The default logic is that AI will just replace all writing tasks, and writers will go extinct.
What actually seems to be happening, however, is this:
- obviously written-by-AI copywriting is perceived very negatively by the market
- companies want writers that understand how to use AI tools to enhance productivity, but understand how to modify copy so that it doesn’t read as AI-written
- the meta-skill of knowing what to write in the first place becomes more valuable, because the AI is only going to give you a boilerplate plan at best
And so the only jobs that seem to have been replaced by AI directly, as of now, are the ones writing basically forgettable content, report-style tracking content, and other low level things. Not great for the jobs lost, but also not a death sentence for the entire profession of writing.
Addressing the wider point, yes, there is still a market for great artists and creators, but it's nowhere near large enough to accommodate the many, many people who used to make a modest living, doing these small, okay-ish things, occasionally injecting a bit of love into them, as much as they could under time constraints.
Translation is an good example. Still need humans for perfect quality, but most use cases arguably don’t require perfect.
And for the remaining translators their job has now morphed into quality control.
This has a bunch of implications that are positive and also a bunch that are troubling. On one hand, it's likely going to create a burst of economic activity as the cost of these marginal activities goes way down. Many things that aren't feasible now because you can't afford to pay a copywriter or an artist or a programmer are suddenly going to become feasible because you can pay ChatGPT or Claude or Gemini at a fraction of the cost. It's a huge boon for startups and small businesses: instead of needing to raise capital and hire a team to build your MVP, just build it yourself with the help of AI. It's also a boon for DIYers and people who want to customize their life: already I've used Claude Code to build out a custom computer program for a couple household organization tasks that I would otherwise need to get an off-the-shelf program that doesn't really do what I want for, because the time cost of programming was previously too high.
But this sort of low-value junior work has historically been what people use to develop skills and break into the industry. And juniors become seniors, and typically you need senior-level skills to be able to know what to ask the AI and prompt it on the specifics of how to do a task best. Are we creating a world that's just thoroughly mediocre, filled only with the content that a junior-level AI can generate? What happens to economic activity when people realize they're getting shitty AI-generated slop for their money and the entrepreneur who sold it to them is pocketing most of the profits? At least with shitty human-generated bullshit, there's a way to call the professional on it (or at least the parts that you recognize as objectionable) and have them do it again to a higher standard. If the business is structured on AI and nobody knows how to prompt it to do better, you're just stuck, and the shitty bullshit world is the one you live in.
Someone still has to choose what to prompt and I don’t think a boilerplate “make me a marketing plan then write pages for it” will be enough to stand out. And I’d bet that the cyborg writers using AI will outcompete the purely AI ones.
(I also was just using it as a point to show how being identified as AI-made is already starting to have a negative connotation. Maybe the future is one where everything is an AI but no one admits it.)
> And I’d bet that the cyborg writers using AI will outcompete the purely AI ones.
In the early days of chess engines there were similar hopes for cyborg chess, whereby a human and engine would team up to be better than an engine alone. What actually happened was that the engines quickly got so good that the expected value of human intervention was negative - the engine crunching so much information than the human ever could.
Marketing is also a kind of game. Will humans always be better at it? We have a poor track record so far.
EDIT: As in, it can make really good derivative works. But it will always lag behind a human that has been in real life situations of the time and experienced being a human throughout them. It won't be able to hit the subtle notes that we crave in art.
It can absolutely do that, even today - you could update the weights after every interaction. The only reason why we don't do it is because it's insanely computationally expensive.
This could change with varying results.
What is average quality? For some it’s a massive upgrade. For others it’s a step down. For the experienced it’s seeing through it.
Every model has a faint personality, but since the personality gets "mass produced" any personality or writing style makes it easier to detect it as AI rather than harder. e.g. em dashes, etc.
But reducing personality doesn't help either because then the writing becomes insipid — slop.
Human writing has more variance, but it's not "temperature" (i.e. token level variance), it's per-human variance. Every writer has their own individual style. While it's certainly possible to achieve a unique writing style with LLMs through fine-tuning it's not cost effective for something like ChatGPT, so the only control is through the system prompt, which is a blunt instrument.
(and shittier software, etc)
