There are thousands of coaches and courses promising to teach amateur writers how to evoke emotion through words. Yet, on a mass level, that principle is being abandoned. The reason is simple: it has become too easy to type in a prompt.
The paradox of human taste
A mid-2024 Bynder ‘Human Touch’ study found that 52% of readers who realise AI has written the words in front of them immediately disconnect. That response – ambivalence, apathy, a mild sense of offence – reveals more about our relationship with language than any sentiment tracker could. Still, the companies and marketers who claim to value authenticity aren’t taking this into account.
As soon as a reader senses the voice isn’t human, most turn away. That recoil is emotional data worth studying. It shows how recognition and empathy still govern communication, even as technology tries to bypass them.
The same Bynder survey reported that 56% of respondents actually preferred the AI-written version to the human one. This confirms what many already know – AI now produces writing most readers find acceptable, even preferable. Yet anecdotally, many marketing decision-makers have noticed something else: in the absence of that familiar, ChatGPT-like phrasing, something feels missing. As large language models loop through their own outputs, and communication teams mirror them, the machine now defines the baseline for tone, even though persuasion should be the goal.
Why words have lost their psychological power
With reading habits declining in the Middle East and all over, our attention has drifted from the stimulation that language once provided. In the post-ChatGPT age, a certain degree of us have forgotten what copywriting is meant to do. A headline, slogan, or caption should move the reader to thought or action. Too often now, they exist merely to occupy space.
The UK’s Institute of Practitioners in Advertising (IPA) analysed the IPA’s database of effectiveness case studies to discover what truly drives tangible advertising success. Their research, published in ‘Marketing in the Era of Accountability’ (2008) and ‘The Long and the Short of It’ (2012), revealed that evocative advertising outperforms all other types. Rational, message-led campaigns produced very large profit gains in only 16 per cent of cases; blended rational-and-emotional campaigns achieved 26 per cent. Purely emotional ads achieved a 31 per cent rate of major profit growth. The conclusion was striking: messaging dilutes emotion, and feeling delivers profit.
Automation has made creation easier, but ease was never the purpose of creativity. For centuries, innovation extended human capability; now it removes effort instead. In marketing, that instinct to simplify has turned the act of noticing into a blur, where the visual rush replaces meaning. Many who hand their writing to large language models seek mere relief and effective freedom from the vulnerability of expression. The machine has become a mask, proof of efficiency rather than imagination, and therefore designed to conceal imperfection, and not create connection.
The culture of acceptable mediocrity
Across advertising, PR, and media, the same pattern plays out. Deliverables are built to avoid scrutiny. The story ends once the box is ticked. Metrics like site footfall, leads, and engagement have been relegated to Narnian fantasies invoked to defend output. Expertise has thinned; impact has flattened. The acceptable has replaced the exceptional, surviving only the quick-glance test.
Promoting products is these days harder for many reasons, not least the saturation of empty language. When marketing teams and decision-makers claim that every phase of construction has ‘redefined the industry’, that every message ‘is not just this, but that’, that every announcement ‘underscores unwavering commitment’, the result is linguistic exhaustion, but at least that’s an emotion. The words cancel themselves out. That’s the real story.
Narratives are stuck in the once upon, dulled by the comforts of convenience. Impact remains out of reach because ChatGPT has no stake in creating it. It is almost beyond argument that these technologies are the greatest writer’s companion ever, yet they’re not storytellers; they’re echoes we’re mistaking for voice.
The measurable middle ground
In four or five years, perhaps, this argument will have expired. The advancement of OpenAI and its competitors will likely have conquered the craft of writing and replicated its best patterns. Applied intelligently – as it rarely is – that future is already here. A/B testers know what a 2024 Cornell University study show. ‘The Value of AI-Generated Metadata for UGC Platforms’ explained that when creators co-edited AI-gen titles, viewership rose by 7.1 per cent and watch duration by 4.1 per cent, which is clear evidence of human-AI synergy.
Some of us have seen that the same applies to email subject lines: when a skilled writer refines AI produced options, read rates consistently outperform human drafts. The advantage lies in data and in defining the task precisely.
In 2025, though, the repetition borders on parody – sentences folding into themselves, phrases chasing their own reflections. The humour lies in the precision: the tech knows most of the rules of writing, yet none of the reasons. What it creates is symmetry without soul. Something is ending, however, because the beanstalk cannot hold the giant for long.
People may never buy your product, yet if they sense you are no longer trying to capture their emotional attention, they will generally disengage. The story of how ChatGPT took over our words may one day prove pivotal. The psychology of why we let it happen is something we can still understand now if we’re empowered with the right facts.