Or, how to stop worrying and love the singularity
It’s 2023, and I, like every other writer out there, have played with ChatGPT. Under the guise of a cynical ‘stick-poking’ (that was, in reality, hiding my acute, gut-withering fear that a robot finally has taken over the task I not only am able to do, but actually enjoy doing), I kicked the tires, prodded around under the hood, and generally hoped to find major problems.
It became distracting, results were mixed, and ultimately I can rest easier. Because, as I say ALL THE TIME (I probably say it in my sleep): authenticity will win.
But that doesn’t stop people from trying to utilise different tools to make the job easier. In fact, ChatGPT is far from the only AI tool helping writers today. Grammarly and Yoast both rely heavily on AI to help writers get the most from their own natural abilities. And it’s not like we don’t all use them (hey, Bertha) as we mine for gold in our oh, so talented brains.
And there are dozens of ways to enlist the help of a machine. Aside from the actual text, there are a plethora of questionably useful tools to analyse the nittiest-grittiest parts of your site. But each one is very specific, and usually extremely limited in its scope. But, as with all tools, you’ll only get the best out of it when you know what it’s for, and what its limitations are.
My current favourites are headline optimisation analysis tools.
Can you optimise authenticity? Sure!
Can you be authentic within optimisation? Probably.
Are those the same things? Who knows?
Does it matter? Not for the purposes of this blog.
If you want to make sure that your already strong and relevant headlines are also ‘optimised,’ I strongly recommend that you do what I did: play with the tool for a while (yippee!) and then use common sense to apply it to what you already know works. And I’ll explain why below.
But first, know that when I teased a headline optimiser analysis tool I found two things:
1. It’s fun.
2. They reward clickbait-y titles.
Both of those should have been obvious to me, but sometimes we have to try things out for ourselves.
I tried inputting different headlines with varying degrees of worrying, emotional, and unnatural language just to see what would happen. And my findings help me sleep at night.
AI is still more A than I.
For kicks, here are some interesting responses to made-up headlines with optimisation results.
I started with what I thought was neutral.
Turns out, the tool agreed, giving me an underwhelming “0.00%” optimised score.
So I jazzed it up by making it a ‘how to’. “How to live your best life” was 16% optimised, so now we know that the ‘how to’s are going to do well.
My next question was, does ‘how to’ trump sense? Only one way to find out:
Shoelace efficiency, never a hugely interesting topic, somehow more spiritual than living your best life. Was it the ‘care’ that did it?
I wondered if there was a way to up the ‘how to’ factor. Knowing how much people love titling Ultimate Guides, my next try was in that vein:
While I was typing this one I giggled, but only for a second, because I realised it probably exists on someone’s webpage. Sorry.
Basically, I went on for awhile doing this. Here are some more of my super silly title tests.
I think you can see why by the end of my AI-baiting session, I was both relieved and disappointed.
The relief comes from knowing that for all its hype, AI still doesn’t have a good grasp of human emotion and how to evoke it. It leaves a lot of room for real human writers to keep plugging away with what they do best, and leave the worry for another decade (hopefully).
My disappointment comes from realising that this cold, hard, robot brain isn’t yet ready to be my best friend, perfect partner, or small-dog replacement for when my inevitable tussle with empty nest syndrome comes around.
In the meantime, at least I can still enjoy knowing that when I want to write an intellectual and spiritual title, I’ve got the skills. Receipt:
I don’t mean to brag, but I managed to get a ‘how to,’ a ‘favourite,’ and a beloved family member in there.