The path you walked shapes the rules you’re willing to follow.

Traditional paths teach us to follow common and sometimes good steps. Non-traditional paths teach us to view those steps differently.

When I was seventeen, I broke my first computer. Twice.

The first time, I sheepishly sent it for repair. The second time, shame drove me to figure it out myself. Creating a boot floppy disk that could access the CD-ROM to reinstall Windows. I think that moment changed everything for me.

I learned three truths that possibly shaped my career:

• Breaking things isn’t fatal
• The path forward often runs through mistakes
• Self-reliance builds deeper understanding than following instructions

In 1999, I entered the tech arena without formal education. There were no YouTube tutorials or Stack Overflow. Learning meant tinkering with HTML, JavaScript, and CSS until something worked.

This ‘constructive breaking’ mindset has carried me through multiple roles spanning development, management, and leadership. Without the identity anchor of formal credentials, I’ve never felt trapped in one specialty.

As generative AI increasingly writes code through simple natural language prompts, with tools like Lovable, Bolt, and Replit that can build entire websites and SaaS products, the most valuable skill won’t be expertise in syntax or current frameworks. It will be the courage to break things, the resilience to fix them, and the adaptability to understand what actually solves the problem at hand.

What we fear in the AI era isn’t just disappearing entry-level coding jobs, but the learning path they represent. If AI handles the ‘easy’ parts, how will developers grow?

The answer might be counterintuitive: we need more people willing to break AI’s output, question its solutions, and develop through deliberate experimentation. The junior role evolves from writing basic code to breaking sophisticated systems, creating a new and equally valuable learning path.

The future belongs to constructive breakers.

It’s very easy connecting dots looking backwards. What made you successful and in which order.

Connecting the dots going forward is what we call strategy.

What often is lacking in strategy is the serendipity and leeway to pivot.

Make room for tactics to alter the destination. The best strategies evolve with new information.

Do the work to become the expert. Don’t wait for others. Don’t wait for completing the class, the book, the seminar. Do the work.

A few days ago I created Ghibli-style art with my kids using OpenAI’s upgraded image generator. “Make one of me” they asked with wide eyes.

As perfect copies appeared with just a few words, I wondered: what makes an “original” anymore?

This got me thinking about authenticity beyond copyright debates. When AI can create anything, how will we know who really made what?

And then it hit me: Non-fungible token aka NFTs . Those blockchain certificates that crashed after their speculative bubble might actually serve their intended purpose now. With greener blockchain solutions emerging, NFTs could finally become something useful: digital authenticity markers when we need them most.

People who guard their one big idea, fear they’ll never have another.

This makes them stuck. Unable to discuss. Unable to move forward.

Their minds become storage vaults rather than creative engines, just like when a midnight idea keeps you awake because you’re afraid to forget it.

When I started as a web developer there was a shortage of talent. You could learn, show your work, get hired. After the 2000 bubble burst, consistency and continuous learning kept careers alive.

Today, people who can effectively prompt AI tools or integrate AI into workflows are finding opportunities—just like those who could build websites in 1998. You don’t need a PhD. Practical skills are what companies need.

The pattern with AI is clear: shortage of skilled people now, then formal education later. Focus on using and learning, and you’ll be fine. Just as with web development before it.

If you have the right audience you don’t need bells and whistles. Focus on the content.

Trust your audience to recognize what’s important without highlighting it for them.

When you manage the context, you manage the results.

This is one of the most important aspects of developing with AI.

When building frontend features that need an API, set clear expectations to and from the backend. Provide only what’s necessary, not everything.

“If you wish to make an apple pie from scratch, you must first invent the universe.” - Carl Sagan

If you understand your organization, you’ll recognize the small ideas your leadership really wants.

The ‘what if we could’ ideas.

These are the ideas your boss or manager or owner wants to test but doesn’t dare ask for since it might be seen as micromanaging.

This becomes especially valuable as companies grow beyond the size where owners can personally know everyone.

Keep track of these ideas because they often surface in various projects. They’re leadership’s pet peeves and concepts that are hard for them to ignore.

Try asking them directly: “Tell me about some ideas that you’ve had for our products that you think we need or should test?”

Remember that listening doesn’t always mean acting. Understanding these ideas gives you valuable insight even if you don’t implement them all. Sometimes leadership just wants to be heard not necessarily have every thought turned into a project.