It’s official — we’re in the AI age. Everyone knows it. I recently listened to a YouTube video discussing how websites are dying because of AI. And honestly, it’s not surprising. In fact, it might be inevitable.
Most websites today exist to provide quick, pragmatic information: how to fix a problem, how to cook something, how to understand a concept. That kind of content — direct, instructional, data-driven — is exactly the kind of thing AI excels at reproducing. And I don’t blame websites for doing that. It’s useful. But here’s the consequence: if your job is delivering sequential information without deeper insight, you’re in danger of being replaced. Even programming isn’t immune. AI-powered code generation tools can write embedded firmware, deploy cloud infrastructure, and scaffold entire web applications — if similar work has been done before.
Now, I know I’m treading on thin ice. Some people will disagree, maybe even strongly. But I’m going to say it anyway: today’s AI cannot truly create. It’s not some mystical force. It’s a predictive machine. Underneath the hype, most large language models (LLMs) are statistical engines trained through self-supervised learning. They don’t “understand” meaning — they generate plausible continuations based on what they’ve seen before.
And what they’ve seen is a lot. We’ve pumped the entire digital world into these models — billions of documents, codebases, images, and conversations. But even so, creativity isn’t just output. Creativity is effort. It’s intention. It’s invention.
Have you ever seen an AI invent something entirely new — not remix, not rephrase, but genuinely originate? I haven’t. I hope one day I will. I don’t believe in halting technological progress. But I do believe we should shape it intentionally. And to do that, we need real discussions — not just around what AI can do, but around what we want from it.
Here’s something most people don’t think about: when you interact with an AI model like ChatGPT, it isn’t starting from a blank slate. There are “system prompts” — invisible instructions baked in before you even begin — telling it to be helpful, informative, safe, and consistent. These are important for reliability. But they also constrain the space in which the model can “think.” What if the kind of creativity we want from AI lies outside those bounds? What if we’ve shackled AI to be predictable before it ever had a chance to be inventive?
That’s not just a technical question. It’s a philosophical one.
Another issue is scale — and energy. AI isn’t just abstract math; it runs on very real, very power-hungry hardware. Modern GPUs, the workhorses of AI training and inference, consume vast amounts of electricity and generate intense heat. As silicon transistors shrink toward their physical limits, we hit diminishing returns. Power usage increases, but efficiency doesn’t always follow. Sure, quantum computing may eventually help — but quantum chips still rely on classical silicon for control, and they’re far from mainstream or reliable.
So, where does all this leave us?
Here’s where I land: AI is a screwdriver.
That’s all it is. A tool. A powerful one — but still a tool. You don’t criticize a carpenter for using a screwdriver, do you? Of course not. The job is to fix, to build, to solve — and if AI can help, use it. Just like there are different screwdrivers for different screws, there are different AI models for different tasks.
But like any tool, misuse can lead to damage — to the tool, to the user, or to the work itself. If you force the screwdriver where it doesn’t belong, or if you forget that it’s the human hand guiding it, you’ll strip the threads and break things.
So think. Choose your problem. Form your idea. Then pick up your screwdriver. Use AI. Build something. Share it with others.
That’s the only way we move forward — not as passengers of the AI age, but as its inventors.
Eren Soylu