AI, AI, AI. I'm pretty sure you've heard this word so often in the past two years that if it had a dollar attached to every mention, we'd all be early retirees by now. It has become the buzzword for 2024, 2025, and probably for the foreseeable future. And honestly, I think that's fair. Because AI isn't just tucked away inside the tech sector anymore. It has seeped into nearly every corner of our lives.
But here's the thing. I'm not a doctor, a farmer, or a chef. I'm in tech, so I can only really speak about the impact of AI in my own world. Specifically, how it has shaped the way I work, the way I create, and the way I think about productivity. I've been using AI tools and agents in ways that save me time, boost my creativity, and sometimes even surprise me with how effective they can be.
And let's be honest. If someone today says they're not using AI at all, my first thought is, are you living under a bubble? Or worse, still manually organizing your email folders? AI is everywhere. The trick is knowing how to actually use it so that it boosts your work rather than creating more of it. Because here's the truth: AI doesn't do my job for me. It doesn't replace the thinking, the decision-making, or the actual creativity. What it does is act like a turbo button. With the right tools and a little bit of patience, I can do twice, three times, sometimes even ten times more than I could before.
And yes, I'll admit it—sometimes that feels like cheating. But if it's helping me, and the results are good, why fight it? My other article is in line to this article, but in that article I talk about where I use AI to build product, whereas in this article, I'll be diving into how I use AI and what tools do I use.
Why Automation Was My Starting Point
Before diving into image generation, fancy video tools, or coding assistants, my first obsession with AI was automation. I've always been the kind of person who thinks: if I'm repeating the same action every day, why not build a system that does it for me?
That's where tools like n8n came into play. If you haven't heard of it, think of it as your own personal assistant that doesn't complain, doesn't need coffee, and can actually follow instructions. I started building workflows for everything—reminders, text responses, email handling, even how I get notified on WhatsApp or Telegram.
It might sound small, like, who cares if you get a reminder through a fancy workflow instead of just putting it on your calendar? But when you multiply these "small" automations over days, weeks, and months, it's a game changer. Imagine the difference between manually sweeping a room every morning and just installing a Roomba that does it while you sleep. You'll notice it.
And the best part? You don't need a huge budget. I run n8n locally on Docker. No payments, no subscriptions, no sneaky "free trial" that suddenly charges you at midnight. Just me, my laptop, and a growing collection of workflows that save me hours.
Once you get a taste of this kind of automation, you don't go back.
The Dopamine Hit of AI Tools
Now, let's talk about AI tools themselves. If you've been around long enough, you'll remember the early days when AI outputs were… well, messy. We used to call it "AI slop." You'd ask for a blog post, and it would give you something that sounded like it was written by a bored alien who skimmed Wikipedia once.
Fast forward to now, and the difference is night and day. These tools can generate polished emails, working prototypes, structured documents, and sometimes even entire applications in a single prompt. And that's where the dopamine hit comes in.
There's something oddly satisfying about dropping a single, well-written prompt and watching an AI generate something that would have taken me hours. Is it always perfect? No. But when it works, it's like hitting a perfect shot in golf—or in my case, like finding an actual working pen when you need one. Rare, but glorious.
The Art of Prompt Engineering
That brings me to what I think is the secret sauce of AI: prompt engineering.
If you've ever been frustrated by bad AI output, I can almost guarantee it wasn't just the model's fault. It's the classic "garbage in, garbage out" principle. Tell an AI, "Build me a to-do app," and you'll probably get something that looks like it belongs in a high school coding assignment. Tell it, "Build me a to-do app with user authentication, local storage, task categories, and drag-and-drop functionality," and suddenly, you've got something usable.
It's not magic. It's about detail. And the more you practice writing prompts, the better you get at it. At this point, I consider prompt engineering a skill in its own right. The same way people used to learn SQL or Photoshop, now you learn how to talk to machines in a way they understand.
And let me tell you, getting good at this pays off.
AI as My Brainstorming Partner
One of my favorite uses of AI is simply brainstorming. Whether I'm exploring a new project idea, analyzing competitors, or figuring out why one of my workflows isn't working, I use tools like ChatGPT, Perplexity, and Gemini.
Do they give me the right answer every single time? No. But if I get even 80 to 90 percent of the value, that's a massive boost. It's like having a mentor who doesn't get tired of your endless "what if" questions. And when I need something written, like documentation or structured notes, I can just throw in my bullet points and let the AI do the heavy lifting.
I've almost stopped using social media, because AI brainstorming has become my new time sink. At least this one actually makes me smarter.
The Creative Playground
For creativity, AI has unlocked a whole new world. I use MidJourney, Nano Banana (yes, that's a real model), and Google's V3 for image and video generation.
Here's how it usually works: I imagine something, I generate an image, and then I start tweaking. Sometimes I animate it, sometimes I refine it, sometimes I just stare at it thinking, "Wow, that looks way better than what I had in my head."
Video generation has been another level. MidJourney now lets you animate, but Google's V3 really set the bar. The quality is just… clean. It's consistent in a way that older tools weren't. I've also dabbled with Cling and Seedance, and they're fun too.
Do I always use these tools for work? No. Sometimes I just create random assets for the dopamine hit. It's the AI equivalent of doodling in a notebook.
Coding with AI Agents
Now, onto coding. This is where my workflow really changed.
At first, ChatGPT was the only option. It helped, but it hallucinated too much. Then came Claude, and for a while it was unbeatable for programming. But recently, I found my perfect combo: ChatGPT-5 and Cursor IDE.
Cursor is an IDE that integrates AI deeply, almost like having a coding partner who can read your entire project and suggest structured changes. Unlike Copilot, which is great but limited, Cursor's agent models can actually restructure codebases, generate new files, and set up entire apps. It feels like cheating, but again, it works.
Do I trust it blindly? Absolutely not. If you're shipping code to production, always review what the AI changes. Otherwise, you'll spend more time fixing than creating. But as a productivity boost, Cursor plus GPT-5 has been a game changer for me.
The Bigger Picture
Beyond automation, creativity, and coding, there are dozens of other AI tools I use here and there: NotebookLM, Julius AI, Anthropic's tools for research and data analysis. Some require more manual input, some are smoother, but the overall theme is the same.
And here's the kicker: most of these tools are either free or have generous free tiers. You don't need to spend a fortune to start experimenting. If you're a student or someone just curious, the free options are more than enough. If you're serious and want to push productivity further, then yes, some pro versions are worth it. Personally, Cursor's paid version was my best investment.
AI Is Not Replacing Us
Let me end on this. People love to say AI is going to take over our jobs. I don't buy it. Not in the way people think.
AI is not here to replace me. It's here to level me up. From level one to level two, from level two to level three, and so on. The more the models improve, the more I improve alongside them.
And honestly? Learning AI has been the most fun skill I've picked up in years. It gives me that dopamine hit without needing caffeine or anything stronger. You can automate small tasks, brainstorm big ideas, or code entire apps. And each time, you come out a little sharper.
So my advice is simple: just start. Automate something tiny. Write a prompt. Generate an image. Build a workflow. You don't need to wait for a company to train you. The tools are already here, most of them free, and the only real barrier is curiosity.
The future belongs to those who learn to work with AI, not against it. And the sooner you start, the sooner you realize: this isn't about AI replacing us. It's about us getting better, faster, and more creative with AI at our side.