AI isn't the problem. Your dependence is.
Every week, someone says AI will eliminate all jobs. Every week, someone else says that's an exaggeration.
As always, the truth is somewhere in between.
AI will change many jobs. Replace tasks. Destroy certain business models. Make some skills less valuable. Force many people to reinvent themselves faster than expected.
But AI isn't the biggest problem. The biggest problem is something else:
Too many people depend on a single income source, a single skill, a single company, and a single professional identity to survive.
That's the real risk. Not AI. Dependence.
The problem didn't start with AI
AI didn't create this fragility. It just made it more obvious.
For decades, the plan was simple. Study something. Get a good job. Get promoted. Save. Retire.
That plan worked for many people. But it also created a quiet problem: it taught us to be good inside someone else's system.
To execute tasks. Wait for instructions. Depend on a company for income, structure, identity, and security.
While the world moved slowly, that could work. But when technology changes the value of skills every six months, that dependence becomes dangerous.
If your stability depends on your role remaining necessary, you're exposed. If your income depends on someone else continuing to approve your salary, you're exposed. If your identity depends on a job title, you're exposed. If you only know how to operate within a structure someone else built, you're exposed.
AI didn't invent that problem. It held a mirror up to it.
Having a job isn't bad
Let's clarify something. I'm not saying having a job is bad.
A job can give you experience. Stability. Teach you how an industry works. Give you relationships. Capital. It can be a very useful stage.
The problem isn't having a job. It's having no other capability outside it.
It's being paid to perform one function but never learning to create complete value.
Knowing how to do “your part,” but not understanding how to find a customer, sell, distribute, build an offer, create trust, or bring something to market.
That's the difference. Someone with a job and the ability to create something of their own has options. Someone with a job and no autonomy has one door.
When you only have one door, every change feels threatening.
AI rewards people with agency
There's a word being used a lot lately: agency.
We don't need to complicate it. Agency is the ability to move without someone giving you permission.
See an opportunity and act. Learn something because you need it. Create something nobody asked for. Try. Fail. Adjust. Try again.
Most people don't lack intelligence. They lack agency.
They wait for someone to tell them what to do. A boss. An algorithm. A company. An education system.
But the coming world won't reward whoever waits for perfect instructions. It will reward the person who can look at a confusing situation and say, “Okay, I'll solve this.”
That's the skill. AI amplifies that kind of person enormously.
If you have agency, AI becomes an incredible tool. It helps you research. Write. Design. Code. Sell. Learn.
If you don't have agency, AI becomes another app you use twice and abandon.
The five skills that will matter most
1. Agency
The ability to start without permission.
Don't wait for someone to hire you to develop a skill. Don't wait for someone to ask for a project. Don't wait until you feel ready.
Start.
2. Taste
Taste means knowing what's worthwhile. What's good. What's mediocre. What deserves publishing. Which idea has potential.
AI can produce infinite versions. Someone still has to choose. Choosing well is a skill.
3. Persuasion
Not manipulation. Persuasion.
Today, every business is a media company with a product or service behind it!
Many entrepreneurs tell themselves a lie: “If my product is good, people will come.”
It doesn't work that way. I'm not saying the product doesn't matter. Of course it does. But many products don't fail because they're bad. They fail because nobody sees them.
4. Persistence
Most people lose because they quit too quickly. They publish three times and say content doesn't work. Try one AI tool and say it's useless.
The people who win aren't those who never fail. They're those who don't give up.
5. Iteration
This is the most important one.
Publish. Measure. Learn. Adjust. Publish again.
Launch. Listen. Improve. Launch again.
That loop is the foundation of almost everything.
The solution is creating something of your own
The best way to become less vulnerable isn't complaining about AI. It's building something of your own.
It doesn't have to be a giant company. A million-dollar app. It can be a service. A tool. A digital product. A consultancy. A system.
Something that forces you to learn complete skills, not isolated tasks.
When you build something of your own, you learn things no job teaches in quite the same way. Defining a problem. Getting attention. Selling. Improving. Handling rejection. Thinking like an owner.
That mental shift changes everything.
Creating content and using software gives you power without depending on anyone
Today, almost anyone can use two forms of leverage:
Media and software.
Media is content. Posts. Videos. Newsletters. Podcasts. Ideas published online.
Software is product. Apps. Automations. Tools. Systems. Agents.
They have something in common: they work while you sleep.
A post can sell for you for months. A video can reach thousands of people. A tool can solve a problem a thousand times without you being there.
But there's a trap. Being able to create something doesn't mean anyone wants it.
That's what many people don't understand. AI lowered the barrier to production. It didn't eliminate the need for distribution and judgment.
Anyone can create an app. Few create an app someone uses. Anyone can write a post. Few write something worthwhile.
That's why media matters so much
If I had to choose one skill for the coming years, I'd choose media.
Not because everyone has to become an influencer, but because anyone who wants to build something needs distribution.
Distribution starts with communication. Knowing how to write. Explain. Tell a story. Build trust. Show how you think. Make someone interested before selling to them.
That's media.
The problem is no longer just creating the product. It's making someone care.
You can have an app, but without distribution, nobody uses it. An offer, but without trust, nobody buys. Talent, but without visibility, nobody knows you exist.
The first step is the most important
Most people freeze because they're looking for “the big idea.” The perfect niche. Perfect product. Perfect content. Perfect business.
Start simpler. Ask yourself:
What do I know and do extremely well? What problem did I solve for myself that others still struggle with? What does my industry consider normal that I think should change? What experience seems obvious to me but would be valuable to others?
There's material there.
You don't need a perfect idea. You need a first piece. Something that puts your thinking in front of the world.
You can't improve something that doesn't exist. Many people stay stuck there.
The final idea
AI won't replace everyone. But it will expose people who depend too heavily on a single structure.
One company. One skill. One income source. One identity.
The solution isn't hating AI. Nor is it believing AI will solve everything for you. The solution is becoming less dependent.
The future won't reward whoever waits. It will reward whoever knows how to act. Learn. Communicate. Build. Iterate.
AI isn't the enemy. Dependence is.
Conclusion
Don't wait for AI to force you to reinvent yourself. Start before that.
Publish something. Build something. Sell something. Learn something. Iterate.
In the AI era, the risk isn't that AI thinks. It's that you stop thinking.
Thank you for being part of this great community.
Pura vida!