The future of writing isn’t a cage match between AI and humans. While pundits obsess over whether ChatGPT will write the next great novel, some writers are already deep in a different kind of draft: prompt engineering. Writers of the next generation will spend less time typing first drafts and more time architecting prompts, curating ideas, and directing outputs.
The skillset is shifting—from penmanship to production design. While today, AI-powered search engines, recommendation feeds, and large language models (LLMs) are increasingly deciding which writers get visibility and readership. The critical fact is that these discovery systems thrive on original, engaging content and filter out the rest.
By now, we’ve all heard the cliché that “AI won’t replace you… a person using AI will.” This bumper sticker advice has cemented a man-versus-machine narrative in pop culture. Equally common is the fear that AI will flood the internet with bland, mass-produced content. The concern is understandable – if everyone starts outsourcing writing to AI, it’s easy to imagine drowning in a sea of cookie-cutter prose. But these narratives miss a key point about digital media platforms themselves: AI-driven systems punish homogenization.
Search algorithms and LLM-based assistants are designed to satisfy users by surfacing content that is unique, credible, and valuable. They actively bury duplicative or derivative material.
Google’s own research into
AI search works similarly – models like Bing’s or ChatGPT’s answer engines favor content with novel insights or up-to-date data and disregard repetitive rehashes. As James Allen writes for Search Engine Land, understanding what gets cited (and why) in AI-generated answers has
Readers crave differentiation, and so do the algorithms that serve them. In a “samey” world of AI-generated text, the moat is originality – your unique voice, style, and perspective. If ten articles say the same thing, an LLM has no reason to choose yours; engagement drops, shares vanish, and your work gets pushed to the fringes of the internet.
This creates a paradox: the easier it becomes for anyone to produce generic content with AI, the more valuable truly original content becomes. Authentic expertise, new information, and a distinctive tone become the currency of the realm. Indeed, the AI revolution in content discoverability is “reshaping our entire information ecosystem,” and the winners will be those who offer genuine insight rather than regurgitated facts.
Adaptation vs. Obsolescence: Augmenting Human Creativity With AI
The writers who will thrive in this new landscape are those who use AI as a creative amplifier rather than a replacement. Think of generative AI as a power tool – it can vastly increase your output and efficiency, but it still needs a skilled operator with a vision.
Forward-thinking writers leverage AI to research faster, brainstorm ideas, generate drafts, and even iterate their writing, all while maintaining their own voice and judgment. They still bring the original angles, the storytelling craft, and the editorial standards that machines alone can’t provide. In other words, they use AI to enhance their creativity and productivity, not to erase it.
On the other hand, talented writers in digital media who refuse to engage with AI face a growing disadvantage. It’s not that their natural writing skill is inferior, it’s that they won’t be able to compete with the scale, speed or quality of experienced writers-turned-prompt-engineers.
Content discovery systems (from Google’s search index to ChatGPT’s training data) are increasingly favoring those creators who can produce high-quality work at a higher volume. A writer using AI can, for example, publish five thoroughly researched, well-written pieces in the time a traditional writer produces one – and do so without sacrificing quality.
If you opt out of these tools entirely, you may maintain craftsmanship, but you lose the race of volume without losing quality. In effect, the new threat is not AI itself but other writers who master AI workflow as part of their writing strategy.
It’s telling that a majority of writers and editors today feel anxiety about having “the most AI-proof job” and yet feel clueless about applying generative AI in their work. That split between fear and understanding is dangerous. Generative AI is reshaping the media landscape faster than most realize, and those who can’t adapt or differentiate will soon find themselves irrelevant.
Simply put, having great writing skills isn’t enough if you ignore the new tools and channels. As I wrote before, AI isn’t the cause of writers’ job insecurity –
The safe haven for writers is not to avoid AI, but to use it better. The moat isn’t built by rejecting the technology; it’s built by mastering the balance between human creativity and artificial intelligence.
Original thought, provocative angles, and a distinctive voice remain irreplaceable. AI can amplify those qualities if you have them – but it can’t generate true originality from scratch. The question every writer now faces is whether you will adapt to write with AI, or effectively write yourself out of relevance by clinging to old ways.
You must tell the world something new if you want to earn attention (and citations) from AI-driven platforms. In a world of auto-generated blandness, your unique perspective is your strongest asset.
This isn’t some dystopian hellscape where Orwellian AIs replace authors wholesale ... it’s more like stepping through Lewis Carroll’s looking glass, entering an alternate world of writing where the rules are flipped upside down.
In this world, the writing process isn’t hijacked by AI but transformed by it. But you must transform your approach to writing itself if you plan to work with AI. Far from losing agency, you gain new capabilities … but only if you choose to use them. And for many writers, this is hardly the first fundamental change in the craft ushered in by technology…
From Typewriter to Thought Partner: How Writing Processes Have Evolved
To understand how fundamentally writing is changing, let’s trace the lineage of writing tools and roles:
- Typewriter Era – Manual and Linear: Decades ago, writing was entirely manual. Typewriters enforced a linear, non-revisable process (no cut, copy, paste). Writers had to plan carefully and often retype entire pages to make edits. The focus was on each word and keystroke, with little assistance beyond basic reference books. The process was slow and required immense labor for research (visiting libraries, scanning microfiche) and revision.\
- Word Processor & Internet Era – Assisted but Time-Heavy: The arrival of word processors and later the internet changed the game. Word processing allowed editing text easily, and the internet (Google searches, online libraries) opened a firehose of information. By the 2000s, a writer’s workflow involved Googling for sources, reading dozens of tabs, and manually synthesizing facts into a draft. It was easier than the typewriter days, but research still took hours, and drafting still meant typing every sentence yourself. The writer was still doing 100% of both the thinking and the typing, just with better tools for editing and lookup.
- AI Assistance Era – “Executive” Writing: Now, with advanced AI like GPT-5 coming, we’re entering a phase of agentic AI partners. Today a writer can input a topic or outline and have an AI model instantly locate relevant sources, summarize them, and even generate a rough draft in minutes. What used to require days of research and writing can be collapsed into an hour-long workflow. The writer’s role shifts from transcriber to director. You sketch the vision; the machine fleshes out the paragraphs. You steer the strategy; the machine handles the scaffolding. In effect, you become an editor or architect of the piece, orchestrating content rather than grinding it out word by word.
This isn’t outsourcing your job to a robot – it’s elevating your role to a higher level. The best writers understand that embracing AI means leveling up from content laborer to content strategist.
Think of it this way: if you could instantly delegate first drafts, research summaries, and tedious edits to a junior assistant, what would you focus on? You’d focus on everything that matters most: your voice, the clarity of your arguments, the narrative structure, the insight and angle only you can provide. Those are precisely the things generative AI can’t do on its own. It has no genuine voice, no judgment, no sense of storytelling or prioritization – it relies on the human to provide those.
In the new model, the human writer provides vision and critical thinking; the AI provides speed and scale. The outcome is a highly efficient generative AI workflow where the writer-editor can produce top-notch work in a fraction of the time.
I’ve started calling this AI writing strategy “executive writing,” because you’re managing the production of content more than line-writing it. You’re delegating to your AI assistant like an executive delegates to staff – but you’re still responsible for the final product. Not coincidentally, this mirrors how many creative industries evolve: as technology automates the lower-level tasks, humans move up the value chain to more strategic, creative decisions.
Importantly, this new process raises the bar, it doesn’t lower it. When research and drafting are semi-automated, the bottleneck becomes the thinking.
The hard part of writing in the AI era isn’t pounding out words – it’s deciding what’s worth saying and how best to say it. The mental heavy lifting of analysis, interpretation, and original insight becomes even more central.
In a sense, final drafts now demand first-class thinking. AI can give you a competent draft of an idea, but only you can refine that into a compelling argument or a resonant story. The writers who succeed will be the ones who double down on human creativity at the final stages – editing with a keen eye, injecting nuance, ensuring the piece has a soul. Those who just copy-paste AI output without adding human value will churn out forgettable content that discovery algorithms quickly learn to ignore.
Final Draft, First-Class Thinking: The Human Touch as Competitive Advantage
There’s a growing misconception that AI-generated content means the end of quality – that automation will flood us with mediocre writing and lower the overall bar. In reality, the effect is the opposite for those who choose to compete at the high end.
When AI handles the boilerplate and the brute-force work, it raises the standards for what humans contribute. The new competitive landscape means anyone can generate a passable blog post or essay with a few prompts. So what sets the best apart? The thinking and creativity behind the words.
If you’re a writer, your job is no longer just to produce words – it’s to ensure those words carry insight, perspective, and purpose. AI can draft your prose, but it can’t determine your point of view. It can’t decide which argument truly matters, or what tone will resonate most with your audience, or which anecdote best illustrates your message. These remain distinctly human decisions. This is your moat of irreplaceability.
In fact, as content gets filtered through AI intermediaries, writing for discoverability means writing with extreme clarity and strong angles. An AI summation or snippet of your piece will only be compelling if the piece itself had a clear, compelling idea. Thus, top writers will start to think more like editors or strategists from the outset.
Before a single paragraph is drafted (AI-assisted or otherwise), you’ll want to outline a sharp structure, have a clear thesis, and think about the takeaway for the reader. It’s like commissioning yourself before you begin writing – what is the story here, and why is it important?
By shaping the narrative early and giving the AI a strong blueprint, you ensure the final output isn’t generic. You then refine the AI’s draft to amplify your unique voice and expertise.
This approach separates commodity content creators from strategic thinkers. A commodity writer might just prompt an AI and post whatever it spits out, adding minimal editing. That content will blend into the mass and likely vanish.
A strategic writer uses AI to generate material, but spends serious time curating and polishing it to editorial standards that stand out. They fact-check the AI, add missing evidence, inject personal experiences or case studies, and cut the fluff. They might use AI to explore multiple angles and then choose the most provocative one to develop further. In the end, the piece is theirs – the AI was a junior collaborator.
It’s worth noting again that content discovery algorithms (whether Google’s ranking or an AI’s answer-picking) are getting better at evaluating quality signals beyond just text matching. They look at user engagement, at whether content gets referenced elsewhere, at credibility of the author or site, etc.
Strategic thinkers who consistently produce insightful, well-crafted pieces will build up those signals – human readers will share and cite their work, and AI models will learn that this author has authority.
As explained in our
You’re no longer trying to capture dumb web crawlers in their own webs; you’re trying to “impress” an AI that actually reads and evaluates your content in a human-like way. It’s a higher standard, but it’s ultimately aligned with what human readers have always wanted: useful, well-written, trustworthy content.
If you focus on substance and structure now, you’re future-proofing your work for both AI and human audiences. Meanwhile, those churning out faceless AI-written fluff will see their work (and perhaps careers) languish in obscurity. In a very real sense, strategic thinkers still have a job tomorrow.
The Prompt Era: Why Writers Will Spend 50% of Their Time Prompting
We’ve arrived at a point where writing great content is as much about how you communicate with AI as how you communicate with readers. Prompting – the craft of instructing AI models to get the desired output – has become a critical skill. In fact, many writers are now devoting about half of their content creation time to prompt design, iteration, and review of AI output. This is the birth of the “prompt era” in writing.
Let’s start with a core philosophy: Prompting isn’t a gimmick or purely technical trick … it’s the modern extension of planning and outlining. Just as outlining a piece forces you to clarify your structure, writing a good prompt forces you to clarify your intent. In a way, prompt engineering is the new brainstorming.
It’s thinking out loud, but in a structured format that a machine can build on.
Prompting Is the New Planning
Good writers have always had to be planners to some degree. Before drafting, you consider your audience, the intent of the story, and the flow of information. In the age of AI, this planning materializes as prompt-writing, setting up the AI with all the context you need to write your story.
On the one hand, generic prompting is ChatRoulette for writers – a basic prompt like “Write a blog post about X” may yield the story you want … but only if your luck is better than your prompt. Advanced writers, on the other hand, practice full-stack prompt engineering: providing a model with carefully chosen examples, defining a role or persona for it (e.g. “You are an expert tech journalist writing in AP style…”), breaking the task into subtasks, and even guiding its chain-of-thought with step-by-step instructions. These complex prompt setups are essentially the new outline. They don’t just tell the AI what to do, but how to approach the topic in detail.
For example, you might feed the AI a structured prompt package: first, an outline of section headings you want; next, a few examples of the tone or style you like (perhaps a paragraph or two excerpted from a personal favorite piece of yours, or a specific quip you wrote to inject some wit); then specific points you want each section to cover. The AI then uses all that to draft the piece. The result is far closer to your vision than a one-line generic prompt could ever produce. You’ve effectively engineered the prompt to get an output that saves you significant editing time.
This is prompt engineering as a real skill – it’s about knowing the right “inputs” to give the AI so that the “output” aligns with your goals.
Basic prompting – like just asking a single question – is becoming commoditized. Anyone can do it. The leverage comes from mastering complex prompt strategies that are tailored to how you think and what you need.
Some content teams even create reusable prompt templates or prompt workflows for different types of content (press releases, listicles, tutorials, etc.), almost like editorial checklists. The evolution of prompt engineering reflects a shift from ad-hoc use to systematic integration in writing projects.
Prompts Are the Skeleton Key for AISO
One fascinating intersection is between prompt design and AI Search Optimization.
If you know that AI-driven search systems value certain content structures – for instance, clear definitions, FAQ sections, summary tables – you can prompt your AI assistant to include those in the draft.
In effect, your prompts can bake in AISO-friendly features from the start. Think of prompts as a way to pre-format and pre-structure content for optimal visibility in AI results.
Traditional SEO taught writers to insert keywords and meta tags after writing. AISO teaches us to consider semantic structure and citeability from the get-go.
For instance, you might prompt:
“Draft an introduction that clearly states a unique thesis (information not commonly found elsewhere). Then provide a bullet list of 3-5 key takeaways with facts or stats. Ensure each takeaway is one sentence and impactful.”
By doing this, you’re creating content chunks that an LLM can easily digest and maybe even quote verbatim. You’re essentially adding a form of strategic markup via the prompt itself.
Another example: if you want an AI answer engine to pick up your content as an answer, you could prompt the AI writer to include an explicit Q&A format in your article (“FAQ: Q: [common question]? A: [concise answer].”). This aligns with how LLMs were trained on platforms like Quora and forums – they love question-answer pairs.
As a result, an LLM scanning your page might find exactly the question it needs and a well-formed answer to quote. In essence, prompting becomes a way of ensuring your final content isn’t just reader-friendly, but AI-friendly too. It’s a new kind of optimization where you’re simultaneously writing for the machines that act as information gatekeepers.
AI Is Your Junior Editor – If You Know How to Manage It
Those who fear that AI will replace writers and editors misunderstand the opportunity. AI can be thought of as a tireless junior editor or assistant on your team. It can check grammar, suggest alternative phrasings, enforce a style guide, or flag inconsistencies. But, like any junior staffer, it needs supervision. It’s good at following rules and patterns, but it lacks true judgment. So you manage it by giving it rules and feedback.
For example, you might maintain a custom prompt or persona for editing. Here’s one I use to preface multiple editing prompts:
“Act as a copy editor who always preserves the author’s unique voice and humor, but correct any factual errors and tighten the text for concision. Go through the selected text and edit for grammar, punctuation, and generally for readability and flow. Use AP Style and spell out numbers less than 10. Use the Oxford comma…”
When you feed your writing through this, the AI will produce an edited version that follows your custom instructions. Many writers are finding that with a well-crafted editing prompt, they can catch issues and improve clarity much faster. It’s like having a second set of eyes that never gets tired – but you remain the editor-in-chief who must approve or tweak its suggestions.
The myth that “AI isn’t good enough to edit or write properly” misses the point. If you get poor results, often it’s a prompt problem, not an AI problem. These models will do exactly what you tell them, if you tell them clearly and specifically enough. A GPT-4 that’s been trained on your own writing (via few-shot examples in the prompt or fine-tuning) and guided by your standards can become a huge force multiplier. But it requires you to step into a management role: defining what “good” looks like, creating guidelines, and feeding those into the AI systematically.
This is why some forward-leaning organizations are building internal tools and prompt systems – to turn prompting into a repeatable system that ensures quality. They use persona prompts (for style/voice), few-shot examples of ideal outputs, and modular prompt blocks for different checks (one for facts, one for tone, one for SEO checks, etc.).
The result is an AI-assisted workflow where even less experienced writers can produce content at a higher standard, because the AI is injecting collective best practices into their drafts. In such a setup, the human writer becomes a manager of an AI ensemble: guiding them, reviewing output, and making high-level decisions.
Spending time on prompts and AI feedback isn’t a waste or an “extra” step – it’s now an essential part of the creative process. In many high-output editorial teams, the breakdown of a writer’s time might be ~30% on structuring and refining prompts (planning the piece via AI), ~20% on reviewing and correcting the AI’s output, and ~50% on final drafting, polishing, and injecting the human touch. In other words, half the time or more might be spent in prompting and handling AI output, and the other half in classic writing/editing mode.
This 50/50 split is not a sign of laziness; it’s a sign of efficiency. By letting the AI handle the heavy lifting of producing a draft or gathering info, a writer can devote a full 50% of their effort to higher-level editorial work – exactly where they add the most value.
Crucially, this means the total output of a writer can be much higher. If you can double the number of quality articles you produce by cleverly using AI, that’s a huge career advantage. And rather than diluting your voice, this practice can actually sharpen it, because you spend more time refining your ideas and less on slogging through first drafts.
When I was a deputy managing editor, the publisher
It’s the principle of editorial leverage: a term that signifies how applying high-level editorial insight across scaled content production leads to much greater impact. By managing an AI assistant, you gain a form of editorial leverage that lets you cut through an age of infinite content with genuine value. You’re still creating and curating the ideas – you’ve just got a tireless helper to execute the repetitive parts.
The writers who flourish will be those who see AI not as a threat, but as a teammate. They hone the skill of giving AI direction (prompting), and in return they gain speed, scale, and even a bit of creative serendipity (AI can suggest ideas you hadn’t thought of). The endgame is not AI-written content that feels generic, but AI-assisted content that feels inspired.