As a writer, I often wonder: will people actually want to read AI-generated content? Increasingly, it seems like the answer is no. There’s even a new social shorthand that’s emerging: AI;DR – “AI; didn’t read”.

Readers are becoming increasingly sensitive to copy that sounds generic because AI could have written it. And they’re not wrong to be suspicious. But as AI gets better and better at writing, will we start to doubt everything we read?

That’s what led me to speak to Alys Key. An independent journalist focusing on tech, business and increasingly AI, Alys has spent the last few years thinking seriously about what it means to work alongside these tools – not against them.

We talked about AI’s evolution, where it’s useful – and why writers and journalists need to get original.

Where is AI useful?

To my surprise, Alys makes it clear she is pro-experimenting with AI from the off. “Industries like art and journalism are reflexively anti-AI, and women tend to adopt AI less,” Alys explains. “But I don’t want to be left behind.”

For Alys the first traces of AI’s impact on journalism weren’t seen in writing but transcription. Back when she was starting out, she used to spend hours transcribing interviews – pausing, rewinding and straining to catch half-heard sentences. Now with AI, that’s become a lot simpler. “I can drop it into Otter.ai and search for a quote in seconds,” Alys reveals.

And as a freelancer, without a wider newsroom to support her, Alys often turns to Claude for spelling and grammar checks or to pick her argument apart. Her top tip: “You need to tell it to be really mean as LLMs are very sycophantic.”

“And you should only take edits that resonate with you,” Alys underlines. “I’ve been in creative writing groups where you get feedback from several people. Maybe only one person is similar to your target reader, but you’re also hearing from someone who is always really critical and another who doesn’t read your genre.”

So take AI’s thoughts with a pinch of salt? “Yes.”

Should we let AI write?

But what about the actual writing? How does Alys feel about AI getting in on the action? “When AI writes an article that answers ‘what time is Strictly on tonight?’ or ‘where can I watch the Super Bowl in the UK?’ I don’t see that as a problem,” Alys starts.

“But you can go much further. Take Velora, a cycling news website co-founded by Peter Stuart and Danny Bellion. It’s a fully AI-powered newsroom, but Peter, as a former editor of Cycling News, is the human-in-the-loop at every stage. They have a very clear AI policy as they want to avoid creating slop.”

It’s an interesting prospect, a lean AI operation overseen by one person. Journalism is an increasingly squeezed industry, and small players have to do what they can do to stand a chance. For already-formulaic news articles like who won a Tour de France stage that day, I get it. But what about being able to access a range of experts from trusted news sources?

With AI and shaky online revenue models, Alys predicts a shift away from the “bloated” centralised newsroom to more local reporting with feet on the ground – something AI simply can’t deliver – and pieces with a distinct voice.

“A good example is The Spectator, which just hit record sales. People are reading it because they enjoy their writers’ voices.”

What writing will come out on top?

For me that hits the nail on the head. AI is pretty impressive, but it lacks one very human quality: personality.

Yes, it can mimic human voices, but it can’t invent. It can’t tell a strange but compelling anecdote. And it certainly doesn’t know the euphoria of watching your favourite cyclist finally win.

As value shifts to lived experiences and original thinking, this is impacting both how Alys and I write. “I'm not a local journalist, but a lot of my newsletters focus on going to an event or an office and speaking to someone, so I try to add some colour that makes it clear that I was actually there.”

For me, I’ve always searched for unusual takes, but now it’s more important than ever. That’s how I got to my latest on how coffee shop chains helped me make peace with AI.

What does this mean for brands?

The pressure to be original is on, and that goes for brands too. Hiding in the “average” shadows won’t cut it anymore. People are seeking a shock of authenticity. Niche ideas. Voices that could only be human.

Alys mentions Ambrook, an American B2B accounting software startup for family farmers that is big on personality. “They have a blog that is just beautiful, nicely written essays about what the American dream means to them, for example.”

The blog is filled with intriguing, first-person angles that create a sense of fun and real community around a brand that could be pretty dry. The title of a particularly striking piece about getting out in the field as an accountant is simply “Heifermetrics” – need I say more?

And it makes commercial sense, too. As Alys explains “investors and tech people love this stuff” as they seek to build the prestige around the thesis of their company: “It’s deeper thinking – almost academic but not quite – and a bit poetic.”

Find your niche

Alys’s advice is clear: lean into what makes your voice distinct. Quote your F. Scott Fitzgerald. Be a little strange.

I’m with her. When creating content is so easy that a machine can do it, only the most human – and authentically intelligent – pieces will capture readers’ attention.

Predictable won’t cut it in the personality era. We are set to see a move towards more complex, literary writing, where us writers need to raise the bar. Because if people suspect you didn’t write it, they’ll label it AI;DR and scroll on by.

‍

‍

‍Alys Key is an independent journalist, covering business, tech and AI in the UK. She was previously features editor at Digital Frontier and deputy money and business editor at The i. You can read her newsletter – UK 2.0 – here.