The internet will soon need warnings for AI-generated content

AI’s strength is automating high-volume, short-term marketing activity, which means social media could become a cesspool of synthetic content.

Neil Patel, the well-known digital marketer, posted a fantastically insane directive last week. Patel has been running the numbers on thousands of accounts across most of the major social media apps. Based on his analyses, he now knows the required posting frequencies to enable you – yes, you – to perfect optimal account growth.

On Instagram, for example, where Patel looked at 15,000 user accounts, his recommendation is to post four to seven stories a day, around three to seven reels every week, images two to four times a week, lives once or twice a week, a carousel every week and between nine and 38 comments every day. Got that?

“It may seem like a lot of work,” Patel noted in his post. “But it isn’t.” His main argument being that you can “repurpose your content” from your other platform activity to save time and effort. Because, of course, Instagram isn’t your only focus. Oh no. Patel has similar gargantuan recommendations for the likes of LinkedIn, TikTok and X too. Each of them similarly hyperactive and equally bananas.

 

I utilise a perceptual sieve for posts from Patel and other social media gurus that I call the ‘Vee Denominator’, or VD for short. It’s named after the ubiquitous Gary Vaynerchuk and, in a nutshell, acts as a bullshit filter. If some digital genius tells you to expect an ROI of 1,000%, apply VD to bring this down to the real figure of 10% to 15%. If a TikTok guru predicts the end of linear TV by Christmas 2025, you make a note to expect significant changes around the start of the 22nd century. If they tell you their company is projected to make a billion dollars this year, you can round that figure down to six figures.

You get the idea.

But Patel’s numbers are so huge and the topic so close to his genuine expertise that, despite my subsequent application of VD, his post still left an impression. Even a fraction of his suggested activity is beyond the capabilities of all but the largest corporate media team. But if Patel knows anything it is surely how to build social media success.

What is a marketer to do in the face of such numbers? Spend the next 10 years in their underpants trying to come up with something – anything – to say across various platforms every few minutes, while adding inane comments to as many other posts as possible? There must be another way.

The bullshit is all going to prove to be true. AI is coming to change everything. Maybe not immediately. And certainly not in the ways old men and women predict with total conviction and zero knowledge.

And, of course, the other way is already on the horizon. We talk a lot about AI without any foundation for its actual impact. Yet, like a giant techno-comet that may or may not hit earth and may or may not change everything, most marketers of a certain age now look to the gloomy skies of 2024 to work out what is coming and what damage it will do their industry and their career.

I say a certain age because for some of us in the greyer part of the discipline there is that rarest of things: an advantage to being middle-aged.

“One of the few bonuses of being in your mid-50s is the relative comfort I take from not being too concerned about the AI revolution that occurs over the coming decades,” I said to one worried marketer who asked me about their place in the post-AI corporate world last week. “By the time the real revolution is upon us I should be all but done. But If I was 29,” I went on, “I’d be shitting myself.”

“I just turned 29,” my now even more worried marketer replied. And I felt genuine pain for a moment. This marketer really wanted to be a marketer. They were enjoying their career so far. They looked forward to the next part of it. A quarter of a century ago I was them, or they would have been me. But in 2024 there is a shadow over the heads of a talented, enthusiastic marketing generation. A genuine fear grips this group that we should talk about more. I feel and fear for them because, while change is inevitable, it is also a dispassionate fucker during the process.

Why the AI bullshit could be true

I have a good friend who is a Japanese business translator. He spends eight hours a day in his Australian bedroom working for a bunch of large corporations turning English into Japanese, or back the other way round. Well, that was what he was doing until he noticed a dramatic slump in his workload at the start of the year. Last week he sat in my kitchen drinking wine and looking glumly out of the window. In a matter of months his assignments halved and his clients are telling him that AI now does the work, in real time, for less, better.

I write an exam each year for the Mini MBA in Marketing. I create a fictional firm, brew up some customer data and then ask my marketing class to build a marketing plan to save the day. To set the scene I need pictures of buildings and products and salespeople and logos. So, I always need loads of stuff that I pay copyright for or that I create in a really shit way that spoils my nice exam. Not anymore. I have a free membership of some AI derived thingy that allows me to type in a few sentences of the image I want and then about seven seconds later it exists. And I paste it into my exam. It’s not DNA sequencing or anything, but it’s impressed the fuck out of me. And saved a tonne of time and money.

I have a fictional sales manager in my current exam called Tom Suarez, for example. He’s originally from Chicago. In his 50s. Hispanic. Plays too much golf. Letting his team down with sup-optimal sales performance. But he’s well connected with the CEO, so he is untouchable. And he knows it. And he knows that we know it. That was what I typed anyway. And here is Tom, seven seconds later.

See what I mean? It’s early days but this shit is at the very top left corner of the cost/quality chart. It’s better and it’s cheaper and it’s easier. The bullshit is all going to prove to be true. AI is coming to change everything. Maybe not immediately. And certainly not in the ways old men and women predict with total conviction and zero knowledge from the stages of every marketing conference venue on the planet right now. But its impact will be – I believe – bigger than most of us have so far imagined.

And in Neil Patel’s bonkers optimal growth suggestions, we might have found one of the first legitimate places where AI can and should play a proper marketing role. No human is going to be able to come up with nine videos a week on Facebook or produce 17 comments a day on LinkedIn. They will turn quickly to AI to do that work for them. Think about it. It’s an ideal application for AI. A repetitive, ongoing information flow that enables a human to look good without doing anything.

Very soon, perhaps already, social media platforms will be deluged with daily content from users who are not actually using the platform at all. Their AI content engine is doing all the heavy lifting. And, as the cliché suggests, once one AI engine ejaculates generic content all over a platform, others will scoop it up and regurgitate it. Shit will beget shit. The digital snake eats its own pixelated tail.

And why stop at social media posting? It’s a relatively minor part of the marketing universe. What about digital advertising? We have been debating whether AI will ever produce better brand advertising than humans and, in doing so, we miss the point. The big, rare, emotional, brand-based communication campaigns – the ‘long’ side of Field and Binet’s effectiveness research – are probably always going to be created by fallibly brilliant advertising people who go to the pub too much and secretly want to be artists. We need this kind of communication to be new, different and so very human.

But the ‘short’ of it – the bigger mass of multiplicitous, rational, product-focused, targeted messages that benefit from a test-and-learn approach and work within the parameters of immediate ROI calculations – are surely a hot zone for AI for the next few years.

By then, much of social media and digital advertising will be synthetic: derived from algorithms, created by machines and then tested and evolved continuously in real time without any human intervention of any kind. Except of course when we read the stuff on our social media feeds. And the intriguing question is whether LinkedIn, already turning into a boring cesspool of faux humility and self-serving hot takes, remains a worthwhile place to visit when you are unknowingly talking with vacuum cleaners and cash machines.

It’s more than likely that at least one social platform will go anti-synthetic and enforce a non-AI policy using some future filter we are yet to develop. It has always struck me as fascinating that America has the very worst food, brimming with preservatives and genetic modification, and – when you walk the aisles of Whole Foods – some of the best stuff too. And, of course, these two extremes create and reinforce each other. The fake chemical awfulness of so much American food ensures there is also a place for pristine, organic fare.

Compare that to Italy, still the greatest food place on the planet. You rarely see mention of the word ‘organico’ over there. They don’t need it. Most food is made with care. With attention and love for the land. When you order cheese or sausages in Italy you assume it will be made by the best human hands in the loveliest possible way.

Perhaps the upcoming AI domination of marketing communications, at least the low-hanging shorter stuff, will eventually usher in an era in which we move from an Italian-style model where we assume all content is human to an American binary system where you either stipulate ‘human, organic content only’ or risk drowning in an cesspool of synthetic recycled ejaculate.

Anyway, just a thought. And my first published thought for more than a week. How the fuck am I going to get the other 87 bits of content done by Friday to keep Neil Patel happy? Perhaps Tom Suarez can help?