"It's not about doing – it's about deciding"

Anna Ryberg works as a communications strategist at RISE and has just completed her master's thesis, in which she interviewed communications professionals about how AI is changing their work. This blog post is based on a conversation with Anna about the results of her research.

What was the clearest change you saw?

That competencies don't disappear, but are reweighted. Almost none of my respondents say a skill has become useless. But the relative importance shifts dramatically. Writing, for example, which has long been the very core of the profession, becomes less decisive. What weighs more heavily instead is the ability to judge what AI delivers. Is the tone right, the level right, is it right for the recipient? You could say the work shifts from doing to deciding.

Which competencies become more important, then?

Above all, professional judgment and critical thinking. Then problem formulation — knowing exactly what you want help with and being able to articulate it. Several compare it to commissioning work from an agency: if your brief is unclear, you get a mediocre result. On top of that comes things like understanding your audience, a feel for nuance, the ability to read social context — things that are hard to put into words but become obvious when they're missing.

Is AI seen as a tool, or something more?

Here I was actually surprised. Almost no one describes AI as just a tool. The words that keep coming up are colleague, sounding board, assistant, sometimes even mentor. AI can sharpen one's own thinking rather than replace it — you become "muscled up," as one respondent put it.

At the same time, the line is crystal clear. AI is allowed to think along, but not to decide. Responsibility always stays with the human, whose role becomes more of a gatekeeper. The communicator becomes the one who reviews, quality-checks, and vouches for the content being true, relevant, and authentic. One driving force behind this is that people actually filter out AI-generated content — material that's technically fine but feels empty. Safeguarding the genuine thus becomes a strategic task, not just an aesthetic one.

If production is no longer the bottleneck, do we risk an overflow of information — more AI slop — rather than better communication?

That's probably the most acute risk in the short term. We're already in a landscape of massive information overload, where dependence on algorithms is high and it takes ever more creativity just to get through at all. Cheaper production doesn't solve that — it makes it worse, unless someone actively pushes back. This is exactly where the professional role becomes more important.

But there's also a new reality to relate to here. More and more of what's produced will be received, interpreted, and further processed by machines before it ever reaches a human or an automated decision system. That requires dual competencies: knowing how to create content that works when the recipient is a machine, and how to design content that appeals to people.

Does that mean fewer jobs, when one person with AI can do what used to take several?

The picture is mixed. There are examples in the data of teams going from three people to two without losing capacity, because AI makes up the difference. But reasoning about efficiency often starts from the assumption of status quo — that the tasks stay the same. That's not the case. Even if AI removes certain tasks, new ones are created too, because the technology enables new things. So I don't primarily think existing teams will shrink dramatically — rather, expectations for what they should deliver will increase.

The study also shows that the breadth of competence required of each individual keeps growing. That points to an important paradox. Steering AI well requires experience-based understanding that's normally built up in the more operational roles that are now at risk of becoming fewer. Where will tomorrow's senior judgment come from, if the path there gets narrower?

What do you do with the time freed up when production speeds up?

That might be the most interesting question going forward. Some put it toward more production; others put it toward analysis and reflection — things AI can't do. That can mean in-person meetings and other things that land in a way algorithms don't. The more content produced everywhere, the more noise. And then the genuine, the thing that stands out, the personal — that becomes a competitive advantage. Several talk about "the last ten percent" of a job. AI can take a text or image to 70 percent done in a matter of seconds. But what makes it good — that final human touch — can't be automated away. And creativity in the sense of creating something genuinely new, not just new combinations of what already exists, is still seen as a human strength.

If you were to sum up what the study says about the future of professions like communications: what is the human role when AI does more of the work?

To be the one who understands why, not just what. AI can answer almost any question, but it doesn't know what matters for your specific organization, your audience, your context. That knowledge — that quiet, hard-to-articulate experience — is where humans still make a difference. It can feel uncomfortable, being forced to think clearly. But those who will do well here are neither the ones who resist nor the ones who hand their thinking over to AI entirely. It's the ones who use AI to think more, not less. And maybe that's exactly why AI companies are hiring philosophers. When execution gets automated, meaning is what's left to wrestle with.

The study "Working methods, competence and professional role – what happens to strategic communication when artificial intelligence enters the workplace?" is Anna Ryberg's master's thesis at Uppsala University, based on 15 interviews with communications professionals in the public and private sectors.