This is discussed all over the place right now: since articifial intelligence, namely the recently-released-publicly ChatGPT service, writes fantastic content almost automatically, should human writers still bother with it? It’s an exciting question and this post won’t attempt to answer it fully.

A general thought: lots of internet content is… let’s say repetitive. I might use stronger language to describe it, since it’s basically loads of rip-off nonsense that has been regurgitated endlessly by people trying to make money from the ads displayed on their fake content sites. Internet search is not broken, but it drowns in the amount of plagiarized content.

If this type of content is what you create, ChatGPT is a great thing. Please go and use it and stop reading right now.

Still here? Glad to see that!

On the other hand, writers such as myself attempt to be more creative when we write. We want something interesting, something potentially unique, to be the subject or main focus of an article we publish. How can we create all these new facts worth writing about, facts that haven’t been described before? Are we going to create alternative facts?

No, of course not. What we plan to do is think. We plan to take what we know, research a bit more about it, view it through our skewed glasses of experience, see it in this and that light, and … form opinions.

Opinions are fantastic. Not all are equally good, and we’ve all been part of one of these discussions where a sadly uninformed person states you can’t disagree with me, this is just my opinion! and thereby ends all useful conversation. I’m joking when I say they are not equally good — they are, in a way, but there are rules that determine whether an opinion you hear or read about is well-formed, well-supported, etc etc… a lot of this happens subconsciously.

Anyway, the point is…

Thinkers have opinions, and that’s what they can write about. They don’t oppose facts — or at least they shouldn’t — but they are often more interesting than the facts themselves. An AI system may one day be able to form opinions, or perhaps this is already possible today… I’m not an expert on AI as such. But whenever it happens, we will need to judge the quality of these opinions in comparison to others, formed by others, human or not.

Meanwhile, I plan to write about opinions, and I also plan to keep using my own perspectives, my own experience, to create content that explains and teaches, describes and illustrates. In a few weeks’ time, I’ll do my bi-annual “perspectives for software architects and developers” workshop again, an event that is based almost entirely on my “take” and my ability to explain, and that has been running successfully since 2015.

The day may come that AI replaces me completely, but I believe that as long as I keep thinking, I’ll compete.

To close things off, I thought I should ask ChatGPT what it “thought”. What can I say, it’s humble… glad to see that AI doesn’t (yet) disagree with me.

Humble ChatGPT