Why You Need to Start "Doing Things"
Imagine it’s the day after Christmas. Your boss calls and says, "I know you’re on vacation, but can you see what this new AI stuff can actually do?"

Most people would groan and go back to sleep. But Jeff Huntley, a tech leader from Australia, decided to test it out. He asked an AI to translate a complex coding library from one language (Rust) to another (Haskell)—a task that would normally take a human weeks of intense brain power. He hit "enter," took his kids to the swimming pool, and when he got back, it was finished. Perfectly.

That was the "oh shit" moment. And if you’re a student looking at a career in tech, you need to know that the rules of the game just changed forever.

The $5,000 Programming Language
Building a new programming language used to be a "Mount Everest" project. It took teams of genius engineers years of work and millions of dollars. Jeff decided to see if an AI could do it alone.

He set up an "agent"—a simple AI program that runs in a loop—and let it work for three months. The result? A brand-new language called Cursed. Total cost? About $5,000 in computer "tokens." That’s the price of a high-end MacBook.

This means the "barrier to entry" is gone. You don't need a million-dollar budget to build world-changing tech anymore. You just need to know how to drive the machine.

Stop Being a "Waiter," Start Being a "Doctor"
For a long time, being a software engineer was like being a waiter: someone told you what they wanted, and you "served" them by typing the code.

That job is dying. AI is the new waiter; it can type code faster than any human ever could. To survive in the future, you have to be the Doctor. A doctor doesn't just "do what they're told"—they diagnose the problem, understand the big picture, and decide which treatment is best.

Your value won't be in knowing how to type "Java" or "Python." In fact, your "identity" as a specialist is over. AI can translate between languages instantly. Your real value will be in your ability to solve problems and "orchestrate" a fleet of AI assistants to do the grunt work for you.

The "Secret Sauce" of Using AI
A lot of people try ChatGPT or Claude once, get a bad answer, and say, "AI is overhyped." Jeff argues that AI is like a guitar. You wouldn't pick up a guitar, strum one bad chord, and say, "This instrument is broken." You have to practice.

Here are the "pro tips" from the front lines:
- The Chalkboard Rule: Every time you start a new task, start a new chat. If you keep talking in the same window for too long, the "context" gets cluttered and the AI starts making mistakes. Clean the board!
- Know the Personality: Different AIs have different "vibes." Some models (like Anthropic’s Claude) work better if you’re very firm and "yell" at them in the prompt. Others (like GPT-5) actually get worse if you use all-caps. You have to learn the quirks of your tools.
- Agents Aren't Magic: An "agent" is just a simple code loop. It asks the AI a question, looks at the answer, uses a tool (like a calculator or a web search), and repeats. If you can understand that loop, you can build almost anything.

The Future Belongs to the "Doers"
There is a "freight train" coming for people who are sleeping at the wheel. Companies like Shopify and Gumroad are already saying that using AI isn't an "extra" skill—it’s a requirement for having a job.

But here’s the good news: the "gold rush" is just starting. If you are a high schooler who starts "doing things" now—building little apps, running AI agents, and practicing with these tools—you will have more power at your fingertips than the biggest tech companies had ten years ago.

The job market might be changing, but for the people who are willing to play, experiment, and do, the opportunities are limitless. Go build something "cursed." Go break things. Just don't stay still.

For Reference: