Human plus AI: how communication really gets better

AI doesn't make weak brands stronger. It makes them invisible faster. Because as soon as everyone uses the same tools, standing out becomes the only thing that counts, and that is something the machine can't do for you.

---

Everyone now uses the same tools

The promise was that AI would make communication faster and better. Half of that promise holds. The work is faster. Whether it gets better depends on something the tool itself doesn't deliver.

By now more than 74 percent of all new web pages contain AI-generated content, and around 85 percent of marketers use AI text tools. That has a predictable consequence. A language model doesn't choose what is different, it chooses what is likely. The same model, fed the same training data, answers the same brief for a hundred companies with almost the same text. Researchers call this content homogenisation: the more organisations lean on the same generators, the more their communication creeps towards one another. Sameness is not the exception, it is the default outcome.

---

AI is a multiplier, not a leveller

Here is the misconception that costs organisations dearly. AI is sold as a leveller, the tool that lets the small player suddenly do what only the big one used to manage. In practice it works the other way around. AI multiplies what is already there.

A brand that knows exactly who it is uses AI to roll that sharpness out across every channel, more consistently and faster than ever. A brand that doesn't know multiplies its vagueness, also faster than ever. The plus sign in human plus AI only pays off if there is something defined on the human side. If there is nothing there, AI simply adds to the noise. Strong brands grow stronger with AI. Weak brands disappear faster because of it.

---

Speed was the old argument, judgement is the new one

Execution speed was long the argument an agency won work with. Whoever could produce faster, won. That argument is largely gone. A well-trained system delivers in seconds what used to take hours.

The labour market shows the same shift. Harvard research found that the number of vacancies in roles suited to automation fell by 17 percent, while roles where AI strengthens the work rather than replacing it rose by 22 percent. In that same study, 94 percent of respondents want AI as a collaborator, not a replacement. Stanford researchers built a yardstick for it, the Human Agency Scale, which establishes per task where human involvement remains essential to the quality.

The question that remains once the machine is done is this: is this good, does this fit who we are, does this say something worth saying. That question hasn't become any faster. That question calls for judgement, and judgement is precisely what a language model doesn't produce on its own.

---

The client can feel whether a human is watching

From 2025 onwards, professionals have been warning about AI slop: the flood of generic, empty AI content that hollows out brand identity for anyone who deploys it without editing. Not out of ill will, more out of an absence of judgement. Someone types a prompt, accepts the output, publishes. The result fits a hundred brands at once.

At Schwung we see this reflected in the work. A careers website built with AI but without a clear employer story sounds like every other one. The candidate recognises nothing and drops off. The work was accelerated, the brand was missing, and the visitor felt it. Not consciously, but at the moment they clicked away. AI as a tool works exactly as well as the foundation it serves.

---

Being found starts with having something to say

There is a new reason why emptiness is expensive. People increasingly search through an AI rather than a search engine, and an AI like that gives no list of links but one answer with a few sources. A page full of sales talk doesn't get quoted. A thin page doesn't either, not even when there are a hundred of them.

We wrote earlier that a content engine is not a writing aid but a findability machine. The same applies here. Being found by human and machine doesn't start with more text, but with having something to say that is worth repeating. The organisations that get good at this are not the biggest or the fastest, but the most consistent. They know what they say, say it well, and repeat it in the places where it lands.

---

What this means for agency and client together

The value of a communication agency shifts along with it. For a long time it was about who could make the work. Now it is about who can judge whether the work is right. An organisation that loads its brand foundation into a brand-style agent or content engine produces consistently on-brand content across channels and over time, without reinventing every expression from scratch. That only works if the foundation is right: if the essence is clear and the positioning is carried by who the organisation truly is.

That is where the real plus sign sits. Not the human beside the machine, but the human who gives the machine a direction no one else has. Judgement beats output. Experience beats speed. AI doesn't make that difference smaller, it makes it more visible.

---

Further reading

• Branding and recruitment communication with Schwung
• Schwung | agency for web design and web development in Tilburg