Remember when everyone was absolutely certain that artificial intelligence was about to wipe out entire professions overnight?
I do. It felt like yesterday everyone from tech conferences to dinner tables was repeating the same terrifying line: “AI is coming for your job.” And honestly, some of the warnings came from people who really should know better.
Then Nvidia’s CEO sat down for a long chat and casually dropped a truth bomb that made me rethink the whole conversation.
The Prediction That Aged Like Milk
Back in 2016, one of the most respected minds in artificial intelligence made a bold call. He looked straight at the future and declared that people should stop training to become radiologists altogether. His reasoning? In five years—by 2021—AI would be better at reading medical images than any human doctor ever could be.
Five years came and went. Then another four on top of that. So… what actually happened?
Here’s the part that should make every “AI doomer” pause: the number of radiologists didn’t crash. It grew. And not just a little—projections now show the field could expand by as much as 40% over the next three decades.
“What’s ironic is that the number of radiologists has actually grown. Today, just about every radiologist is using AI in some way.”
– Jensen Huang, CEO of Nvidia
Think about that for a second. The exact job that was supposed to vanish first didn’t disappear—it got better, bigger, and more in demand.
It’s Not About Replacing the Human—It’s About Supercharging Them
This is where most people get the story completely backwards.
The real purpose of a radiologist has never been to stare at grainy scans for hours. The real job—the part that actually matters—is detecting disease and saving lives. Studying images is just a painfully slow step in that process.
AI didn’t eliminate the radiologist. It eliminated the boring, time-consuming part that nobody liked anyway.
Suddenly one doctor can review triple the number of scans in the same shift. Accuracy goes up. Patients get diagnosed faster. Hospitals make more money. And guess what hospitals do when they have better margins and higher patient volume?
They hire more radiologists.
It turns out abundance creates demand, not scarcity.
A Pattern We’ve Seen Before (And Somehow Keep Forgetting)
In my view, this radiologist story is just the latest chapter in a very old book.
Go back to the 19th century. People were terrified that tractors would put millions of farm workers out of a job. And they were right—fewer people were needed to grow food. But food became so abundant and cheap that entirely new industries exploded into existence.
Fast-forward to the ATM. When those machines started popping up everywhere in the 1970s, bank tellers were supposedly doomed. Actual result? The cost of running a branch dropped so much that banks opened more branches, and the number of tellers actually increased for decades.
- Tractors → fewer farmers, more food, new jobs everywhere else
- ATMs → fewer transactions per teller, more branches, more tellers
- AI image analysis → fewer hours staring at scans, more patients helped, more radiologists hired
The pattern is almost boringly consistent once you see it.
The Real Risk Isn’t AI—It’s Refusing to Use It
Here’s the line that stuck with me the most from the entire conversation:
“You’re not going to lose your job to AI. You’re going to lose your job to someone who knows how to use AI.”
Ouch. But also… fair.
We’re already seeing this play out. The lawyers who embraced AI research tools didn’t get replaced—they started handling twice the caseload and charging premium rates. The designers who ignored Midjourney and Stable Diffusion are now competing against people who can produce ten concepts in the time it used to take to sketch one.
The tool isn’t the threat. Complacency is.
Yes, Some Jobs Will Actually Disappear (And That’s Okay)
Let’s not sugar-coat everything. Some roles really will shrink or vanish.
Any task that is repetitive, rules-based, and doesn’t require human judgment is on the chopping block. Data entry clerks who spend all day copying numbers from PDFs? Rough outlook. Junior coders whose entire job is writing basic CRUD apps? Better start leveling up fast.
But here’s the part people miss: every single time we’ve automated the drudgery in the past, we didn’t end up with mass unemployment. We ended up with new categories of work we couldn’t have imagined before.
The Robot Revolution Will Need Millions of Humans
Take the coming wave of physical robots. When humanoid robots finally become practical (and yes, that day is closer than most people think), we’re not just going to have millions of silent metal workers quietly replacing humans.
We’re going to need:
- Technicians who maintain them
- Programmers who teach them new tasks
- Entire supply chains making robot “clothes” and accessories
- Mechanics specialized in robotic joints and actuators
- Insurance adjusters for robot accidents (yes, really)
- Trainers who help robots learn physical intuition the way we teach apprentices
One breakthrough doesn’t eliminate work—it multiplies it across a dozen new specialties.
In many ways it feels like we’re standing at the edge of the biggest job-creation event in human history, and half the conversation is still stuck on fear of loss.
What This Means for Anyone Worried About Their Career
If you’re feeling anxious about AI right now, here’s the mindset shift that helped me sleep better:
- Stop asking “Will AI take my job?” Start asking “How can AI make me ten times more valuable?”
- Focus on the parts of your work that require judgment, creativity, empathy, and taste—those are the parts machines still struggle with.
- Treat AI like the world’s most over-eager intern: give it the boring stuff and keep the interesting decisions for yourself.
- Invest time in learning prompt engineering, tool chaining, and basic data literacy the same way previous generations learned Excel or PowerPoint.
The workers who thrive won’t be the ones fighting against the tools. They’ll be the ones who figure out how to conduct the orchestra.
Looking back at that 2016 prediction, I can’t help but feel a little grateful it was so spectacularly wrong. Not because the technology failed—it succeeded beyond what most people expected—but because it succeeded in exactly the way that makes humans more capable, not less needed.
Maybe the future of work isn’t a dystopian nightmare after all. Maybe it’s just the first time in history when almost everyone gets to spend more of their day doing the parts of the job they actually went to school for.
And honestly? That sounds pretty good to me.
