My take on AI is a mixed bag. I love using it. I also see a lot of ways it's being misused — by companies, by individuals, and honestly sometimes by me. Let me explain.
So, AI in 2026. I love using AI. As an IT professional for the past 20-plus years, I think it's one of the most life-changing technologies to come along. But like all new technologies, companies have kind of taken advantage of it. I see a lot of posts and articles talking about how companies are choosing to replace people with AI. I think that's a huge mistake — and I think we're seeing the consequences. 2025 was full of mass layoffs, with lots of companies getting rid of employees to replace them with AI. 2026, so far, seems like some of those companies are taking a step back.
Was the real reason for all those layoffs actually AI? Or was AI just the excuse companies used as a corporate strategy to justify laying off so many people, rather than simply saying "we hired too many people after the pandemic and now we don't want to pay for them all"?
That said, I do think AI has done a great job in some areas at helping increase some individual’s capacity — and I mean 10 to 25%, not tripling output. It's helped on the productivity front in a similar range. Thinking back to my own experience using AI in a corporate setting, a lot of times we had to go back and rework things because the output was just okay, and there were things it got wrong. Keep in mind: the output is only as good as the input. If you don't put good information in, you're not going to get good results out.
Another factor is governance. What is AI allowed to access within a corporation, and what shouldn't it have access to? One thing we did — and I think it was a mistake — was we didn't start with governance. We started with the implementation, only to realize about six months in that there was a real lack of oversight. It was accessing things it really shouldn't have been, and presenting information to employees in departments that had no business seeing it. IT shouldn't know what Finance is working on, and vice versa. That's the kind of exposure that can pose serious problems down the road. It's not just a corporate issue either — it applies at the individual level too.
From a personal perspective, using AI as an IT professional at home — helping plan my day, drafting articles, answering emails — it's great. But it's a starting point. If you go into it thinking you can just use AI and publish whatever it produces, that's where the problem starts. AI isn't the end solution. It's an assistant, not a product producer.
I think back to a colleague who had an idea for a project. While the idea had merit on paper, the entire output was done with AI — maybe 90–95% AI-generated with 5% of his own thinking. When asked basic questions about how the application would work, how it would be secured, how it would benefit the company, he couldn't answer them. He'd been hoping to market the idea internally, maybe get a promotion out of it. Unfortunately, it amounted to nothing — which was sad, because he was genuinely looking at this as a career opportunity. The ease of AI made it appealing to skip the hard work of fully developing something, and that's a big part of the problem.
Which brings me to vibe coding.
Vibe coding has been around for a while now. It's pretty cool, I'll admit — I've been playing around with it myself. I am not a coder by any means. I have a basic understanding of HTML, CSS, Python, and a couple of other languages, but we're talking rudimentary. So I've been experimenting with using AI to write code for me, and it works. At least it works well for me.
Here's the thing though: while I don't have a lot of knowledge, I have enough to know what I should and shouldn't be doing. My experiment was really just to play around. I built a completely local web app for myself — a Next.js frontend with a PocketBase database, only ever going to be local, no monetization possibilities. It was really just for me. It's a kind of dashboard for a lot of things I do day-to-day, and I enjoy using it. I enjoy building it.
What the experience taught me is that while agentic AI coding is genuinely impressive, and the fact that a fairly basic developer like me can build a functional application is remarkable, it's really not meant for just anyone. It is not a replacement for a developer in any way, shape, or form. If you go into it as a salesperson, a marketing executive, or a gamer thinking that you can now create your own apps and games because an AI assistant will just do it for you — you've got another thing coming.
I'm not saying it's impossible. I'm saying you still need to know enough to know what you're putting in and what you're getting out. That's one of the major problems with AI: people see it as a replacement for doing the work, so they think they don't need to know how to do the work. That's a huge mistake.
You need to know how to do the work. More importantly, you need to know how to read what the AI produced so you can spot where it makes mistakes and fix them. There is nothing more dangerous than someone who creates something without understanding what they've created — and then releases it to the world. Your application might be fully functional. AI is great at making things that work. But there's a big difference between functional and safe. If you don't know how to check whether your database is storing sensitive information in a secure format, if you don't know how to spot that your system is exposed, you're opening a Pandora's box. Serious liability. Serious problems for real people — especially if you go to market with it.
A few weeks ago I was at a business orientation session offered free to entrepreneurs in my area. One individual was talking about their plan to build and sell an application entirely using AI, to solve a problem they'd experienced in their own business. I don't normally speak up in these settings — I'm pretty introverted — but I had to say something.
There's a big difference between understanding a problem from a usability perspective and understanding what goes behind the solution. I studied industrial design for several years before getting into IT, and one thing they drilled into us was that understanding how something is used is completely different from understanding the engineering and infrastructure behind it. What AI does is make complicated things seem simple. But that complexity is still there — it's just been glossed over. If you go in thinking you can create anything you want without understanding what a properly functioning application actually requires — secure, functional, not exposing sensitive data, connected to the right systems — and without the ability to go into the code and identify problem areas, good luck.
AI is not perfect. If you can't spot the imperfections and you release something, people will spot them for you. And if you don't know enough to fix them, you're in trouble.
So that's my take on vibe coding. I'm not saying don't do it. I'm saying go in with an open mind. Practice on your own system, locally, in an isolated environment — a VM or container if you can. Play with it there before you try to build something you're going to put in front of people.
And remember: AI isn't free. There are free versions that are fine for general use, casual questions, replacing a basic Google search. But once you really get into it, you start to see what you get for free has usage limitations and capability ceilings. Don't get roped into paying for something right away. Play with the free tiers, find one you're comfortable with, understand its limitations, then decide if it's worth paying for.
For myself, right now I'm experimenting with coding, and I've also built a Markdown-to-HTML preview module in my dashboard. I can write a blog post in Markdown and instantly see exactly how it will look on my website — not just generic HTML output, but styled with my own CSS. Small thing, took maybe half a day, but it saves a lot of back-and-forth. I use AI for general planning, bouncing ideas around, researching tools and products, stuff like that.
I genuinely enjoy AI. I just think you have to go into it with the right mindset, understand that it has limitations, and be ready to adapt. If you're younger and entering the job market soon, especially in fields that will heavily use AI internally, the more you know now, the easier your transition will be. AI experience is transferable. What you learn using it at home carries over into the workplace.
Just please — don't do what my former colleague did. Don't try to market yourself through AI work you haven't vetted and don't fully understand. If you use AI to produce a report, that's fine, but be transparent about it. Most companies have governance policies around this — follow them. Never claim work that AI did as your own.
AI is an assistant. It's a helper. It's not you.
Hope you found this interesting. If you have any questions, feel free to ask. See you next time.