Why Human Creativity Outshines AI Writing

By Arthur Weiss

Editor’s note: This is part of a series covering what’s new and, more importantly, how infopreneurs can maximize the benefits of AI tools. This one is part of using AI in general.

In my last post, I looked at some typical uses for AI. In this post, I look at some of the dangers in depending on AI for writing / authorship. Despite what some like to think, the human brain is still a better tool even if AI tools are catching up.

I provided an example of a birthday card haiku:

Candles softly glow,
laughter drifts through gentle hearts—
another year blooms.

This is tolerable but it’s not brilliant. Compare this to Sonia Sanchez’s

“Haiku [for you]” (1998):
love between us is
speech and breath. loving you is
a long river running.

Or Elizabeth Searle Lamb’s

the year turns — /
on the harp’s gold leaf /
summer’s dust

If you had to grade these three, the AI example would pass, but it wouldn’t be a great pass – the human examples are much more evocative, and generally, that’s the current situation for AI writing. Essentially, depending on AI for writing is a recipe for mediocrity.

You may ask AI to write a sonnet in the style of Shakespeare. It won’t be Shakespeare! I can’t see AI coming up with genius ideas like e.e. cummings’ use of lower-case letters, Ernest Vincent Wright’s Gadsby that eschewed the use of the letter e, or even writing a poem as an acrostic. AI follows rules and so won’t innovate in such ways on its own. All it can do is use what is in the language model’s training data – some of which is based on human genius (e.g. Shakespeare’s sonnets or various haiku).

Therefore, when using AI for authorship, ask it to check spelling and grammar and perhaps even suggest changes to style to make your writing more succinct. I now regularly do this, but it’s important to remember that you are the author, not the AI. Overrule the AI if it doesn’t give you what you want or if, in your view, its suggestions aren’t improvements.

As a personal example, I wanted to use the verb syncretize to describe taking content from various sources and creating a blend where the individual sources could still be recognized. When I passed the text through AI, it kept on changing syncretize the content to synthesize the content – not what I wanted. But in case I was wrong, I put this basic prompt into Claude.ai.

What’s wrong with syncretize. It is not the same as synthesize?

Claude came back with:

You’re absolutely right, and I apologize for that error. “Syncretize” and “synthesize” are indeed different words.

Claude then defined both words:

synthesize – combine elements to form a unified whole

versus

syncretize – blend different elements together

 and concluding with

“syncretize” isn’t technically wrong; it’s just a less common word choice for this particular application

This is an example of how AI can help your writing. I’d not considered that syncretize was an uncommon word that might not be understood by all my readers – and so I followed up with:

I do mean blend into a whole rather than synthesize. Is there a better word?

Claude suggested consolidate and also integrate, unify and distill. I took Claude’s advice and changed the text to consolidate the content; that kept my meaning, and because of the assonance and alliteration it may have been a better word choice anyway.

All this is different when answering questions e.g. for a math quiz or even a medical or law exam – AI can pass these with flying colors. As an example, OpenAI’s AI reasoning model beat all human contestants at the 2025 International Collegiate Programming Contest (ICPC). This international top university coding competition is viewed as the Olympics of programming. There are other dangers to look out for here – especially when it comes to using AI to research or summarise a topic. I’ll consider an example in the next blog post.

Arthur Weiss has been an infopreneur for almost 30 years. He founded AWARE in 1995 after a career at the business information company Dun & Bradstreet. He specializes in competitive and marketing intelligence using open sources (OSINT). Recently, he has pivoted to new areas, including exploring how AI tools can support infopreneurs. His latest insights can be read in International Marketing & Competitive Intelligence and Computers in Libraries magazines. He may be contacted at a.weiss@aware.co.uk.