AI is making communications faster. That part is obvious. What’s less clear is whether the work itself is getting any better.

A few years into all this, AI is already part of how most teams operate. It’s there in the background, pulling together research, drafting first cuts, summarising long threads, helping you get from nothing to something in minutes instead of hours.

And yes, that’s useful. Anyone who says otherwise is probably overstating the case.

Less friction. Faster turnaround. Work that used to drag now moves.

But speed has a way of passing for progress. That’s where things start to feel slightly off.

Because faster isn’t the same thing as better.

Where the value actually sits

The value of communications, particularly at the advisory level, was never really about the output on its own. It’s about the thinking behind it. What you choose to say, what you leave out, and why. Reading a situation properly before deciding how to respond. Knowing when something needs to be addressed, and when it’s better left alone.

That’s the work people actually pay for, less the act of producing communications and more the judgment behind them

And it doesn’t get easier just because the tools do.

Judgment in this line of work is rarely neat. It’s deciding whether a response will land well or quietly erode trust. It’s picking a line that works in one market and realising it could fall flat, or worse, backfire in another. It’s looking at a flood of signals and figuring out which ones matter, and which ones don’t.

Sometimes it’s knowing that the safest answer isn’t the right one.

That part is still human. And it’s still where the value sits.

What the tools don’t do

Most of the tools we’re using don’t really touch it. They’re built to help you get things done. Draft faster. Summarise quicker. Pull information together in a clean way. Useful, yes. But they don’t help you decide what matters. They don’t hold context over time. They don’t really challenge how you’re thinking about a situation.

They help you move. They don’t necessarily help you think.

So the work speeds up. The underlying judgment doesn’t always move with it.

When everything starts to sound the same

There’s another effect here as well.

The tools don’t just help produce work. Over time, they start to shape what good work looks like. The phrasing, the structure, the way arguments are built. Those patterns aren’t neutral. They come from the data, from the training, from how the models are put together.

Use them enough, and you start to absorb them.

You see it in small ways first. A familiar turn of phrase. A certain rhythm to how something is written. It all feels clean, complete, a bit polished.

Then it spreads.

When enough people are working from similar prompts and similar outputs, the work begins to converge. You get more content, but less distinction. Things move faster, but they start to sound alike.

And in this business, that’s not a small issue.

Because once language starts to converge, thinking tends to follow. The range of what feels like a “good” answer quietly narrows. The less obvious angle becomes less likely to show up, not because it’s wrong, but because it doesn’t quite fit the pattern.

The work gets smoother. It also gets safer.

Easier to approve. Harder to remember.

That’s where the real risk sits.

Why sameness is a problem

Communications doesn’t create value by being generically right. It creates value by being specifically right. By understanding nuance. By recognising when two situations that look similar are actually very different. By helping an organisation say something that cuts through, instead of blending into everything else around it.

Once that starts to slip, the impact isn’t just creative. It shows up commercially as well.

Brands start to sound interchangeable. Advice begins to feel familiar, even when it’s technically sound. Over time, it gets harder for clients to see where the value is actually coming from.

Sameness isn’t neutral. It’s how things quietly start to lose relevance.

What needs to change

None of this means AI isn’t useful. It is. It can broaden the field of information. It can help test language, stress ideas, get you to a more developed starting point faster. Used properly, it can make the work better.

But that depends on what you’re using it for.

If the goal is just speed, most teams are already there. If the goal is better thinking, then the bar is higher. It means asking whether these tools are helping you connect the right dots, hold the right context, and arrive at better decisions—or whether they’re just helping you produce more, a bit faster, within the same frame.

That’s a different question.

It also changes where the focus needs to be. Less obsession with volume. Less satisfaction with speed for its own sake. More attention on what actually informs judgment.

What helps you make sense of conflicting signals instead of just summarising them? What helps you hold context across conversations, markets, moments? What helps you test whether a line will stand up once it’s out in the world?

Those questions are harder. They don’t come with neat answers.

But they’re closer to where the value actually sits.

The real question

AI can produce something that sounds convincing. It doesn’t have to live with it.

In communications, that distinction matters. Because what we say doesn’t just get published. It gets acted on.

And the responsibility for that doesn’t sit with the tool. It sits with us.

So the real question now isn’t whether we’re using AI. That part is done.

The question is what it’s doing to the thinking underneath the work.

Is it making that thinking sharper? More grounded? More distinct?

Or is it quietly nudging us toward the same language, the same ideas, the same answers?

AI will keep making the work faster. That’s not going to change.

Whether it makes the work better is still up to us.

Because sameness might scale.

It just doesn’t win.