We're ruining AI by using it the way we are.

You know that we can all tell that you're using AI for your posts, right??

And when it comes to safety, our limbic system really kicks in. It's the part of the brain involved in our behavioural and emotional responses, especially when it comes to behaviours we need for survival: feeding, reproduction and caring for our young, and fight or flight responses.

When we feel unsafe, that system sparks up like Christmas lights. And I'm going to go out on a limb here and suggest that the same system fires up when we notice something not quite right about something we're reading online.?

To test this theory out, I decided to measure my own responses.

I did a test on myself, and it was very eye-opening.

I took note of my own responses over about six weeks this year when it came to reading things in my inbox that seemed to be scams.?

Thanks to my wearable fitness smart watch, I could see that my heart rate would increase an additional ten beats per minute when I had read a scammy email or saw a scammy ad online.??

My oxygen saturation level would also increase from 95 to 97%

I was having a stress response when I'd read these emails and ads.

But here's the kicker, I'd have the same response when I read a post on LinkedIn that had clearly been written using AI.

The three cases I used were:

1. A person who had recently taken up a new job with a career change into sales. Her posts had gone from selfies and inspiration to a sudden pivot towards emoji-filled informational posts about her new employer with her talking about things she has never had any experience in doing.
2. A business owner who only ever posts promotional posts about his business in correct English that is too correct, if you know what I mean. I know this guy, and this isn't how he speaks or writes an email.
3. Another individual who recently positioned herself as a LinkedIn expert and coach, but who has never really had much engagement or momentum on LinkedIn or as a coach.

With all three of them, I'd read their posts and have the same measurable response. Their posts would stress me out. And in every case, they were clearly using AI to write their posts.

I know all three of them well enough to know that this wasn't them talking. And that cognitive dissonance was giving me "the ick."

For all their sales messaging and efforts to promote whatever they were promoting, it didn't match who they were in real life, so it not only read like a copy and paste from ChatGPT, but it would cause an actual stress response.

Which I'm pretty sure is not what any of them intended.

And this is happening all across the web and social media right now.

You'll see it in those blog posts and articles that have a distinct pattern - words you know that no actual person would use. Around twenty words per sentence. Three sentences per paragraph. Three paragraphs per main point. No dot points. Generic information that contains no real world examples or case studies - or if it does, those examples are completely anonymised so that you know immediately that they're not real.

You'll also see it in social media posts.

A first line that reads like a headline, accompanied by an emoji. Three short paragraphs of highly generic information. A question to spark engagement at the end. Three hashtags.

It's a set of patterns you can't unsee once you know what you're looking at.

And it's getting worse.

We are using ChatGPT and other Ai writing tools to do the heavy lifting for us, but then we just copy and paste the results to save time.

What we don't realise we're doing, is causing a stress response in people that gives them enough of an aversion to you that you annihilate all trust they had in you, and you will unlikely ever get that trust back.

After all, you just presented yourself as a fake, a fraud and a danger to them.

Not a great first, second or forty-fifth impression.

If this is the wrong way to use AI, what is the right way?

AI tools work best when used like a watch.?

You don't make decisions based off a watch. You use the information on your watch to inform you of the time, which then helps you make decisions about what to do next.

AI is a tool just like a watch is.

When it spits out answers to what you want to write about, that's just an outline of what to do next - it's not what you actually put in a post or article. Especially in the case of the free version of ChatGPT which has a terrible habit of just making up answers that are often grossly wrong at worst and obviously generic at best.

The result in an AI writing tool is the start. It lets you skip the generation of ideas and the creative block - and it gives you a first draft. NOT the final version.

You take that first draft and personalise and humanise it.

You add real world examples and case studies. You remove the weird words and replace them with what you'd normally use in conversation.?

You take the robotically consistent sentences and paragraphs add in that burstiness that people write with - where we might write a five word sentence, but follow that up with a 50 word one.?

It adds feeling, pace and excitement.

AI tools are very, very good. I use them every day to help me reduce the amount of time I spend researching, generating ideas and writing first versions of things. ?

But there's no way in hell that I'd let them loose on this newsletter of mine.

You'd see straight through it.

And I'd struggle to deal with the guilt of causing all those elevated heart rates.

That's all for this weekend. Just one short read about how things aren't what they seem and people aren't what they are on social media.

If you're not getting value out of these tips, please consider unsubscribing.

I won't mind and there are no hard feelings.

And if you are enjoying this newsletter, the best compliment you could pay me would be to share it with one person who you think would benefit from it.

See you again next week.

Cheers,

Dante

P.S. Here are a few free ways I can lend a hand if you need some help.

On Monday April 29 I'm running a small group on Mastering Meta Ads for Facebook & Instagram in Darwin. As one of just five people in Australia trained by Meta to deliver this, it's worth attending. Book your place here.

My next Saturday public speaking workshop is being held on May 4. We'll be looking at how to plan, practice and deliver your first keynote speech. It's free and very interactive, so book your in-person place here.

If you've been thinking about how to present yourself online with more of a strategy and plan, my free Personal Brand mini course will get you started. It's delivered once a week over five weeks to your email. Get it here.