AI won't take all IT jobs — or at least not all of them. It removes manual, repetitive tasks and frees those who use it to focus on what we're actually good at: problem-solving, creativity, and seeing the big picture.

At the same time, there's massive hype. The bubble risk is real, but the biggest players — Microsoft, OpenAI, Google, Amazon, Meta, Nvidia — are investing billions. Not for fun, but because they see a path to automating vast amounts of current work and creating an entirely new level of productivity. In other words, replacing unpredictable humans with loyal robots.

On the other end, regular users like us are currently enjoying free lunches from AI services. Investment money is being poured into AI the same way it was into Wolt's early years: drivers could earn huge sums and customers got endless discounts — until the service was widely adopted and prices went up.

AI is a massive advantage for those who use it right — and a trap for those who outsource all thinking to it.

In this article, I'll share practical examples of how I use AI almost daily.

Coding: what used to take an hour, now takes five minutes

In coding, AI works at its best as a "senior developer" pair who writes boilerplate code, explains what the code does, suggests better solutions, fixes bugs, creates alternative implementations, and generates test cases.

I've done programming projects where AI produces the skeleton in seconds based on my description of the goal. Then I fine-tune the code myself. The benefit: I can focus on logic, not typing syntax.

System maintenance and troubleshooting

Maintenance and troubleshooting have traditionally been IT tasks where enormous amounts of time are spent. Before, this meant an hour or two of puzzling: Googling, browsing support forums, reading system documentation, interpreting error codes, and guessing the root cause. Often the actual cause was found only after extensive digging.

With AI, this process has sped up significantly. When I feed a log dump or an Azure/M365 error code to AI, it can summarize the log to essentials, identify the likely root cause, suggest fix options, and provide the necessary commands or scripts.

What used to take an hour or two now takes five minutes.

This doesn't remove the need to understand the big picture — quite the opposite. AI provides quick analysis, but the expert must decide which solution actually fits the environment. But the time sink of troubleshooting, the most frustrating and tedious part, is almost completely eliminated.

Web development: CSS fixes, debugging, and content development

Website maintenance involves a lot of small, time-consuming tweaking. That's exactly where AI excels.

CSS and HTML fixes in seconds

Before, fixing small issues meant digging into developer tools, browsing forum answers, and trial-and-error cycles. Now I just paste the code snippet to AI and describe what's wrong. It identifies the problem, suggests a fix, and provides a ready CSS/HTML version. Significant time saved.

Clear and user-friendly copy

Raw text in — clear, polished text out. I ask AI to clarify, shorten, create an SEO version, and adjust the tone (professional, casual, concise, persuasive). This improves both readability and search engine visibility without spending hours on editing.

Site sparring

I ask AI directly: "How does this page look from a user's perspective?", "What's missing?", "Where should I link to?", "Is the structure logical?" It quickly provides improvement suggestions and points out things you easily become blind to.

Graphic work: logos, infographics, and technical diagrams

With AI, I can produce visually high-quality material without spending an entire evening on it. For example, Microsoft-style diagrams are created in minutes. This way, many small businesses can create quality visual content without an external partner.

Documentation

Documentation is the work that many IT professionals would happily outsource to a secretary. AI makes this easy. I use it for writing architecture documents, change documentation, project final reports, and formatting miscellaneous configurations.

Often I just provide raw notes and ask AI to turn them into a clear, client-appropriate document. Of course, no sensitive information should be fed to AI.

Project work: plans, risks, and justifications

In project work, AI helps particularly well with the following:

Risk analyses

"What risks are associated with this change?" — a clear list with causes and mitigations.

Roadmaps

"Create a 6-month roadmap for M365 deployment." — I get a framework that I refine.

Justifying technical decisions

When you need to explain a technical solution in plain language to a client, AI is incredibly useful.

Conclusion

AI doesn't make you a better expert if you let it think for you. But it can make you many times more effective if:

You can solve problems without AI too — otherwise you don't know if AI's answer is correct or completely off.

You can create new things on your own — AI speeds up work, but you can't build expertise on generated content alone.

You can break problems into clear, solvable parts — AI works only as well as the problem you give it.

You know what you want and can ask for it precisely — a vague request produces a vague answer.

You use AI as an assistant, not a replacement — the best result comes when humans make decisions and AI handles the routines.

It's not about AI replacing the worker. It's about the worker who knows how to use AI replacing the worker who doesn't.

// Kalle Huttunen 14.11.2025

Closing joke: don't do this

A Pakistani newspaper accidentally published AI placeholder text verbatim. The lesson: always review AI output before publishing. Always.

If you want, I can also create an even snappier "front-page style" version with punchy one-line stats and a bold, infographic-ready layout — perfect for maximum reader impact. Do you want me to do that next?

Source: Dawn newspaper, Pakistan. AI text ended up directly in the printed paper.