This article will shed light on the kind of support AI offers in the daily work of a Web Developer.

The revolution is underway.

That a revolution is underway due to all AI-based tools is certain, and much has been written about it.

Let me give you an overview of how AI works as a web developer’s assistant in day-to-day work. If you are new to this topic, I will familiarize you with it, and if you use this tool in your work, you certainly know what I am talking about.

We, developers of all kinds, yes those at The Humans too, are often faced with problems so unusual, that it can take us hours, even days to find a solution. We are no strangers to frustration and fatigue, and if you think you’re the only one experiencing this, believe me, you’re not. Despite being quite at ease daily with the technology in question, a tripping point can appear out of nowhere. And it rarely concerns something difficult. I’m guessing you’re familiar with situations where you reach for a calculator when quickly calculating the width of a 960px container divided into 5 columns. Analysing things for a couple of hours gets to the point where we’ve had enough, and you are getting nowhere.

How do you deal with this?

In early Stone Age (like 5 years ago), the rubber duck method was popular. You know what this is, don’t you? If you don’t, I’ll remind you: you put a rubber duck in front of your monitor and explain to it one by one what’s going on in your code and what doesn’t work in it. If the situation is complicated, you’ll have a chance to have a quality talk with the rubber animal, but the technique isn’t as silly as it sounds.
This is because while explaining, you’ll be more likely to locate and remove the cause of your troubles.

This was the way in the prehistory.

Now, at least when I am concerned, the duck got lost somewhere (which is a shame, because it was cool).

Today, there is a deus ex machina solution in the form of assistance from AI. You can call it whatever you like. Someone outside the industry might call it playing with codes. But it works a little differently.

When you run out of ideas to eliminate 'headers already sent’, or your WordPress multisite returns a 500 error and the database query doesn’t return results, try specifically describing the issue to the AI chat.

Conversing with a chat is no different from talking to a human. To get prompts, you need to be precise about the intended outcome you hope for and provide any possible input to create a satisfactory response.

What is common to a human interlocutor and a chatbot is the possibility of making a mistake at some point.

It is not perfect. Quite often, it turns out that the “ready-made” code we get in response will not work. And this can sometimes be seen at first glance, before clicking “copy code”, and sometimes it only turns out “in the wash”. Further conversation, clarifying the issue or asking for a different solution to be proposed very often solves the problem and will save you time and nerves.

While we all know that security is extremely important, I will remind you of good practices when using AI chatbots. When requiring help with a script, do not include real login details, domains or URLs containing sensitive addresses. For such a conversation, I recommend substituting example data foo bar or lorem ipsum.

When receiving code from a chatbot, take the time to understand the comments within it.

When receiving code from a chatbot, take the time to understand the comments within it. A very good thing is that, in addition to the verbal description in the AI-presented response, the code is enriched with comments with information on what the use of a function or parameter implies. This not only allows us to learn how our code works, but also how to develop it if necessary or apply it in another application.

Keep in mind that you should always take a sensible approach to what you get from AI. Even when you start to feel comfortable with your favourite chatbot, I recommend relying on your experience and on knowledge, which is good to develop by coming to solutions yourself. After all, we want to be better and better specialists, right?

Although it’s convenient to outsource the writing of a blog post to a chatbot, it’s a good idea not to fall into the trap of getting lazy. This text was written independently 🙂 If you liked it, please share it online.