The end of the working writer
A couple weeks ago, the Bloomberg writer Joe Weisenthal went mini-viral on Twitter for something I think he’s largely on the money about. I refuse to go into the details of an argument that unfolded on a social media platform where people are incentivized to dunk on one another, but I do think the most interesting thing that emerged are two entrenched perspectives where one side thinks it’s insane to use LLMs for anything and the other that think it’s insane not to use LLMs for everything
Sometimes, I think we as writers overestimate how much people care about the quality of writing outside of specific venues like fiction, criticism, and narrative journalism. But the balance of written words produced in the world are not valued for their quality, but rather for their existence. White papers, product copy, SEO blog posts—these are all hypercommercialized pieces of writing meant to one serve specific end purpose: buy this and not that. This is writing as a product and a large portion of that taxomony is, also, already being created by writers using LLMs. The same, to Joe’s point, will start happening to the little quotidian notes we have to send back and forth as part of daily life, the confirmation emails, the thank you notes, the scheduling and rescheduling. I cannot imagine using AI to do any of that, but I also think that a whole mass of people would rather click a button to tell their colleague they’re running 5 minutes late or cancel a reservation at a restaurant. Writing is extremely important to me; that does not mean it is important to everyone. I have a lot of respect for Adam Johnson’s work, but I think his QT was the perfect example of selection bias that many writers have:
Here’s something that I think more professional novelists, journalists, and critics should understand: the vast majority of people who write in a commercial capacity are already using LLMs to help them write both because they’re being told to by management and because it’s a really fast way to turnaround a draft. It’s happening to content teams at tech firms, it’s happening to freelancers working as contractors, it’s happening to large corporate marketing teams who need to churn out copy. This doesn’t mean people just slamming a prompt into Claude and shipping it, but rather that they’re using AI programs to produce the bulk of the work and editing down from there which has become one of the major responsibilities of corporate writers in 2026. I’d argue this is what people should actually be worried about, rather than the glut of shitty stories and novels emerging from people who have no interest in writing to begin with.
A few years ago, someone from a creative agency emailed me asking if I had any bandwidth to take on writing a white paper on some extremely esoteric technical topic for a major technology company. I had no idea what the hell any of the materials they sent over said, so obviously I told them of course I could do it and when did they need it by. Two weeks they said, which was a slightly tighter timeline than I usually like so I asked for $20,000. They said yes immediately which annoyingly told me I could have asked for more. But the lack of hassle was always its own blessing.
The best part about this corporate work is that it never needed to be that good and everyone understood that. The people at the agency knew it, the tech company knew it, I knew it. I just needed to make it easy enough so that non-technical executives could read it and understand, like, 60% of the topic. The agency handed me some more background materials: PowerPoints, speeches, PDFs. I think the paper I wrote was about edge computing or something, but I honestly couldn’t tell you anything about it at this point. I wrote it, someone from the agency edited it and handed it back to me, rinse, wash, repeat. It took about a month to get published and I couldn’t tell you if the damn thing even exists anymore. It’s likely lost to some CMS memory hole, another piece of content thrown into the Great Internet Garbage Patch.
Work like this was my living for a long time. I would have to learn something quickly and then write about it from a place of at least passable fluency for a general audience. I did not consider that work part of my creative oeuvre, and in fact made significant efforts to partition it from how I wrote for newspaper and magazines.
Now that I don’t have to do it anymore, I wonder if that commercial writing should be treated the same as, say, this blog? Is all writing real writing? I think often of the wonderful John Warner’s comment that these LLMs “can’t kill anything worth preserving.” (His book-length treatise on this topic, “More Than Words,” is excellent and you should buy it. John is also no doubt grimacing at many of the arguments I’m making in this essay lol) In this case, the argument is about chatbots in classroom settings so the contours are slightly different, but I don’t think any professional writer sets out to be a great content marketer. If these pieces of writing are susceptible to robotic writing because they had very little humanity in them to begin with, then I’m not sure whether to mourn them or not. The one major caveat to that I have is an economic one. I think chatbots destroying these writing “products” is roughly similar to what AI companies like Suno are doing to working musicians who have paid their rent creating stock music or commercial jingles. They were not creating, strictly, art, but they were able to at least adjacent to it and making a living. It often feels like one major economic outcome AI is the squeezing out of middle class creatives.
Depending on your perspective, it’s either fortunate or troubling that AI is extremely well-suited to doing the sort of stupefyingly boring work I used to do and seemingly quite bad at the rest of it. If you feed Claude the exact same materials that the ad agency gave me and give it a halfway decent prompt, you’ll get something like 70% of the way to what eventually got sent to the IBM team for final review. Sure, it’ll be full of the vacant platitudes and annoyingly predictable LinkedIn rhythms that all AI writing has, but when it takes all of 30 seconds to produce a draft there is a lot of time left over to edit for style. (Not that style matters a terrible amount here.)
From what I’m hearing, a lot of that work has dried up while rates have cratered. Projects I would have charged several thousand dollars for because they were a function of hours spent are now worth a fraction of that because a client’s first question will be whether you use AI or not to write a first draft. I made it out in time, but I fear that a lot of younger writers who could have done that sort of work to support themselves will be denied those stabilizing paychecks. It was a sturdy if boring way to exist as a “working writer” and something the vast majority of freelancers I know had to do to make ends meet even if they didn’t like talking about it. But if someone working in content at a major technology firm is able to get Claude to write something 70% as good for 10% of the cost of a freelancer or a staffer then I’m not sure how human writers compete. Even all of the “storyteller” and “narrative” jobs that have become voguish in the last few years may not survive the next real downturn because “taste” isn’t as vital as “site reliability engineering” no matter what the vibe coders tell you.
I genuinely hate this and, for the record, I think the world would be better without these writing tools because people are using them to replace the kind of “thinking” writing necessary to produce something human. The volume of robotic submissions to literary journals and awards competitions is overwhelming, and the unscrupulous morons doing this are making it harder for the rest of us. We’re also quickly losing the kinds of jobs that have sustained writers for decades. It is remarkably harder for younger writers to make a living because the writing that didn’t take any real thinking to begin with has been taken over by unthinking machines. I do not like these tools, but people need to stop dismissing things they find repulsive or uncomfortable. It is part of the responsibility of writers to be curious about why something bothers you in order to understand the entire shape of the problem.
Joe’s tweet certainly struck a nerve because, of course, things are teetering and many writers see themselves as part of a vanishing class under threat from people who have no interest in the humanity of thought and creativity. People also clearly do not like AI writing, especially in places they expect to encounter organic thoughts written by people on the other side of a screen or page. Max Tani wrote last night in his indispensable Semafor media newsletter about how Reddit has been a sort of Ground Zero for understanding that dynamic. “The way people react to AI writing on Reddit clearly shows that people on the internet still want original, well-informed thought that feels human. Mostly,” Tani writes. The problem is that so much of writing created by people was the kind of stuff no one cared about. There is more a general, immediate economic issue that is a lot more concerning to me than the spiritual one. I am less threatened by the presence of bad writing created by robot-equipped dullards than I am by the specter of there being no way to make a living as a writer at all.








I keep wondering what role the local and in person will play in this new environment. Will theater see a resurgence? Are there types of in person or more physical jobs that will replace content marketing roles for more creative people? Anecdotally, have noticed that my creative friends are having good luck in things like Waldorf teaching, woodworking, reiki. These are people who are first and foremost artists & writers and who even a year ago were copywriters etc. anyway, mangled thoughts here but wondering what, if anything will fill the vacuum.
I would describe my face shape as more "furrowed brow" than full grimace while reading this, but in truth, I don't really disagree with much.
I think what we're looking at is a shift from needing what I'd call "writers" to "automation-assisted text producers." The categories of artifacts you discuss lend themselves to automation-assistance in various ways and at various levels. The fact that they're judged on their ultimate utility to the end audience makes it relatively straightforward to judge whether or not the automation is harming the final product.
I would argue that the best automation-assisted text producers are likely to have strong writing practices and so if you want to become one of those people you should still learn to write, but the bigger question is whether or not the people commissioning this work will continue to pay for human involvement.
Some of this stuff probably is "bullshit writing" which I explored in a deleted chapter from the book that I published here (https://biblioracle.substack.com/p/bullsht-writing?utm_source=publication-search).
If I'm a company that's commissioning this stuff, the first question I'm asking is whether or not I actually need this stuff. As you say, it often seems to drop down a hole. I can think of all kinds of artifacts I had to produce in my time in market research and can pretty clearly map what I would need to write, what I could automate almost in total, and where automation-assistance would be useful, but again, that's because I know what writing is.
I think Wiesenthal is overstating the case because I think this tension is just as likely to clarify where human involvement is necessary as it is to erase the occasions where humans remain involved, but there's no doubt that we are in the midst of a shift.