Is Learning AI the Only Way to Prepare for the AI Era? | AI Survival Diary #31

AI Survival Diary Series | #31

Is learning AI the only way to prepare for the AI era?

 

In post #30, I left off with this question:

 

"So what should an individual actually prepare?"

 

Ask AI experts these days, and the answer is pretty much always the same.

 

"Start with vibe coding. You don't need to know a programming language — you can build what you want just by describing it in plain language. Try an AI agent too. It can book your flight, reserve your hotel, and even handle the payment."

 

Hearing that, it feels like I should be doing this right now.

 

"Shouldn't I hurry up and try this myself?"

"If I don't learn AI, will I end up like the elderly folks who wander off from a kiosk because they can't figure out how to order?"

But underneath that, one question kept nagging at me.

So if I learn it, does that actually solve the problem of making a living? 

●   ●   ●

I'd seen this scene before, somewhere.

 

A few years ago, everyone said: build your own income stream through YouTube, through a smart store.

 

"Even when people share exactly how to make money, 95% quit. Just don't quit, and you'll make it."

"Alright. Maybe it's time I built something of my own, outside a regular job."

 

But when I actually looked into it later, the issue wasn't simply about quitting or not. The reality itself turned out to be harder than that.

 

₩1.29B

Top 1% of Korean YouTubers, average annual income (2024)

₩24.6M

Bottom 50%, average annual income (2024)

 

And this is only the story of people who actually made money from it. People like me, who never got to that point at all, are probably the vast majority.

 

Those creators weren't lying. They really did succeed.

 

But

 

no one emphasized how rare it actually is to reach that kind of income.

 

What if, from the very start, someone had told all those aspiring YouTubers just how brutal that competition really was?

 

At least I would have known what I was actually signing up for.

 

●   ●   ●

AI felt similar, somehow.

 

Everyone says, "Learn AI."

 

Learn vibe coding, use an AI agent, try every new AI tool the moment it comes out.

 

But this time, it's not just a matter of time.

You have to pay a monthly fee to really do it properly.

You have to keep learning new tools, keep tracking features that keep changing.

 

Whenever I watch one of those videos, the same thoughts hit me:

"Should I be doing this too?"

"Do I need to learn that one as well?"

 

And then I'm back to the same question.

 

If I learn it, can I make money? Can I survive at my company? Or am I just learning it so I don't fall behind?

 

And an actual answer to that question was hard to come by.

 

What came back instead was always, "Just try it."

 

●   ●   ●

So — why was I trying to learn AI in the first place?

The biggest thing driving me was fear — fear of ending up like those elderly folks stuck at the kiosk.

Underneath that was a vague hope: maybe learning AI would somehow solve the problem of making a living.

 

But thinking about it again — not being able to do vibe coding, not being able to use an AI agent, isn't really going to turn me into that person at the kiosk.

 

Using the AI models that are already widely available, reasonably well, is probably enough.

 

And the vague idea that learning AI alone would somehow solve how I make a living?

 

That, too, was just fear talking — reaching for a tool before I'd even decided what I was going to do with it.

 

I was rushing, without thinking things through in order.

 

I don't think the AI era changes the basic way people make a living all that much.

 

When the internet arrived, when YouTube took off — in the end, the question was always the same.

 

What do I actually need to do, so that making a living isn't a problem?

No matter how far AI advances, deciding "what to do" is still something each of us has to figure out for ourselves.

 

So is it really the right order — driven by the anxiety that "I have to learn AI to survive" — to pay real money for vibe coding and an AI agent before I've even decided what I'm actually going to do?

 

Right now, what we're really wrestling with is:

"Learn AI → figure out the rest later."

"Decide what I'll live on → learn only as much AI as that requires."

 

I think the second order is the right one.

 

This isn't an argument against learning AI. I think I should learn it too.

 

I just want to say — learning it isn't supposed to come first.

 

●   ●   ●

Looking back, none of this worry is new, or unique to the AI era.

 

I asked myself the same question back when I was facing pressure to retire.

 

"How much longer can I hold on at the company? Shouldn't I be figuring out now what I'll do after I retire?"

 

Back then too, it came down to "What will I actually live on, going forward?"

 

AI didn't invent a new question. It may just be handing us the same old question, faster and louder than before.

 

So I've decided to think about it a little differently now.

 

Preparing for the AI era doesn't mean rushing to learn every new tool.

 

It means thinking more concretely about how I'm actually going to make a living, figuring out what that work would require —

 

and once I have some sense of direction, learning whatever AI I need for that.

 

That feels like the right order.

 

●   ●   ●

I still haven't found the answer.

 

I haven't decided what I'll live on, and I don't know how much AI I actually need to learn.

 

But one thing feels a little clearer now.

 

The AI era didn't change the question people actually need to ask.

"What will I live on?"

 

If I keep chasing what to learn without answering that question first,

 

I'll end up right back where I started — standing in front of the words, "Just try it."

 

So in the next post, to find work I can actually do, I want to follow one more question:

 

what kind of value still holds up, even in the AI era?