When I started out as an editor, when giving an uncomfortable piece of feedback I would start with, “This is fine, but…”
In any publishing environment, ‘fine’ can be a dangerous, almost insulting word. Worse than damning with faint praise, and definitely worse than just saying what you mean to say, which is that it needs rewritten.
But what I meant by ‘fine’ was that the copy was technically sound, the spelling and grammar largely correct, the structure makes sense, essentially nothing is obviously wrong… and yet something is missing.
As I gained experience leading editorial teams, I developed an editor’s eye: a sensitivity to the gap between ‘fine’ and compelling, readable, clickable. The article reads smoothly, but it doesn’t quite grab. It informs, but it doesn’t persuade. It ticks all the SEO boxes, but it doesn’t look beyond that. It’s competent – and forgettable.
That instinct has become sharper with AI firmly embedded in editorial workflows.
Most content and editorial teams now use generative AI in some form. The benefits are obvious, and I’ve outlined them in previous blog posts like this one on using AI for content writing. It helps you get started, and it can turn a rough outline (or even stream-of-consciousness word soup, as I am prone to do) into something coherent in minutes. It’s particularly good at clearing the basic hurdles: readable, structured, tonally neutral.
AI can make writing acceptable very quickly. It’s much worse at making it interesting, and acceptable is never the benchmark, not on my watch!
The eye for flatness
Editors know when something is emotionally flat. It’s hard to define precisely, but you know it when you see it. The sentences are tidy, the transitions are logical, the vocabulary is current but generic – the article feels as though it could have been written by anyone.
That’s not accidental, of course: AI systems are trained on enormous amounts of text. When asked to produce ‘clear’ or ‘professional’ writing, they will gravitate towards statistical averages of what good writing looks like. The result is clean, competent, and often indistinguishable from everything else produced in the same way.
Interesting writing doesn’t exit algorithmically, it lives in specificity, in slight tension. It pulls you this way and that, feeling deliberate rather than optimised.
An editor’s job has always involved asking: where is the sharpest version of this idea? What’s the sentence that carries weight? Where is the beating heart of this article? AI can assist with structure and fluency. It struggles with ‘feel’.
The eye for veracity
There’s another instinct that matters even more. A good editor doesn’t buff up copy, they interrogate it. If a piece hinges on a claim presented as fact, the first response is to ask: is that actually true?
Where did that statistic come from? Is it current? Is it being interpreted correctly? Is correlation being presented as causation? Does this quote need context? Has someone with subject-matter expertise validated this argument? Editors are trained to be suspicious of confident prose.
AI, by contrast, is often extremely confident in tone. It produces sentences that sound authoritative and complete. That surface fluency can make weaknesses harder to spot.
The risk is that AI can and does produce plausible assertions that haven’t been properly verified. It blurs nuance and simplify complex issues into clean summaries that feel convincing but lack depth. If no one applies editorial judgement at that stage – if no one pauses to question the premise – then ‘fine’ copy crosses over to misleading copy; a complete red flag.
The eye for consequence
In many organisations, readability has quietly become the primary quality test. If something is clear, structured, and professional, it passes. That’s understandable: time is limited and an obsession for some. Volume and production ‘at scale’ are the buzzwords du mode, meaning workloads are high. AI makes it possible to move faster, and faster often feels better.
But writing does more than fill blank pages with crawlable KBs, it shapes perception, influences reader decisions, and ultimately it reflects on a brand or publication.
If a piece rests on an unverified assumption, exaggerates evidence, or removes important caveats in the name of clarity, it may still read smoothly, and it may even perform well traffic-wise, but it erodes trust over time.
The editor’s role has always been as the protector of brand credibility. AIs answer prompts, editors question premises. That distinction matters even more now.
The eye for when to wrap up a blog post
As I’ve said before, none of this means AI is a threat to writing. It’s a great tool, a powerful tool. Used carefully, and sparingly, it can be genuinely useful in outlining, summarising, and restructuring, and free up time for deeper thinking.
But tools don’t exercise judgement. Interesting writing still requires emotional input in order to elicit an emotional response. And responsible writing still requires verification. Credible writing still requires someone willing to say, “Hang on – is that actually right?”
The editor’s instinct has become more valuable in 2026 because in a world of acceptable content, the difference will be made by those who are prepared to move beyond ‘fine’.
