Why reliance on artificial intelligence is killing your content
Personal anecdote time.
Have you ever used Originality.ai or another AI analysis tool to see whether a writer has used artificial intelligence in producing content?
It doesn’t work.
Many of my contacts in the industry have been judged against this tool by, rather ironically, the clients that don’t want to use AI-driven content. Props to them – there’s a reason why fully AI-driven content is a bad idea, and we’ll get onto that.
But it has cost colleagues their clients. Purely because clients are putting so much trust into AI in one way or another.
My experience with AI content detection software
Until recently, I hadn’t come across these tools myself. That is, until I produced some ad-hoc work for a new client in financial services. The agreement? Each short-form article would be assessed against Originality’s AI content detection tool.
My first pass scored 99% human, allaying my fears somewhat.
My second article – 90%.
On my third article, my client was told that there was a 55% chance AI had written my article.
I’d done nothing differently. I’d used the same keyword research tools. I’d followed my usual style of writing. There was nothing I could’ve done differently to produce a different piece of work.
And honestly?
I was irritated. This doesn’t really convey how I felt, but I like to think that this is a professional blog, so let’s leave it at that.
I’ve been doing this for ten years, and I’ve been watching AI kill off numerous sources of work – particularly those websites that serve people who want to get into this industry but need to build a portfolio. The content mills. Now, it’s telling prospective clients that their writers are “cheating",” and clients are taking that as gospel.
But that’s just one of the problems – from a writer’s perspective.
The real problem is much more serious.
AI has begun to cannibalise itself
Do you remember being at university and completing your first long-form essays? You were just a student, not an experienced researcher. Your arguments were made off the back of existing research into complicated subjects, be that politics, science, religion, history, and so on.
So, what did you do to avoid getting chucked out of university for plagiarism?
You found new takes on the subject matter. You quoted, but you paraphrased, and you wrote things with your own distinct and unique voice.
Crucially, you checked your facts, because an essay that used incorrect or entirely fictional facts was going to get you a resit at best or thrown off the course at worst.
Artificial intelligence cannot do that reliably.
AI tools are scouring the web for existing content from which to learn. What happens when that content has already been indexed, digested and assimilated into the AI’s lexicon?
It begins to cannibalise itself.
AI tools begin scouring AI content, which is flooding the internet at an alarming rate. This degrades the quality of the writing and risks introducing factual inaccuracies.
This is a problem for more than one reason:
Artificial intelligence is often confidently incorrect. While Google Gemini just landed, their first attempt at AI made factual inaccuracies in its first demo.
Microsoft’s Bing AI had a similar SNAFU moment in its first demo.
Students have begun using AI to take shortcuts in their work, potentially missing out on crucial learning points.
Lastly, and this is a huge one folks – a US lawyer referenced multiple fictional cases to set precedent in an argument in court.
We’re in unusual times right now. Like any new technology, we’re going through the boom of excitement whereby every company and its dog wants to get on the gravy train and reap its rewards.
But we should be moving forward with caution. And personally, I think we should be using AI to focus on the more important applications – as a supportive tool in areas such as medical research and diagnostics, among others.
For now, though, I just hope that clients begin to realise that there is still immense value in the human writer. Artificial intelligence should be used as a tool to optimise content creation as opposed to trying to replace people who have infinitely more ingenuity.
And do you know what I find truly hilarious?
I’m willing to bet an AI detection tool will tell me that this blog post was written by an artificial intelligence.