Each of us today has the opportunity to benefit in various ways from the vast amount of free knowledge available on the Internet. Despite this, many people use only a small portion of this technology.

With artificial intelligence, more or less the same pattern is repeating: some people fully exploit its potential, while others limit themselves to asking simple questions to ChatGPT, much like what happened with Google.

We have in our hands a tool with enormous potential: generation of images, videos, software, and music. Activities that previously could only be performed by professionals are now accessible to anyone, making the process more “democratic.” The only real limit is imagination.

Progress in this field is exponential, and with each new update, increasingly extraordinary possibilities open up.

How I use AI

Since the early days of ChatGPT, I have had the opportunity to observe the evolution of the models. The technological leap has been impressive. What previously required numerous human-machine iterations now, in most cases, requires only a single request to obtain a perfectly coherent result.

Naturally, I have refined my prompting technique, but it is clear that the latest available models seem to immediately understand the user’s intent.

My first experience with vibe coding was the creation, still ongoing, of a web-based 2D strategy video game. In about two weeks, I managed to develop a map, the logic for placing structures, the resource accumulation mechanics, and the entire movement system using mouse and keyboard. Every milestone reached was a rush of adrenaline and amazement, not only for the result achieved but especially for the time spent. Without AI, the same work would probably have required at least a year of development.

Currently, for this video game, the entire graphic component still remains, which I am not yet able to replicate as I would like with AI (also because I do not have deep graphic design skills). This aspect requires much more time compared to simply generating code, so the project is progressing slowly, but it remains an excellent training ground for refining technique.

The project that made me reflect the most, however, was the development of a SaaS application as an internal tool for my company. I managed to replicate most of the functionalities of similar paid tools available on the market. Time spent? One weekend. Additional features? Ten minutes in the evening.

Minimal mental effort, maximum result. And this changes everything.

The market is full of SaaS software that, through subscriptions, allows the use of platforms that, in most cases, can be replicated quickly. This is not just my personal opinion: it is something that is already happening.

Do you work for a company? This could be the opportunity to save your employer a lot of money and earn a salary increase. Or you can create your own company with a very small team while generating revenue like a much larger organization.

Obviously, what you can do, others can do as well. However, returning to the initial point, not everyone is using AI in depth and, above all, not everyone is a programmer. Technical skills remain fundamental in order to develop software and guide AI in the right direction.

That said, some clarification is needed.

Limitations and Risks of AI

My experience has also shown me the critical issues with these tools: it is necessary to pay very close attention to what AI generates. AI tends to comply with our requests and tries in every possible way to provide the requested solution, even at the cost of neglecting fundamental aspects such as security.

It has often happened that I rejected proposed solutions due to potential vulnerabilities. An experienced programmer, equipped with critical thinking, can avoid passively accepting everything AI proposes. Beginners, or those outside the field, instead risk considering these solutions as automatically valid.

It is possible that in the future this weakness will also be reduced or eliminated, and it would not be surprising, but for now it is an aspect that must be considered very carefully.

The Superficial Programmer

Many entered the sector during the period of strong market demand for programmers. We can consider the interval between 2010 and 2021 as a particularly favorable era.

The famous bootcamps come to mind: real training and recruitment schools for companies, which today are facing major difficulties, shaped by a historical context that will probably not repeat itself.

AI will drastically reduce the need for roles that are limited to writing code without truly understanding what they are doing or without a genuine interest in the discipline.

What I Recommend

In this scenario, I believe it is essential for an intelligent programmer to focus on the fundamentals:

Curiosity is the fuel that allows exploration, maintaining an open mindset and, above all, connecting the dots. Many ideas are born precisely from the ability to create connections between different fields.

It is not enough to simply do your job from nine to five: it is necessary to dedicate time to continuous personal learning. Only in this way is it possible to become better programmers and maintain relevance in the job market.