My perspective on AI and generated writing

As you may know if you’ve been following the blog, over the past few days I did an experiment with AI to write an article in English titled A Guide to Composer Autoload. In this article, I will share my point of view on AI and the generated writing in that post.

My thoughts are varied, and some are contradictory, since AI performs well in several aspects.

As for whether AI will take the job of a technical writer? I can already confirm that it won’t.

Strong points

As the main strength of AI when writing a technical article, I can say that it provides many perspectives on what we want to write, which is a good thing.

Secondly, it generates a lot of data, and even though it might not be entirely reliable, it’s still work done. Therefore, you must be knowledgeable about the subject or review it using an official source.

This information can serve as a brainstorming session for what you have or want to write. There’s nothing wrong with AI generating text for us as long as we don’t use it directly and without checking

Weak points

AI tends to write articles in a sensationalist way, and you’ll need to mitigate that as much as possible. How much? A lot, I’d say, like those tabloid newspapers that try to evoke an emotional reaction from the reader.

It can make serious mistakes. Like any auto-generated content, AI can produce complete nonsense or combine two unrelated points. It might even claim that something exists when it actually doesn’t.

You have to recheck too much the content, you do not want content that say the same thing or confusing or difficult to read.

How can I leverage AI for generated writing?

What I can tell you is that with AI writing in these areas, the best approach is to let it generate content, see what titles it creates, and from there, start building your own content, starting with the titles.

Language Translation: Another thing that IA is really good is at creating translation, 90% of the translation is wonderful make by the IA.