Apparently, AI is ruining social media.
Everything sounds the same. They’re all filling up with AI generated social media content. Captions are becoming predictable. Thought leadership is starting to feel suspiciously like the same thought being passed around by 400 different people.
And yes… There’s definitely some truth to that.
But we’re not letting humans off that easily! Boring content existed long before ChatGPT came along.
AI might have made it considerably easier to produce it, but perhaps the problem isn’t that we’ve suddenly got machines writing our content. Maybe we’ve forgotten to have something genuinely worth saying in the first place?
Let’s get one thing out of the way. We use AI.
There we said it.
We use it for research, we use it to bounce ideas around. We use it to interrogate things, organise thoughts, look at different angles and occasionally help us find the words for something that’s been rattling around our brains for far too long.
And we’re not about to tell businesses that using AI generated content automatically makes their marketing bad. It doesn’t.
Even Google isn’t inherently against AI-generated content. It’s guidance focuses on whether content is accurate, relevant, useful and created to add value, rather than simply whether AI had a hand in producing it. Using generative AI to churn out pages without adding value, however, can fall foul to its scaled content abuse policies. (You can find out more here)
LinkedIn is taking a remarkably similar position.
This year, it started actively tackling what it calls ‘AI slop’, low-effort content that might look polished but lacks a clear perspective, substance or anything particularly original. LinkedIn says content that appears AI-generated and lacks perspective is less likely to be distributed beyond someone’s immediate network.
In September of this year, LinkedIn said its efforts had already reduced views of content it classifies as AI slop by 40%. More than one million members had also used its new ‘seems like AI slop’ feedback option within the first two weeks of it’s launch! (Stats came from here!)
But Google and LinkedIn aren’t saying ‘Don’t use AI’. The problem isn’t the tool itself, it’s how people are using it to create things.
You know the stuff…
‘We’re delighted to announce…’
‘Quality is at the heart of everything we do…’
‘Our dedicated team always goes the extra mile..’
‘Here at [insert company name here] we pride ourselves on…’
Nobody needed AI for any of that. We managed just fine coming up with these things, as boring as they were! The difference now is scale. Something that once required a person to sit down and write can now be generated in a few seconds.
That means the barrier to creating content has practically disappeared. Which sounds brilliant, until everybody starts creating more of it, without adding their own personalities, stories, etc resulting in everybody sounding the same.
This is where things get interesting…
For years, businesses have been told:
And now we’ve handed everyone a machine capable of helping them do all those things, way faster. So theoretically, marketing should be easier than ever.
…Except there’s one small problem.
Your customers haven’t suddenly developed more hours in the day to consume all of it.
We’re producing more content for roughly the same pair of eyeballs. And if creating something is becoming easier, simply creating something becomes less impressive. The value inevitably shifts somewhere else.
Somewhere along the way, having an opinion became a content formula. There’s practically a recipe for it now.
Except, what was the thought? This part seems to have gone missing.
Generative AI learns patterns, from enormous amounts of existing material. So when AI generated social media content starts sounding formulaic, we should probably acknowledge that humans created quite a lot of those formulas first.
We decided what a social media post was supposed to sound like. We copied hooks that performed well, adopted the same storytelling structures, started writing in ways designed to encourage engagement, and then fed an enormous amount of that content into the digital hemisphere. And all of a sudden we’re now surprised that AI can reproduce that pattern. It’s like teaching someone your secret recipe and then complaining when they use it and it tastes familiar!
Let’s strip ‘thought leaderships’ marketing jargon shall we?
You don’t need to call yourself a thought leader. You need ACTUAL thoughts!
Good B2B thought leadership usually starts with something a person genuinely knows, believes, has experienced or noticed.
That might come from a director who has watched their industry change over the last 20 years. It might come from an engineer who keeps seeing customers make the same mistake. Or a salesperson who hears the same objective every week. A marketer questioning advice that everyone else seems to repeat without thinking about it.
That’s the raw material. The content should come afterwards.
We come across this quite a lot with our own clients at Brainstorm.
A client will tell us they want to push a particular product or service. Which is fine, that’s what we’re here for, but sometimes that’s pretty much all the information we’re given. And while we could quite easily ask AI to write us ten social media posts about that product or service, what would we actually be telling people?
That the product exists? That it’s available? That the business would quite like people to buy it?
There’s only so much you can say about something when you don’t fully understand it yourself. So we have to dig a little deeper. We start asking questions about the product, the process behind it, what customers use it for, why they need it, and what makes it different. Sometimes that means speaking to the people manufacturing it, understanding the intricacies of how something works or finding out what problems customers have been having that this service or product could help solve.
And quite often, that’s when we find something worth talking about. It might be part of the manufacturing process that most people never get to see, a problem that’s been solved in a way we hadn’t considered or a detail that the client hasn’t thought to mention before. But to someone outside the business, that information could be exactly what makes the product interesting.
AI can help us turn what we’ve learned into a concept to give our topic more legs, explore different approaches or to make sense of our scribbles, our broken sentences and give us a basis to work with. But if we’d skipped the questions and gone straight to generating posts, we’d have missed the very things that make the content worth creating in the first place.
Well, we can’t exactly put the genie back in the bottle now can we? So, no.
AI can be incredibly useful. Use it to research, organise your thinking, challenge your argument, identify questions you haven’t answered, help structure a blog, turn a rambling voice note into something coherent, explore different angles, interrogate the data, repurpose something you’ve already created.
But there’s a difference between ‘Help me communicate this idea.’ and ‘Give me an idea’
And an even bigger difference between that and ‘Write me something to post today.’
If you’re giving AI absolutely nothing to work with except your industry and the fact that it’s Tuesday, don’t be surprised when it gives you something everybody else could have posted too.
Start before the prompt! Before asking AI to create anything, ask yourself these questions:
That’s where the useful content starts to form.
The same applies if you’re working with a marketing agency, freelancer, or internal marketing team. Don’t just ask them to ‘do some thought leadership.’ Give them access to the people who actually have the thoughts.
It means we probably need to stop treating SEO, social media, content marketing and AI visibility as completely separate little boxes. Google still says traditional SEO best practices are still a foundation for visibility in its generative AI search features. (If you want some proof, check this out)
You still need technically accessible pages, clear language, useful headings, relevant search terms, internal linking, properly structured websites. But no matter the optimisation, you can’t rescue content that doesn’t offer anything useful.
Google’s people-first guidance explicitly asks whether content demonstrates first-hand expertise and whether someone leaves having learned enough to achieve their goal. (you can check it yourself here)
So yes, we’ll put a keyword in this blog. We’ll write a sensible SEO title, structure it so Google and AI systems can understand what we’re talking about. We’ll also answer questions people might genuinely search for. But there still needs to be something underneath all of that worth finding. Otherwise, what are we optimising?
We all want to be more productive, AI helps us do that.
But somewhere along the way, have we started losing our humour, our stories, and the little things that make us…well, us?
Have we become lazy because we can type something into AI and get an answer in 60 seconds? Or is it simply that we haven’t quite figured out how to get the best out of it yet, so we ask it basic questions, take whatever it throws back at us like it’s gospel?
And then there’s the whole conversation around authenticity!
People are craving it. Businesses are being told to be more human, more relatable, more real.
But isn’t it funny that we’ve reached a point where we’re now having to prove we’re being authentic? We’re questioning whether someone’s actually written their own social media posts. Whether that photo is real, or if the person talking in the video is a real person. We’re even starting to change the way we write because apparently certain words, sentence structures and punctuation are now enough to make someone say “That’s definitely AI.”
So now we’re deliberately trying to sound less like AI, which has learned how to sound like us, while we’re trying to sound more like ourselves… Does anyone else see how ridiculous this is getting?
And where does that leave authenticity? Surely being authentic isn’t about proving you haven’t used a particular tool, but whether what you’re saying actually comes from you. Your experiences, opinions, humour, your way of looking at things, the stuff you believe in, disagree with, and the things you’ve learned along the way.
If AI helps you put those thoughts into words, does that suddenly make them any less you?
If you’re asking AI to come up with the thoughts, opinions, personality AND the words, then what part of that content is actually coming from you or your business?
Use the tools, keep the personality, but for goodness’ sake, don’t outsource your entire brain!
Are company social media pages becoming redundant? It's a fair question. Open LinkedIn and some…
"Just be consistent." There it is. The answer to almost every social media problem on…
Remember when everyone said websites were dead? That social media was supposedly going to replace…
Let's start with an uncomfortable truth... Your customers probably aren't comparing you to the business…
Somewhere along the way, 'being visible' turned into 'being everywhere.' LinkedIn? You should be on…
There's a version of marketing that exists in people's heads. It looks simple enough. Post…