In online marketing, companies are often advised to be authentic. However, AI-generated texts, images and videos mean that they lose precisely this. Or is it only a matter of time before AI can also be authentic?
Recently, a building cleaner wanted me to help him optimize his website. No problem. While looking through the website, I laughed heartily when I came across the service “maintenance cleaning”. Of course I meant “maintenance cleaning”. Naturally, I corrected the error (and many others), graphically revised the website and optimized it for search engines. As amusing and charming as the one slip-up was, the website as a whole was unprofessional. And that is certainly not the impression the building cleaner wanted to give potential customers.
But cases like this are decreasing, while the opposite problem is increasing: Websites are free of spelling mistakes and look smart, but no longer have anything to do with the company – even though the services are described correctly and the “About us” is accurate in terms of content. Website generators, ChatGPT and other AI tools send their regards.
Do customers even want authenticity?
Before lamenting the loss of authenticity, we should first clarify whether customers are interested in authentic marketing at all. If they were, companies that post boring, uncreative and aesthetically unappealing, but all the more authentic content on Instagram would have to generate a lot of attention and reach. Do they do that? No. Why? Because authenticity in marketing is not a value in itself. Authenticity satisfies neither a longing nor any other human need. It is merely a vehicle for conveying credibility. In marketing, reality must therefore be shaped and staged in a way that suits the company, the brand or the product. After all, the vast majority of customers want exactly that: real people, real products, but please staged in an exciting, funny, creative or spectacular way and preferably peppered with an attitude to life or a narrative with which they can identify.
Staged reality is an old hat
You don’t win customers with a snapshot of a Big Mac lying carelessly in a cardboard box, slightly squashed and losing a few iceberg leaves. That’s why McDonald’s doesn’t print it on its advertising posters and doesn’t post such a photo on Insta, but instead stages it. This is the most normal thing in the world, and McDonald’s makes no secret of it – as you can see from this making-of product photo on YouTube. Customers know this too, which is why they don’t complain about eating a Big Mac. Because they want to be seduced. And legally, they would have the short end of the stick anyway, because as much as the Big Mac on the McDonald’s poster differs from the real thing, it doesn’t lie, because the ingredients are there: Burger patty, cheese, lettuce, sauce, bun. They just look a little different. And we see this form of staged reality not only at McDonald’s, but in fact at every advertising company, from car manufacturers to pet shops.
Can the AI be made more authentic?
Back to the topic of AI. “It’s all about the prompt!” is sometimes said when an AI tool does not produce the desired results. Or to put it positively: if you specify your input, you get better results. At this point, a distinction must certainly be made between text generation, which already works very well, and image and video generation, where there are still a number of detailed errors. But let’s stay with text generation. By providing your prompts with stylistic information such as “more casual”, “less promotional” or “simpler”, you can adapt the texts to your own writing style. If you tell her to put in a few spelling mistakes or stylistic blunders, she will do that too, and will get carried away with oblique comparisons such as “a good cleaning plan is like Swiss clockwork”.
The more data, the more authentic!
You can get even more authentic results by uploading your own content to train the AI. The more, the better. If a lot of data is available, it is comparatively easy for AI tools to recognize patterns and produce new content in a similar style. This applies equally to texts, images and videos. That’s why ChatGPT, Gemini and other Large Language Models (LLMs) can imitate Goethe’s writing style better than mine – because they don’t just have access to a few blog articles, but to all of his works. Everyone has to decide for themselves how much personal data they want to feed into an AI tool (and the legislator in the case of copyright-protected works), but it won’t fail because of the technology. This also applies to image and video generation. If an AI tool has access to all Wes Anderson films and ideally also to expert analyses that analyze the style and break it down into its components, it is comparatively easy for the AI to create a trailer for a Harry Potter film in Wes Anderson style:
The more data, the scarier!
It’s obvious, but still creepy: the more the AI knows about you, the better it can imitate you. So if you not only feed the AI with technical texts, but also provide it with personal data, tell it about your own opinions, attitudes, values, feelings and thoughts, the more open you are towards it and the more you reveal about yourself, the more the AI becomes like you. As a result, she then only needs a basic text or basic information and can then, for example, write an opinion piece that could not only be stylistically but also in terms of content. That doesn’t just sound scary, it is scary. And that quickly leads to the big questions in life: What is man? Who am I? And if so, Richard David Precht?
This is what ChatGPT says about this blog post:
The blog post “AI and the loss of authenticity” impresses with its clever structure, clear argumentation and a successful mix of anecdote, specialist knowledge and social reflection. The language is pointed, entertaining and yet factual – the anecdote about the “Unterhalsreinigung” charmingly loosens up the text without drifting into the banal. The differentiated examination of the question of whether authenticity is a relevant value in marketing at all is particularly successful – with comprehensible arguments and a good feel for target group reality. The examples (e.g. McDonald’s, Wes Anderson trailer) are vividly chosen and strengthen the argument.
What the text lacks in places is a somewhat stronger central theme – the jump from loss of authenticity to prompting to philosophical questions at the end seems slightly overloaded. Nevertheless, it remains worth reading and is thought-provoking without being preachy.
Grade: 1- (very good minus)
The text is very strong in terms of language and content, but loses some of its formal clarity at the end due to thematic expansion.
