explainable — giving employees ethical guardrails so they could adopt AI confidently.
• Wells Fargo is driving a large-scale AI literacy push, with Saul Van Beurden framing reskilling as a “mutual responsibility.”

Dimension 1 — Leadership: from sponsorship to orchestration. 78% of organizations now use AI in at least one business function, but only 25% of banks are truly ready for the AI era. 95% of corporate AI projects fail to create measurable value. CEOs spend 47% of their time on issues with a horizon under one year and only 16% on strategy beyond five years AET Leadership. That’s the disease. The treatment: anchor AI in business strategy, rewire whole domains rather than isolated use cases, and install centralized governance such as an AI Control Tower. Goldman Sachs’s “One Goldman Sachs 3.0” is the model — AI as a core strategic pillar, coordinated across business units rather than enjoyed independently by nine departments who, charmingly, don’t talk to each other.

And here’s the leadership point nobody enjoys hearing. If leadership doesn’t explicitly define what human contribution now matters, employees will invent their own answer. It will be speed, volume and perfection — the three fastest routes to burnout, and a curriculum nobody designed on purpose. Swap the ambiguity for a sentence.

Dimension 10 — Workforce: human-in-the-loop. The most effective transformation doesn’t remove people; it elevates them. Skills shift toward prompt design, AI oversight and high-level judgment — what I call cognitive upskilling. And replacement was never on the table anyway: in 2022 nearly two million Americans reached retirement age against roughly 40,000 entering prime working age, and US workforce participation is projected to fall to 61% by 2033 AET Workforce. We don’t just want augmentation. We need it. The robots are not taking your job. They’re becoming your intern — and unlike the last intern, they don’t need a desk, and they don’t schedule a standing meeting to ask you what “context” means.

Dimensions 8 and 7 — the ones that decide whether you manufacture this problem or dissolve it. Organization moves to Agile AI Cells: distributed authority to experiment, centralized accountability for ethics and security. Financial Leadership moves to Cognitive ROI: capacity creation, cycle-time compression, operating leverage. If your internal success metric is “how many people can we remove,” you have manufactured Imposter Syndrome 2.0 and you should expect the bill. If your metric is “how much judgment did we unlock,” you have dissolved it.

Cognitive offloading is a door, not a tragedy

I want to be honest about the research, because the honest version happens to be the optimistic one.

The warnings are real. Gerlich’s 2025 study of 666 participants found a strong positive correlation between AI tool use and cognitive offloading (r = +0.72) and a strong negative correlation between AI tool use and critical thinking (r = −0.68) MDPI. MIT Media Lab’s Your Brain on ChatGPT found that across 54 participants, brain connectivity “systematically scaled down with the amount of external support” — a cumulative effect the authors call cognitive debt arXiv. They can measure the mortgage you take out against your own thinking. Which is a genuinely uncomfortable sentence to read aloud in a boardroom, so I’d suggest reading it silently.

But the positive framing of the same mechanism is equally well established: cognitive offloading meaningfully improves problem-solving and learning efficiency because it frees finite mental resources for higher-order processing Nature.

Both are true. The difference is entirely what you do with the capacity you get back. The spreadsheet didn’t dull accountants; it freed them to model scenarios. The ATM didn’t end tellers; it freed them to build relationships. Offload the arithmetic. Never offload the judgment. That’s the entire rule, and it fits on a Post-it.

One caveat, and it’s the most important line in this piece. Doshi and Hauser found that generative AI enhances individual creativity while reducing the collective diversity of novel content Science Advances. Read that carefully: more productive individuals, less varied portfolio. The machine is a convergence engine. If everyone offloads the same thinking to the same models, your organization quietly becomes a well-formatted version of everybody else. Your job is to be the divergence.

And a final number that reframes the whole problem. Only 8% of employees have received formal AI training. 57% say AI tools weren’t part of onboarding at all. Just 34% believe their company provides sufficient AI training EWU. People are not failing because they can’t learn. They’re failing because nobody taught them. We bought a Ferrari for the entire workforce and skipped the driving lessons. That’s a leadership gap, not an employee deficiency — and it’s the most fixable item on this list.

The five things I would actually do as a leader

1. Say the quiet part out loud. In your next all-hands, name what only humans now do: judgment, ethics, client trust, exception handling, the call that isn’t in the training data. Ambiguity fuels Imposter Syndrome 2.0. Replace it with a sentence. One sentence. You have given longer speeches about the parking situation.
2. Make learning public. Reward the person who shows the prompt that failed, not just the prompt that worked. Cohen’s guidance — normalize uneven adoption — is the whole game.
3. Steal shamelessly. Morgan Stanley’s practical use cases, DBS’s guardrails, Wells Fargo’s mutual responsibility. None of it is secret. Strategy has never been about having the idea; it’s about the follow-through.
4. Close the training gap deliberately. Build the cognitive upskilling curriculum: prompt design, AI oversight, exception handling. Then give people somewhere to point the time you just handed back to them. “Go be more strategic” is not a curriculum. It’s a shrug with a budget.
5. Change the metric. Cognitive ROI, not headcount reduction. Measure judgment unlocked. Because the fastest way to make your whole workforce feel like frauds is to tell them, through your numbers, that the goal was fewer of them.

So here’s the question I’d put to you

Whether you lead ten people or ten thousand: are you building a culture where your people are told what only they can do — or are you quietly letting them infer that the answer is “less”?

The organizations that answer that well won’t simply survive this era. They’ll be the ones whose people walk into it feeling like painters who just discovered color — a little disoriented, briefly terrified, and about to make something no camera on earth could have captured.