AI Is Not the Problem. Unedited Slop Is.

A practical fix for publishers, authors, and readers

The argument about AI writing is getting loud, but I do not think the real problem is hard to name.

The problem is not that a writer used AI.

The problem is that some people are publishing unedited, low-effort, poorly checked work and expecting readers to treat it the same as human judgment, craft, research, revision, and accountability.

That is where the line should be drawn.

There is a meaningful difference between an author using AI as a drafting partner, research organizer, grammar helper, brainstorming tool, or editing assistant, and someone using AI to mass-produce books or articles they did not meaningfully write, check, revise, or understand. One is assisted authorship. The other is content dumping.

Readers are not wrong to be irritated by AI slop. Publishers are not wrong to worry about trust. Authors are not wrong to worry that their work is being buried under low-quality machine-generated material. The market is already circling around the same basic issue: disclosure, quality, and trust.

The fix does not have to be mysterious.

Publishing already has models for this. Academia has originality reports. Journalism has corrections policies. Book publishing has editorial review. Schools have rubrics. Grant writing has compliance checks. We do not have to invent an entirely new ethical universe just because AI entered the room.

We need standards.

Start With an Originality Report

One practical requirement could be simple: if a manuscript, long-form article, book, curriculum product, or submitted publication used AI assistance, the author should submit an originality report or documentation of originality as part of the process.

That does not mean the work has to be 100 percent “untouched by AI.” That is probably unrealistic now, especially when tools are built into grammar checkers, search engines, word processors, phones, and editing platforms. The better question is this: how much of the work reflects original human thought, judgment, organization, experience, research, revision, and accountability?

A college paper is not judged only by whether the student used a spellchecker. It is judged by whether the student did the thinking.

Publishing should move in that direction.

A reasonable system could require authors to disclose whether AI was used for brainstorming, outlining, drafting, editing, image generation, translation, research support, formatting, or summarizing. The author should remain responsible for every final word.

That is not censorship. That is accountability.

Set a Minimum Human Originality Standard

I have said before that created work should be at least 50 percent original. I still think that is a reasonable starting point for public-facing writing, especially books, educational materials, essays, and paid products.

The exact percentage can be debated, but the principle matters.

If a person cannot honestly say that a substantial portion of the structure, insight, examples, argument, revision, and final judgment came from them, then they should not present the work as fully authored by them.

A minimum originality standard could include:

1. Original thesis or central argument.

2. Human-selected structure and organization.

3. Human examples, lived experience, reporting, analysis, or interpretation.

4. Verified sources where factual claims are made.

5. Human revision beyond accepting machine output.

6. Clear disclosure of AI involvement when it materially shaped the work.

This would not punish responsible writers. It would separate thoughtful AI-assisted work from lazy content dumping.

AI Detectors Are Not Enough

There is one major caution: AI detection tools should not become judge, jury, and executioner.

AI detectors can be wrong. They can flag polished human writing. They can miss AI-generated writing. They can punish people who write in formal, structured, second-language, academic, or highly edited styles. Recent literary controversies have shown how quickly AI accusations can become messy when detector results are treated as proof instead of a screening signal.

So yes, originality tools can help.

But they should not be the whole system.

A better review process would combine:

1. Originality reports.

2. AI-use disclosure.

3. Human editorial review.

4. Source checking.

5. Random audit requests.

6. Revision history when needed.

7. Clear consequences for fraud, plagiarism, or deceptive mass production.

That is much more defensible than pretending one percentage score can tell the whole story.

Require an AI Use Statement

Every serious platform should have a clear AI-use statement option.

Not every use of AI is the same. There is a major difference between:

1. “AI helped me correct grammar.”

2. “AI helped me organize my notes.”

3. “AI generated a rough draft that I heavily revised.”

4. “AI generated most of this book.”

5. “AI generated the cover image.”

6. “AI summarized sources that I independently checked.”

7. “AI created fictional content, characters, or scenes under my direction.”

Readers deserve to know what they are buying or reading.

Transparency builds credibility.

Hiding AI use destroys it.

Create Different Labels Instead of One Big Accusation

Part of the problem is that people are using “AI-written” as if it means only one thing.

It does not.

A better system would use categories, much like food labels, academic contribution statements, or content ratings. For example:

1. Human-written, AI-edited.

2. Human-written, AI-assisted.

3. Human-directed, AI-drafted, human-revised.

4. AI-generated with human review.

5. AI-generated and lightly edited.

6. AI-generated image or cover.

7. AI-assisted research support, sources independently verified.

That is a much more honest framework than pretending all AI use is identical.

A writer who uses AI to clean up dictated notes should not be treated the same as someone who uploads 300 machine-generated books with no meaningful editing. Those are not the same ethical act.

Platforms Should Penalize Slop, Not Tools

The real target should be low-quality, deceptive, unverified, mass-produced content.

Platforms should focus on whether the final work is useful, accurate, original, edited, labeled, and accountable. They should not punish responsible writers for using modern tools any more than they punish writers for using spellcheck, dictation, citation managers, grammar software, or formatting templates.

The standard should be the finished product and the honesty of the process.

A strong platform policy would ask:

1. Is the work original enough to justify publication?

2. Is the author accountable for the final content?

3. Are factual claims checked?

4. Are sources properly credited?

5. Is AI use disclosed when it materially shaped the work?

6. Is the work spammy, repetitive, or mass-produced?

7. Is the reader being misled?

That is where the ethical line belongs.

Not “Did AI touch this?”

But “Did a responsible human stand behind this?”

Publishers Should Audit the Worst Offenders First

The publishing industry does not need to waste time hunting down every writer who used AI to improve a paragraph.

Start where the harm is obvious.

Audit high-volume uploaders. Audit suspiciously repetitive books. Audit fake experts. Audit books with fabricated citations. Audit children’s books with strange AI images and no real editorial care. Audit medical, legal, financial, educational, and safety-related content where bad information can cause real harm.

That is where enforcement should begin.

The worst actors are not hard to spot. They are often publishing too much, too fast, with too little variation, weak editing, fake authority, and shallow content. That is not innovation. That is pollution.

Authors Should Keep Process Records

Responsible authors can protect themselves by keeping a simple record of their process.

That does not need to be complicated. It could include:

1. Notes or outlines.

2. Draft versions.

3. Source lists.

4. AI prompts used for major assistance.

5. Human revision notes.

6. Disclosure language.

7. Final fact-checking steps.

This would help honest writers defend their work if questioned. It would also discourage dishonest writers from pretending a machine-generated dump was a carefully authored manuscript.

A process record is not about fear. It is about professionalism.

Readers Deserve Better

Readers are not asking for perfection.

They are asking not to be tricked.

They want to know whether a book was written by a person with expertise, experience, imagination, research, or lived understanding, or whether it was quickly generated to capture clicks, royalties, search traffic, or algorithmic attention.

That is fair.

Readers spend money, time, attention, and trust. Those things matter.

The writing world should not respond to AI by pretending nothing has changed. It also should not respond by attacking every person who uses AI responsibly. The better path is clear standards, honest labels, originality expectations, and consequences for deceptive low-effort publishing.

AI did not create the slop problem by itself.

Humans chose to publish slop.

Humans can also choose to fix it.

A Simple Standard Going Forward

Here is the standard I would support:

AI may assist the work, but it should not replace the author’s responsibility.

If you publish it, you own it.

If AI materially shaped it, disclose it.

If facts are included, verify them.

If sources are used, credit them.

If the work is mostly generated, label it honestly.

If the work is low-effort slop, do not blame the tool. Blame the decision to publish it.

That is not anti-AI.

That is pro-reader, pro-writer, pro-ethics, and pro-quality.

Robyn L Brown

Robynsnapsback

WeThePeople

For civic literacy, peaceful accountability, community resilience, and repair.