The Economics of Writing Just Changed

For most of human history, writing was scarce. Producing text required a person to sit down, think, and write. That constraint meant everything: it set the price of content, the value of authorship, and the assumption that behind every piece of writing was a human mind.

That assumption is gone.

Large language models can now produce text that is grammatically correct, factually plausible, stylistically flexible, and available on demand at near-zero cost. The supply of "good enough" writing has become effectively infinite. And when supply becomes infinite, price approaches zero.

It is already happening. Content farms are replacing human writers with AI pipelines. Marketing agencies are cutting copywriting budgets. Newsrooms are experimenting with automated articles. The raw output of words, the kind you could hire a freelancer to produce for a hundred dollars, is now available for fractions of a cent.

So what happens to human writing?

The Handmade Premium

When machine production makes something abundant and cheap, the handmade version of that thing does not disappear. It becomes a luxury. It becomes a signal.

Consider furniture. IKEA made functional furniture accessible to everyone. Custom woodworking did not die. It became a premium category. People pay more for a hand-built table not because it holds dishes better, but because it was made by someone. The craft is the value.

Consider music. Streaming made recorded music essentially free. Vinyl records, objectively inferior in convenience, hit their highest sales in decades. Live concerts became the primary revenue source for musicians. The experience of something real, something with physical presence and human performance, became worth more, not less.

Writing is following the same pattern. AI-generated text is the IKEA of prose: functional, available, indistinguishable from better-made alternatives at a glance. Human writing, with its perspective, its lived experience, its specific voice shaped by a specific life, is becoming the handmade table. The vinyl record. The live performance.

But there is a critical difference between writing and furniture. You can tell a hand-built table from an IKEA table by looking at it. You cannot always tell human writing from AI writing by reading it.

That is the verification problem.

The Verification Problem

In every other domain where a "handmade premium" has emerged, there are visible cues that distinguish the authentic from the mass-produced. The grain of the wood. The slight imperfections of a hand-thrown bowl. The warmth of a vinyl recording. These are perceptible signals that the thing was made by a person.

Writing has no such signal. A paragraph written by a human and a paragraph written by GPT-5 can be lexically, syntactically, and semantically identical. You cannot hold them up to the light and check. The "grain" of human writing, the idiosyncratic word choices, the rhythm of someone's thinking, the traces of lived experience, can be mimicked to a degree that defeats casual inspection.

This means the handmade premium in writing cannot function on trust alone. If anyone can claim their AI-generated text was human-written, and there is no way to verify the claim, then the premium collapses. The market cannot distinguish the genuine from the counterfeit.

AI text detectors were supposed to solve this. They failed. The tools produce unacceptable false positive rates, especially against non-native English speakers and writers with clear, direct prose. They are probabilistic guesses, not proof. No serious institution should stake reputations on them, and increasingly, none do.

The verification problem is real, and it does not have a software-only solution. You cannot determine authorship by analyzing text alone.

What AI Writing Means for Different Fields

The impact of this shift varies by context, but the direction is the same everywhere.

Journalism

Readers are already skeptical of media. Add AI-generated articles to the mix and the trust deficit deepens. Newsrooms that want to maintain credibility will need to prove their reporting was done by humans who were actually present, who actually interviewed sources, who actually investigated. Bylines will not be enough. Verification will matter.

Academia

The university essay is in crisis. Students face false accusations from unreliable detection tools. Professors cannot tell who did the work. The entire assessment model (assign writing, grade writing, assume the student wrote it) depends on an assumption that no longer holds. Academia needs new ways to verify authorship, not better versions of tools that guess.

Publishing

Publishers are already drowning in AI-generated manuscript submissions. Agents report a surge in query letters for books that were obviously machine-produced. The slush pile, always difficult, is now nearly impossible to sort through. For authors who write their own work, standing out means proving it.

Content Marketing

This is the field most immediately disrupted. If AI can produce a serviceable blog post in seconds, the value of a serviceable blog post is zero. What retains value is content tied to a real person's expertise, experience, and reputation, the kind of content that only works when the audience believes a human produced it.

Legal and Medical Writing

Some contexts cannot tolerate ambiguity about authorship. A legal brief must be attributable to a licensed attorney. A medical report must reflect a physician's professional judgment. These fields will be early adopters of authorship verification, not because they want to, but because liability demands it.

The Infrastructure Gap

We have authentication infrastructure for almost everything except writing. Digital signatures verify documents. Blockchain verifies transactions. SSL certificates verify websites. Two-factor authentication verifies identity.

But we have no widely adopted system for verifying that a piece of writing was produced by a specific human being. This is the infrastructure gap.

The gap exists because, until recently, it did not need to be filled. The assumption that all writing was human was so deeply embedded that nobody thought to build systems to verify it. Now that assumption has evaporated, and we are left with no tools to replace it.

Filling this gap does not mean banning AI writing. AI-generated text is useful, and it is here to stay. The goal is not prohibition; it is transparency. People should be able to prove, when it matters, that their writing is theirs.

One Approach: The Seal of Humanity

The Seal of Humanity is LyteWriter's attempt to build a piece of this infrastructure.

It works through two paths. For handwritten or typewritten documents, LyteWriter captures photographs of the physical pages. The handwriting itself is biometric evidence (pressure patterns, letter formation, natural variation) that ties the text to a human hand. For text composed digitally, LyteWriter's editor captures keystroke dynamics: the unique timing patterns between keystrokes that form a behavioral fingerprint no AI can replicate.

Both paths produce the same result: a cryptographic certification linking a specific text to verified evidence of human authorship. The Seal is publicly verifiable at lytewriter.com/verify. Anyone can check it. No account required.

This is not a detector. It does not analyze text and make a probabilistic guess. It captures evidence at the point of creation and makes that evidence independently verifiable. The difference matters. A detector says "this text is probably human." A Seal says "here is the evidence that this text was written by a human, and you can verify it yourself."

The Future Is Not AI or Human. It Is Both.

The anxiety around AI writing often frames it as a replacement story: AI writes, so humans stop writing. That is not what is happening and not what will happen.

What is happening is a sorting. AI will handle the writing that nobody wanted to do by hand anyway: boilerplate, summaries, first drafts of routine documents, templated content. Human writing will persist everywhere that authorship, perspective, and authenticity matter, and those contexts will value human writing more than ever, precisely because AI made the alternative so easy.

But this sorting only works if the market can tell the difference. If human writing and AI writing are indistinguishable and unverifiable, the premium for human work collapses. The incentive to write, to do the slow, difficult, deeply human work of putting your own thoughts into your own words, erodes.

Verification is the infrastructure that makes the future of human writing viable. Not by fighting AI, but by giving humans a way to prove their work is real.

The Line to Remember

AI writing is not going away. Human writing is not going away either. What is going away is the assumption that all writing is human.

We need to replace that assumption with proof.

The Seal of Humanity is one way to start. Whether you write by hand, type on a typewriter, or compose in a digital editor, the Seal gives you a way to say, with cryptographic certainty: I wrote this. Verify it yourself.

For a deeper look at where this is heading, read what happens when you can't tell if anything was written by a human. For a personal perspective on this choice, read why our founder — an AI researcher — writes on a typewriter.