What AI Can’t Fake
Chris Rosato
What AI Can’t Fake
Why knowing how to learn now matters more than knowing what you know
Three months ago I watched a junior designer feed a napkin sketch into a GenAI tool and get a polished logo back in ninety seconds. It would’ve taken me a weekend in 2010. She stared at the screen and whispered, “So… what’s my value now?”
I’ve been there. The first time an AI pair-programmer rewrote my code before I’d finished reading it, I asked the same thing.
Short answer: your value isn’t in what you know. It’s in how fast you learn from what the machine gives you—and then layer in the judgment, taste, and story it still can’t fake. Generative AI hasn’t killed the expert; it has rewritten what expertise means.
The Knowledge Fallacy
For centuries we treated knowledge like treasure. You hoarded it, guarded the key, and charged admission. Then AI walked through the wall.
Harvard Business Review calls this shift “domain knowledge on demand,” arguing that value moves from storing facts to shaping them in real time.⁴ You can already see the pivot:
In law, the model drafts the memo; the associate earns her keep by spotting nuance it missed.
In medicine, AI reads the chart; the doctor builds trust by reading the patient’s face.
Teachers draft lesson plans in minutes, then rewrite them based on what lights up a room of ninth-graders.
The human edge isn’t gone—it’s just moved upstream.
The New Stack of Expertise
The World Economic Forum’s latest skills survey ranks curiosity, empathy, and influence alongside analytical thinking.¹ McKinsey forecasts that demand for social-emotional abilities will soon outpace many technical ones.⁵ Put those trends together and a four-layer stack emerges:
AI Literacy – prompt, interpret, sanity-check
Domain Context – spot when the model is confidently wrong
Sense-Making & Story – turn raw output into insight humans trust
Ethical & Emotional Intelligence – know where the data ends and people begin
Remove any layer and the tower wobbles. Combine them and you’re the person everyone wants around when the model hallucinates at 4 p.m.
Not technical? That’s the point. The fastest-growing roles are hybrids: domain depth + people sense + just enough machine fluency to steer the output.
New Roles (That Would’ve Sounded Made-Up in 2018)
Job boards now feature:
AI Trainer – shows models what “good” looks like
Prompt Architect – builds reusable language scaffolds
AI Ethicist – checks bias, enforces guardrails
These hybrid jobs top the World Economic Forum’s Jobs of Tomorrow growth charts.⁶ Somewhere out there, a Prompt Architect is explaining to their parents that—yes—it’s a real job, and no, it isn’t a cult.
My “Keep → Curate → Coach” Audit
| Category | Guiding Question | Last Week’s Example |
| Keep | Does it rely on judgment or trust? | Final sign-off on a leadership curriculum |
| Curate | Can AI draft 80 % and I finesse the rest? | Synthesising a market-trends brief |
| Coach | Could AI handle this with light oversight? | Formatting citations for a blog post |
“Coach” tasks get automated or delegated. “Curate” work is co-piloting. “Keep” work is your human edge. The audit turns AI from threat to tool—and feels oddly like cleaning out a mental junk drawer.
The Inequality Problem
About 60 % of jobs in advanced economies are highly exposed to AI; in emerging markets the share is roughly 40 %.² If reskilling lags, the gap between those using the tools and those used by them widens quickly. Treat AI literacy like basic infrastructure or risk widening that divide.
Takeaway
When knowledge is cheap and speed is infinite, expertise must change shape.
The old expert knew things.
The new expert learns faster than the model can fake it—and still leads the room when the music stops.
Start this week. Audit your work. Sharpen your AI fluency.
If learning is the new knowing… how fast can you learn?
Footnotes
The Future of Jobs Report 2023, World Economic Forum. https://www3.weforum.org/docs/WEF_Future_of_Jobs_2023.pdf
Aiyar, S. et al. “Artificial Intelligence and the Future of Work.” IMF Staff Discussion Note 24/01, 2024. https://www.imf.org/-/media/Files/Publications/SDN/2024/English/SDNEA2024001.ashx
Manyika, J. et al. “Generative AI and the Future of Work in America.” McKinsey Global Institute, 2023. https://www.mckinsey.com/~/media/mckinsey/mckinsey%20global%20institute/our%20research/generative%20ai%20and%20the%20future%20of%20work%20in%20america/generative-ai-and-the-future-of-work-in-america-vf1.pdf
Davenport, T. H. & Mittal, N. “AI Will Transform How Organizations Define Expertise.” Harvard Business Review, Nov 2023. https://hbr.org/2023/11/how-generative-ai-will-transform-knowledge-work
Madgavkar, A. et al. “Skill Shift: Automation and the Future of the Workforce.” McKinsey Global Institute, 2018. https://www.mckinsey.com/featured-insights/future-of-work/skill-shift-automation-and-the-future-of-the-workforce
World Economic Forum & Burning Glass. “Jobs of Tomorrow: Mapping Opportunity in the New Economy.” 2020. https://www3.weforum.org/docs/WEF_Jobs_of_Tomorrow_2020.pdf
Subscribe to my newsletter
Read articles from Chris Rosato directly inside your inbox. Subscribe to the newsletter, and don't miss out.
Written by