Rage Against the Machine, Not the Author

I do not write with AI. My sentences are mine. My bad sentences are mine too. The strange metaphors, unnecessary tangents, midnight revelations that look considerably less revelatory in daylight, unfortunately all mine. I care about that distinction because writing, to me, is thinking made visible.

But I also think the growing witch hunt around AI and creative writing is becoming intellectually lazy.

Panic Is Not a Standard

Publishing has a legitimate problem: generative AI can produce prose, characters, dialogue and entire narratives. If you prompt a machine to create the substance of a novel and then sell the result as something you wrote, there is an obvious question of authorship. That question matters. Panic does not help answer it.

We are already watching careers collide with this uncertainty. Jerry Falade’s debut crime novel, Call Me, I’ll Hide the Body, reportedly received an offer exceeding $2 million during a 14-way U.S. auction before his agents withdrew it. They said they could no longer authenticate how the manuscript evolved from beginning to end. Falade denied using AI (Creamer, 2026).

Mia Ballard’s Shy Girl met a similar fate. Hachette canceled its planned U.S. publication and discontinued its U.K. edition after investigating allegations of generative AI use. Ballard said she did not personally use AI and suggested problems may have arisen during editing (Trachtenberg, 2026). H.M. Wolfe’s Daggermouth has also faced AI allegations, which Wolfe denied. The larger publishing industry still has no universal standard defining precisely where acceptable AI assistance ends and unacceptable AI authorship begins (Trachtenberg, 2026).

That distinction is everything.

The Machine Can Be a Tool Without Becoming the Author

If I ask AI to write a novel about a widowed astronaut discovering God inside a black hole, I did not write that novel. I commissioned one. Maybe the premise was mine. Maybe I selected the output and rearranged some sentences afterward. But if the machine invented the characters, dialogue, prose and narrative architecture, calling myself the sole author becomes difficult.

But suppose I write 100,000 words myself and ask a machine to find continuity errors. Suppose I ask it to organise research, locate contradictions, explain a grammatical problem or challenge an argument I have already made. The machine has not become the author. It is a tool.

Those two things should not be treated as morally or creatively equivalent.

Human beings have always extended their minds through technology. Dictionaries compensate for memory. Calculators handle arithmetic. Spellcheck catches mistakes. Search engines transformed research. Word processors transformed editing. Every generation eventually decides that the tools it grew up with are perfectly natural while whatever came afterward constitutes cheating.

That is nostalgia disguised as ethics.

The Part I Refuse to Outsource

The human mind may be the greatest creation in our universe. It experiences, remembers, suffers, loves, desires and imagines. It can take childhood, heartbreak, mathematics, death, some stupid conversation overheard fifteen years ago and sunlight coming through a kitchen window and turn them into something that has never existed before.

That is the part I refuse to outsource.

There is a failure in asking AI what to believe, what to imagine, what argument to make and then what words to use to make it. Eventually there isn’t much of you left. If every difficult act of thought is surrendered to the machine, the machine may become more useful while the person becomes less capable.

But there is another kind of failure: possessing one of the most powerful intellectual tools humanity has created and refusing to touch it because someone might accuse you of cheating.

A telescope does not invalidate astronomy. A microscope does not make the biologist fraudulent. A calculator does not make the mathematician a fraud. The meaningful question has never been whether a tool was involved. The question is what intellectual work the human surrendered to it.

Suspicion Cannot Become Evidence

Publishing should protect human authorship. Publishers have every right to establish contractual rules around generative AI. Readers deserve honesty about what they are buying. Writers deserve protection against having their work copied, imitated or replaced without consent.

But suspicion cannot become evidence.

Especially when the accusation amounts to saying something “sounds like AI.” Generative AI was built to imitate patterns in human language. We have therefore created the wonderfully absurd possibility of teaching machines to imitate humans and then accusing humans of being machines because the imitation became convincing.

We built the parrot and are now interrogating ourselves for sounding like the parrot.

AI-detection systems should therefore never become literary polygraphs. Publishing itself remains divided over their reliability and appropriate role, and the industry has not settled on a universal definition of permissible AI assistance (Trachtenberg, 2026). The Shy Girl controversy alone has produced arguments over whether detector results should carry enough weight to damage an author’s reputation or career (Taranto, 2026).

Once someone’s name becomes associated with “AI author,” proving innocence becomes almost philosophically impossible. How exactly does a writer prove that they imagined something?

Drafts help. Notes help. Version histories help. Research records help. A publisher can reasonably ask for evidence when substantial questions arise. But eventually we arrive at the private country where writing actually happens.

The mind.

There is no audit log for imagination.

Draw the Line at Authorship

So draw the line at authorship, not mere contact with technology.

If AI creates the work and you claim you created it, there is an authorship problem. If AI substantially generates your prose, disclosure is reasonable. If it researches, organises, proofreads, checks continuity or challenges something you already created, that is something fundamentally different.

The distinction isn’t whether AI touched the process.

It is whether AI became the process.

I don’t write with AI because I want the struggle of writing. I want to stare at a sentence, hate it, delete it, rewrite it, walk away and suddenly understand three hours later why it wasn’t working. That frustration is part of creation. The difficulty is not an unfortunate obstacle between me and the finished work. Sometimes the difficulty is where the thinking happens.

But my choice does not need to become someone else’s commandment.

A dyslexic writer may use AI to examine grammar. Someone writing in a second language may use it to interrogate syntax. A researcher may use it to organise thousands of pages of notes. A novelist may use it to search a massive manuscript for continuity problems.

None of those acts inherently surrender authorship.

The alphabet is technology. The book is technology. The printing press is technology. The computer is technology. AI is technology.

The ethical question is what we choose to surrender to it.

Do Not Confuse the Hammer With the Carpenter

Protect writers. Protect human creativity. Protect copyright. Require disclosure when authorship genuinely becomes ambiguous. Punish deliberate deception when it can actually be demonstrated.

But don’t create an inquisition around the mere use of a tool.

The greatest creative instrument we know is still sitting between our ears. Failing to use that mind would be the greater tragedy. Technology should make its reach longer, its questions deeper and its possibilities greater, not make the mind unnecessary.

Use the machine.

Do not surrender yourself to it.

And don’t confuse the hammer with the carpenter.

References

Creamer, E. (2026, July 31). $2m crime novel deal collapses amid questions over AI use. The Guardian.

Taranto, J. (2026, April 4–5). The “AI detector” as defamation machine. The Wall Street Journal.

Trachtenberg, J. A. (2026, March 20). Publisher pulls “Shy Girl” horror novel after AI allegations. The Wall Street Journal.

Trachtenberg, J. A. (2026). AI has plunged the book publishing industry into utter chaos. The Wall Street Journal.