Relax, ChatGPT is not going to steal your programmer job
Taking advantage of the hype that this subject has raised to bring my reflection that ChatGPT, or whatever AI it may be, is nnot going to steal your job.
You have probably already seen, read, or heard several opinions on this same subject that spread at high speed after OpenAI released ChatGPT like a bomb.
Me too. Some expressing conservative opinions and others expressing aggressive opinions. Mine is one of the conservative ones and I will tell you why.
All of these AIs are cool and interesting, but not self-sufficient — and they are far from it.
I can spend a good time talking to ChatGPT and asking random questions. Not only that! It has been quite useful during my workday to ask a doubt about some technology or an error I have been facing.
But to be honest, I can’t see much beyond that. AIs still make a lot of mistakes and can’t go much further than something specific. They struggle, I would even say they are not capable, of dealing with large product requirements, complex analysis, and significant coding efforts.
“CDome on! Just train more! Just increase the number of parameters…” — it could be a good argument. Is it really?
Maybe we are facing the biggest challenge only now, which is: the curve of the AI’s “intelligence” does not follow the curve of growth of the number of parameters.
This means that a model with 5 trillion parameters is not necessarily 5 times smarter than one trained with 1 trillion parameters.
Sam Altman, CEO of OpenAI, recently said that GPT-5 is not in training and “won’t for some time” given various reasons, including a very special one. He states that he sees “the end of an era” for the race to increase the number of parameters in models and no longer sees the number of parameters as a good performance indicator. “I think we’re at the end of the era where it’s gonna be these giant models, and we’ll make them better in other ways”.
Reading this, I believe the next challenge, and in my opinion the biggest for an even greater leap, will be to find solutions for even more efficient training that deals better with large product requirements, large project contexts, and complex and subjective analysis.
When that arrives, then I believe it will be plausible to replace you as a programmer. But this should not happen in the short or medium term.
For now, leverage of what the tool provides to become better. Because in the short term, I see good professionals using the tool productively and kicking your ass.
Being a good programmer is not just about programming. It’s also about being analytical, a product and business expert. There is a great work that needs to be done before typing an if-else.
Not only that, I believe that with these models being integrated with Excel, PowerBI, etc., and with the emergence of various plugins, it will reduce the demand for potential projects.
So what may happen is that the bar will rise for you — and maybe more.
That being said, don’t despair or get discouraged. Just don’t fall asleep at the wheel for the next few years. I believe we programmers will find more demands in the future — now I am saying in the short term.
Before I finishe, I would like to share one more prediction. In the long term, I still find it difficult for programmers to become extinct. What I believe may happen is that the programming language becomes even more high-level.
We had C and Assembly, which are examples of low-level languages. Today we have Python and JavaScript, which are examples of high-level languages. Perhaps what may happen in the future is that we will use natural language to program — and that’s okay.
Or maybe we will find intelligent frameworks that do exceptional work, and you will only have to develop the business value — and that’s okay.
You will still be needed. Again, just don’t fall asleep at the wheel.
Anyway, those are my predictions. And obviously, being predictions, they may all be wrong.
If you’ve enjoyed, consider buying me a coffee as a token of appreciation.
Thanks for reading!
