{"id":88980,"date":"2025-08-08T09:50:39","date_gmt":"2025-08-08T09:50:39","guid":{"rendered":"https:\/\/mailinvest.blog\/index.php\/2025\/08\/08\/editors-notebook-ai-and-the-writing-craft\/"},"modified":"2025-08-08T09:52:07","modified_gmt":"2025-08-08T09:52:07","slug":"editors-notebook-ai-and-the-writing-craft","status":"publish","type":"post","link":"https:\/\/mailinvest.blog\/index.php\/2025\/08\/08\/editors-notebook-ai-and-the-writing-craft\/","title":{"rendered":"Editor\u2019s Notebook:\u00a0AI and the Writing Craft"},"content":{"rendered":"

<\/a>\n<\/p>\n

\n

Nothing brings me extra pleasure in my job than seeing a author flip in a draft of outstanding high quality.<\/p>\n

I\u2019ve been reviewing writers\u2019 work, formally or in any other case, for over half my life. Typically, it might be the school roommate who wanted one other set of eyes on his work, or working as an editor for the sports activities desk at my faculty newspaper. Or because the managing editor and, later, vice chairman of content material for a community of then-upwardly cell sports activities web sites.<\/p>\n

For the final virtually 4\u00a0years, doing so for Kuno\u2019s content material workforce has paid my payments and offered me ample alternatives to really feel that sense of delight for a author. I nonetheless get fired up when anyone delivers me excellent work.<\/p>\n

So when generative AI \u2013 particularly, ChatGPT \u2013 broke out in Spring 2023, I had not-entirely-unjustified considerations about how it might have an effect on our day-to-day. I didn\u2019t consider it might put us out of labor, and I nonetheless do not. However my fear was and is that the work we writers put in would develop static and joyless, and that we might be buying and selling our inventive and strategic edge for productiveness and effectivity.<\/span><\/p>\n

Writers are creatives, and the cold machinations of an algorithm are inclined to run counter to issues of and merchandise from the center. Even advertising and marketing copy and industrial whitepapers.<\/p>\n

There are the actual worries right here: the moral ramifications of utilizing LLMs to supply content material that’s in any other case not unique to the author (and because of this, U.S. regulation prohibits purely AI-generated work from being copyrighted). Ecological considerations endemic to AI\u2019s proliferation for any goal (notably, right here in my house state of Wisconsin, using Lake Michigan to offer water for a Microsoft knowledge middle close to Kenosha and the Illinois border has any variety of potential\u00a0unintended penalties to each the habitat and surroundings).\u00a0<\/p>\n

Then there are the threats of hallucination, disinformation, bias or outright errors from LLMs. ChatGPT\u2019s propensity to patronize the person, as much as and together with encouraging self-harm. The research indicating physiological indicators of dependancy to AI that resemble substance abuse.<\/p>\n

After which there\u2019s the evolving impression on the ever-changing content material panorama<\/em>. \ud83d\ude80\ud83d\udcaa\ud83d\udcc8<\/p>\n

For those who\u2019re even remotely focused on AI use and on LinkedIn, there\u2019s little doubt your feed is bloated with any variety of individuals combating over what phrases, phrases and phrases represent hallmarks of gen AI. And as a content material skilled waist-deep in content material day in and time out, once I see sure phrases and phrases, I\u2019m educated to look nearer and hunt down tone.<\/p>\n

I\u2019m dissatisfied to report that I see these hallmarks in all places <\/em>in society. Social media, radio spots, advertising and marketing content material, press releases, drafted laws; I even learn a latest piece in The New York Instances <\/em>that seems to have been crafted in no small half from the byproduct of a ChatGPT immediate. Individuals are much less focused on doing the work than they’re in getting consideration, by no means stopping to think about that if everyone seems to be looking for consideration, nobody will get it \u2013 at the least, not in the way in which they essentially need. After all, that is nothing new and positively not unique to the office.<\/p>\n

I am additionally inspired by the truth that, right here at Kuno, we have averted the worst impulses AI permits for.<\/p>\n\n

\u201cLanguage failed this week.\u201d<\/em><\/h3>\n

New York Instances <\/em>literary critic Michiko Kakutani penned this Hemingway-grade lede in her Critic\u2019s Pocket book \u2013 from which I drew inspiration for this column\u2019s title \u2013 days after the September 11 attacks<\/span><\/a>. Studying that commentary for a category on journalistic ethics whereas mud and smoke nonetheless rose from Floor Zero haunts me to today.<\/p>\n

She closes that column thusly: \u201cIt could appear trivializing — even obscene — to speak about films in the identical breath as this week’s tragedy, however the truth that so many individuals did was a symptom of our incapacity to get our minds round this catastrophe, our incapacity to seek out real-life precedents, real-life analogies for what occurred within the morning hours of Sept. 11.\u201d<\/p>\n

Granted, we don\u2019t write breaking information tales from the uncooked feed of present occasions. We\u2019re primarily B2B entrepreneurs, and our job is to attach shoppers with potential patrons in ways in which imply one thing to the latter. What does a literary critic writing a few nationwide atrocity a quarter-century in the past must do with us?<\/p>\n

Since that point\u00a0and, extra to the purpose, because the arrival of AI for the lots, it might appear that we as a individuals received\u2019t even attempt to get our minds to metaphor or corollary, cinematic or in any other case. Gen AI permits us to outsource our takes to gadgets that string collectively phrases based mostly on likelihood and a rudimentary sense of context. And all these inputs can be found for the supplier to additional feed their LLMs. The result’s a rhetorical race to the underside, from bulletins to weblog posts. Nice work brings me pleasure; equally, work that blatantly exhibits its GPT-stitched seams, properly, it\u00a0brings me down.\"Callout<\/p>\n

Writing is connection and neighborhood, issues synthetic intelligence essentially can not domesticate or foster. Because the editor right here, my function has more and more been working with Kuno stakeholders and writers right here to determine firmer guardrails round acceptable use of AI. (To make certain, there are acceptable makes use of, particularly within the palms of well-trained individuals.) We\u2019re an employee-owned outfit, and somebody who takes these two phrases actually, I really feel a weight of accountability to ensure AI doesn\u2019t flip our output into one thing resembling the Old \u201897<\/span><\/a>: quick and out-of-control to off the rails.<\/p>\n

The purpose isn\u2019t to name anybody out right here; it\u2019s a name encouraging us to lift expectations for our writing. In any case, patrons have excessive expectations of these nonetheless within the operating on the determination section. We owe it to them to ship at that stage, as a lot as we owe it to ourselves as advertising and marketing professionals with a way of integrity and delight in our craft.<\/p>\n

These are a few of the purple flags for AI use I\u2019ve discovered over the latest months and years. You could have your personal, through which case, I\u2019d love to see your examples<\/a>!<\/p>\n

The Biggest Hits<\/strong><\/h3>\n

‘Evolve,’ ‘Panorama,’ ‘Right this moment,\u2019 \u2018Fashionable,’ ‘Drive,’ \u2018Unlock’<\/strong> \u2013 LLMs are primarily fueled by chances: variables are decided based mostly on most-likely-common use.<\/p>\n

Briefly, gen AI makes use of averages to create what quantities to common content material.<\/p>\n

It\u2019s not designed to have a degree or point-of-view, however to be middle-of-the-road, even merely satisfactory. These phrases are so entrenched in gen AI output to the purpose of being cliche and switch readers off. My guess is that it received\u2019t be lengthy earlier than engines like google comply with swimsuit and begin searching for phrases like these, penalizing accordingly within the spirit of helpful content<\/a>.<\/p>\n

‘Not solely…but additionally…’<\/strong> \u2013 This was one in every of my unique hallmark phrases for AI, not a lot as a result of it was prevalent (although it was), however as a result of the both-and seldom truly works from a technical standpoint. AI typically hallucinates these, or conflates with an A-to-Z language machine reasonably than a price and value-added 1-2 punch.<\/p>\n

‘In the end’<\/strong> \u2013 One other case of lots of people not understanding what an final is and polluting the LLM. Usually, AI doesn\u2019t do adverbs properly. I can\u2019t let you know what number of items I\u2019ve seen that use in the end for various topics in a chunk and I\u2019m left to marvel, Wait, what number of finish objectives are there right here, anyway? <\/em>AI doesn\u2019t know as a result of it will possibly\u2019t do the strategic work people do.<\/p>\n

‘Ensures’<\/strong> \u2013 This one is admittedly difficult as a result of, 1) there aren\u2019t loads of good synonyms, making it troublesome for writers to differ, and a pair of) frequency shouldn’t be essentially a marker for gen AI output. The important thing right here, as with checking towards plagiarism, is ensuring that context stays intact, on this case, the verb and topic are appropriately aligned. Not each X<\/em> really ensures Y<\/em>.<\/p>\n

Talking of keys\u2026<\/p>\n

\u2018Key\u2019 <\/strong>\u2013 I blame this one on the hyper-SEOification of content material over the previous decade. Key this, key that; not like \u2018ensures,\u2019 there are lots of good adjectives capable of\u00a0be deployed as a substitute. Merely put, this showing throughout AI output is a failure of creativeness from AI (which is, in fact, incapable of it).<\/p>\n

And I\u2019m not dragging the worth of writing in ways in which make crawlers joyful (although methods outdoors authoritative, credible content material seem like dwindling). Writing to be seen consists of each writing properly and writing in ways in which permit for indexing. Today, when the latter takes priority over the previous, we are able to moderately conclude that AI has finished the work for the author.<\/p>\n

Now, we flip to technical and tonal purple flags:<\/p>\n

Passive voice (‘The function of <\/strong>Y<\/em><\/strong> in <\/strong>X<\/em><\/strong>‘) <\/strong>\u2013 As a result of writing properly is much less and fewer of a precedence for educators, we see increasingly more passive voice showing in all places from information articles (the place it’s traditionally verboten) to titles and headings in content material (dragging readability down). There are few causes to make use of passive voice as a substitute of energetic, for instance:<\/p>\n

\u2018The function of passive voice in dragging readability down\u2019 versus \u2018Passive voice\u2019s function in dragging readability down,\u2019 one is stilted and formal, like a teen making an attempt to ship their first tackle in speech class. The opposite will get to the center of the matter, sustaining efficient tempo for the reader.<\/p>\n

Constant, even awkward use of energetic tense (Ing syndrome) <\/strong>\u2013 Even with the significance of energetic voice, not each verb must be energetic. That\u2019s why I confer with this as Ing syndrome, the place every part that occurs is rapid and pressing. The biblical guide of Mark is written like this within the unique Greek, making your entire quantity roughly a giant, awkward run-on sentence. Your weblog content material isn\u2019t going to be studied as sacred textual content centuries after the actual fact, on the identical time, that doesn\u2019t excuse writing on the stage of an historical close to jap commoner.<\/p>\n

Once more, all of it is a product of what fashions are fed, and what\u2019s fed is essentially common at greatest writing.<\/p>\n

Misaligned singular verb to plural nouns <\/strong>\u2013 This is without doubt one of the newer ones I\u2019m seeing, and it\u2019s difficult as a result of that is much less a consequence from generated output itself and sometimes from making an attempt to therapeutic massage AI-generated content material into one thing with a extra human (and humane) tone. Like a contemporary coat of paint, that odor signifies one thing in a room has modified.<\/p>\n\n

The Most Necessary Factor<\/strong><\/h3>\n

What issues most is that we\u2019re not residing and dying by these hallmarks. Sure, they assist establish AI content material, however we ought not miss the forest for the timber. Each important reader runs the danger of lacking the worth in a chunk as a result of they see an AI purple flag, myself however.<\/p>\n

Not way back, I flagged a phrase for doable AI use in a chunk I used to be reviewing from a long-time freelance author. Relatively than getting indignant over it, I left a margin notice with my concern, and it prompted an awesome dialog about writing, phrase selections and our views on the craft. On this case, the author lamented that AI had co-opted a most popular phrase, one which by the way seems above within the biggest hits.\u00a0<\/p>\n

And that\u2019s the place significant development occurs for each author and editor \u2013 within the belief and conversations round our work that permit us to debate writing and our respective approaches to it. The\u00a0writer-editor relationship brings readability and perception to the event course of, enriching each the content material and people who produce it, saying nothing of worth for an meant viewers. Typically, the byproduct delights an editor as a lot as it should a reader.\"Callout<\/p>\n\n

For those who\u2019re searching for efficient content material that comes from an intensive understanding of your model and enterprise, Kuno\u2019s workforce of content material strategists and writers are right here to assist. We responsibly use AI to tell our writing course of and develop real work that meets actual, human prospects at any (or each!) section of the client lifecycle.<\/p>\n

In an age when content material is more and more homogenized by probabilistic fashions, considerate content material stands out that rather more. Make Kuno content material a trademark of your advertising and marketing technique. Contact us to learn how.<\/p>\n

\"get<\/a><\/span><\/span><\/p>\n

Editor\u2019s Pocket book is a brand new recurring column with commentary and perception on content material and communication from Kuno Digital Content material Editor Brent Sirvio. All em dashes are the writer\u2019s.<\/em><\/span><\/p>\n<\/div>\n