What's the deal with AI?
The truth about artificial intelligence, how it's used, and how to actually put it to work in your business.
These days everyone talks about artificial intelligence. Few people, though, understand how to use it, and even fewer realize what such AI actually costs.
It has to be said that AI is a powerful and capable tool. It can automate countless redundant tasks for us that waste our time and energy. But we often inflate our idea of what AI can do and raise our expectations. AI is not a replacement for human work. Nor is it some autonomous being that will miraculously solve every problem. It’s more of an amazing tool that democratizes, and above all reminds us of, the reality that these days you no longer have to do everything tediously by hand.
Where AI shines
Let’s walk through how we use artificial intelligence here at FRGTN. No fluff, just specifics.
Search
A well-worn use case by now, but deservedly so. Google isn’t what it used to be. The internet is bloated with an unbelievable amount of content, pages, and assorted other clutter. Finding anything on it is becoming an almost superhuman task.
Plenty of people already use AI for direct answers to questions, but if you take it one step further, a whole range of new possibilities opens up. A simple prompt: “Please find me a few sources on the topic of the new 2026 law on electronic sales records.” can on its own hand you the actual sources you can use to look the information up yourself. Suddenly you don’t have to blindly trust a short answer.
But we don’t have to stop there. Adding the prompt “Please rate the credibility of the sources based on the reputation of the sites they come from, and sort them from most to least trustworthy.” can enrich the results with useful data that lets you reason about the topic more precisely.
And to finish: “Now take the most and least trustworthy sources, compare them against each other, and write up where their information differs and diverges.” Suddenly you have a quick overview of the topic and of the public’s perspective, while also understanding which pieces of information might not be the most accurate and where there’s room for closer investigation.
If a topic truly matters, don’t be afraid to go deep. AI can save you time and organize your approach to the topic in a matter of minutes.
Documentation and problem solving
Very similar to search is exploring documentation and live support. With a simple connection to your internal documentation, AI can become a very powerful support for your team. Suddenly there’s no need to spend half an hour searching where in the docs your problem is addressed. Instead I send a screenshot and a description of the problem into the conversation, and the AI searches, compares, and then offers a solution or a place in the documentation.
A great tool that can come in handy for anyone. It often happens that a problem appears in multiple places in the documentation; with AI’s help we can get a single, unified record of where to find everything, saving us a lot of time.
Here you do need a bit of setup and preparation, but for us it absolutely pays off.
Quick data visualizations
These days AI handles writing short programs very well. Combined with a web browser, it can prepare overviews, visualizations, and other aids for working with data.
It has to be said, though, that these quick tools are more of a stopgap, just to give you an idea. For full-fledged data visualization it’s always better to rely on real tools and professionals. But for a quick summary and a better sense of things, AI can be put to use.
It’s often handy to quickly line up costs over time and find out where the money is leaking. AI is very good at navigating data and spotting anomalies and the patterns that emerge from them.
If you find something, it’s then not hard to go to the assigned team and ask them for a detailed review. This way you can at least identify a problem if you suspect one, and get your bearings in your data faster.
How not to use AI
Sometimes we might think that thanks to artificial intelligence we can do things blazingly fast and don’t have to give them as much time. We don’t have time to dawdle, after all, and things have to move quickly.
This is the road to hell. It often happens that higher management positions misread the power of AI and then demand impossible tasks from their teams.
A quality product takes time and care. If you sweep things under the rug and your solution is a generated web page that looks like every other AI slop someone spat out from the same mold, you’re digging your own grave.
Use AI as a tool, not as some cheap employee. By its very design it remembers little on its own in the next conversation in a new window. So it drags you into an endless loop of recurring problems and mistakes.
If you take one thing away from this article, please let it be this:
- A simple rule: if a task is short and only needs to be done once, not repeatedly, AI is probably a good tool for it. But if it’s long, requires several iterations, and is waiting for you again next week, you’re better off solving the problem with proper automation, or handing it off to someone.
- AI is a great filter. People who want to cut corners and live by the rule “Better than nothing.” will swarm to it like flies to dung. For those of us who think our product is worth investing some time and energy into, all that’s left is to watch where we step.