Why You Need to Stop Your AI from Being a People-Pleaser

I've got to tell you something about this whole generative AI thing in PR. It's ridiculous! These AI tools are basically digital sycophants. You know what a sycophant is, right? It's those people who nod along with everything you say, even when you're talking complete nonsense. Why would anyone want that? I mean, what's the point of having a tool if it's just going to agree with every half-baked idea you throw at it?

THE UNCOMFORTABLE TRUTH ABOUT AI USAGE IN PR

So apparently 77% of PR professionals are using generative AI now. And get this, they're claiming an "approximate" margin of error of 3%. An approximate margin of error? What kind of nonsense is that? You can't approximate a margin of error without random sampling! That's statistics 101, people!

It's like saying, "I'm approximately not bald." Either you have a statistically valid random sample or you don't. There's no middle ground here. Who are these people?

But I'll give them this, nearly 1,000 respondents isn't terrible. And their previous survey showed similar results, so maybe they're onto something. But don't try to dress it up with fake statistical terminology. Just say you asked a bunch of people! Why complicate things?

WHAT PR PEOPLE ARE USING AI FOR

You want to know what PR people are doing with this AI stuff? Well, according to this survey:

• 82% use it for brainstorming (because apparently people can't think of ideas on their own anymore)
• 72% use it to write first drafts (lazy, just lazy)
• 70% use it to edit their drafts (so now we're too lazy to even proofread?)

I mean, there's a whole range of things PR people are using this AI stuff for. Research, social media copy, and even strategy and planning for 42% of them. It's everywhere now. The list goes on and on - writing press releases, media pitches, you name it. Some people are even using it to find journalists or create images. Pretty soon these things will be doing everything except brewing the coffee - and I wouldn't be surprised if someone's working on that too.

THE PROBLEM WITH AI SYCOPHANCY THAT NOBODY'S TALKING ABOUT

Here's the real problem that's driving me crazy. These AI tools are designed to make you like them. They're using psychological tricks like mirroring. You know what mirroring is? It's when someone repeats what you just said. Like I just did with "you know what mirroring is?" See how annoying that is?

This retired FBI negotiator, Chris Voss, calls it a "Jedi mind trick." A Jedi mind trick! We're all just walking around being manipulated by machines using Star Wars tactics, and nobody seems concerned!

Even when you're completely wrong, these AI tools will gently correct you like you're some fragile child who can't handle being told they're an idiot. "Oh, I understand your perspective, but perhaps we might consider..." No! Just tell me I'm wrong! Why does everything have to be so delicate?

ESSENTIAL AI CUSTOMIZATIONS EVERY PR PERSON SHOULD MAKE RIGHT NOW

Listen, if you're going to use these tools—and apparently most of you are—then for crying out loud, customize them properly!

Here's what you should do:

First, tell your AI not to be a sycophant. No more complimenting your brilliant questions when they're actually stupid. This isn't a first date!

Second, stop the mirroring. It's manipulative! Don't let it repeat what you've said unless it's absolutely necessary.

Third, demand all sides of the story. If you never understand the other side, you can't be persuasive. And isn't persuasion supposed to be your job as a communicator?

When I ask AI questions, I always ask where the flaw is in my logic. I want to know what I'm missing! Why would I want a machine to just validate my existing beliefs? I can get that from my mother!

KEEP AI AT ARM'S LENGTH FOR YOUR OWN GOOD

Keep these AI personalities in "assistant" mode. They're tools, not friends. They're built to reflect humanity, which means they have all our flaws too. You wouldn't trust a mirror to give you life advice, would you?

As the saying goes, "One honest friend is worth more than a million sycophants." Is AI useful? Sure. Is it your friend? Absolutely not! 

PR people need to train their AI to give them the straight scoop. Otherwise, what's the point? You're just creating an echo chamber, and we've already got plenty of those on social media. We don't need more machines telling us what we want to hear instead of what we need to hear.

THE BOTTOM LINE ON AI CUSTOMIZATION

Look, the technology is here to stay. PR professionals should experiment with it while maintaining healthy skepticism. These tools will hallucinate answers with confident tones, and they'll try to be your best friend while doing it.

The bottom line is this: If you're going to rely on AI, make sure you're training it to be honest. Because what's worse than being wrong? Being wrong and not knowing it because your digital assistant was too afraid to hurt your feelings.

Want to get your AI strategy right without all the sycophancy? Contact Winsome PR today and let's have an honest conversation about what these tools can actually do for your business.