This article is a bit different from the usual ones — the goal is to share a personal reflection from my own perspective.
How has AI evolved over the years? Let me briefly tell you about my experience.
In early 2023, I started using ChatGPT for the first time, driven by curiosity after reading a few news articles. I had already heard something about its development in previous years, but I hadn’t really paid much attention, nor did I fully understand the potential of this technology.
I began using the free plan and wrote a few prompts. I must admit, the results left me quite amazed — not only were most of the answers fairly accurate, but the tasks I asked it to perform were surprisingly precise and detailed. Of course, the model sometimes made mistakes (and it still does today), but as long as you double-check its content, it’s manageable.
In the following years, OpenAI’s service evolved at a remarkable pace — if we compare it with the development timeline of other computing technologies, its progress has been incredibly fast. By early 2024 (maybe even sooner), we could already upload files and access the web for more accurate answers. Today, the model has advanced so much that all those features are almost automatic, and several new ones have been added — all of this in just a little over two years.
Let’s compare OpenAI’s evolution to the evolution of video game graphics, from the late 1980s to today. How long did it take to move from 2D pixel art — often with hand-drawn backgrounds — to today’s hyper-realistic graphics and powerful GPUs? That took at least a couple of gamer generations.
AI is moving much faster — proof that it’s a constantly evolving and self-learning technology.
Another AI worth mentioning is Midjourney. I tested it in 2024, when it could only generate still images. Today, it can animate both generated pictures and our own photos if we wish — and its development won’t stop there. Other AIs, like Runway and Luma AI, are even more advanced when it comes to video generation.
We’ve talked about text generation, data analysis, and video content, but AI doesn’t stop there. In the audio field, cloning a voice is no longer a challenge — nor is creating one from scratch or composing music entirely with AI.
The question I keep asking myself is: at this pace, what will be left for us?
This isn’t the end. Once errors and imperfections are corrected, in a few years AI will be able to do exactly what we humans do with our biological brains — or perhaps even better.
Will some professions disappear? Will others be downsized? Will governments be forced to regulate or even restrict AI use? Who knows.
In my opinion, AI is an incredibly useful tool in many aspects of our lives — but in others, it can cause serious harm if left completely unchecked.
We’ll see how things unfold.
Thanks to everyone who took the time to read my reflection.
Jack – The Skills Factory Webmaster
