The Writer Is Not the Pen: What the AI Debate Is Still Missing
Listening to Kerry O’Brien’s Thea Astley Address, Lessons from History, delivered at the Byron Writers Festival.
“Acclaimed journalist Kerry O'Brien has seen it all, but nothing like right now. And he's not mincing his words. On AI, neoliberalism, education, equality, gutless governments, and more. For over 50 years, he covered the world's biggest stories and change-makers, and asked tough questions of power. Hear why he thinks a perfect storm threatens our society and our politics, and what he thinks can be done about it.” – ABC Radio National
I recently listened to Kerry O’Brien’s Thea Astley Address, Lessons from History, delivered at the Byron Writers Festival. It is a brilliant speech, and one I think everyone should listen to.
The legendary journalist Kerry O’Brien at the 2026 Byron Writers Festival.
Photo Jeff Dawson
O’Brien moves across decades of technological change, neoliberalism, declining government capacity, inequality, concentrated corporate power, social media, journalism, artificial intelligence and the future of democracy.
Much of it deserves serious attention. This is a journalist who has spent decades with his boots on the ground, interviewing political leaders, watching institutions change and seeing technological disruption arrive in successive waves. His concern about the concentration of technological power in the hands of a small number of extraordinarily wealthy corporations should not be dismissed as fear of change. Nor should his concern about governments arriving late.
He talks extensively about workers potentially being displaced, automation, robots, factories and the meaning people derive from work. That concern is legitimate. But as I listened, I kept returning to the other side of the equation, one that receives far less attention in the speech:
What happens when individuals suddenly acquire capabilities previously available only to institutions?
O’Brien argues that digitisation and neoliberalism have developed alongside one another over the past half-century. One radically increased technological capability while the other, in his view, diminished the capacity of governments to regulate markets and act decisively in the public interest. Artificial intelligence is now emerging into that environment. I do not think he is wrong about that. I think it should be taken very seriously.
The writer and the pen
Near the beginning of O’Brien’s address, he quotes Australian novelist Thea Astley, and for me it opened another door entirely.
“To be a writer has nothing to do with pen and paper.”
Astley locates writing somewhere much deeper, in the ability to see what others do not see, to hear what is not being said, and to possess the emotional capacity to confront what you find.
I immediately thought: what an extraordinary quotation to use in a speech partly concerned with artificial intelligence.
Because if the writer is not the pen, what exactly is the writer?
A lot of what O’Brien says about AI is important. The compelling tension, for me, is that he is viewing it predominantly from the outside, through the history of technology, government, labour, concentrated power and journalism. What is largely missing is the view from someone sitting down with the technology and discovering what it allows an ordinary person to do.
O’Brien does not establish whether he personally uses generative AI. Listening to the address, however, he approaches AI primarily as a societal and political phenomenon rather than exploring the lived experience of people using it as a creative, practical and learning tool.
AI & Author Kevin Tumlinson
One example appeared beneath a discussion by author Kevin Tumlinson. Kevin Tumlinson is a bestselling independent author and publisher who says he has sold more than 10 million books over a twenty-year career. Rather than rejecting artificial intelligence outright, he has spoken openly about using it as an assistive creative tool while maintaining that the writer, not the technology, remains responsible for the work. More recently, his discussions have also turned toward writers with disabilities and others for whom AI can provide access to forms of expression that were previously difficult or impossible.
A commenter living with multiple sclerosis described how AI had helped her finally put into words stories that had existed in her head since 1981. Her frustration was not that she lacked imagination or understood storytelling poorly. The barrier sat between what she knew and her ability to express it fluently.
Tumlinson’s response was striking. He told her he was glad she had found a tool that allowed her to share those stories, then added: “That’s an important aspect of this that I think we’re all overlooking. I know I was.”
That exchange captures something easily lost in arguments about whether AI makes writing less authentic. The technology did not supply this woman with forty-five years of imagined worlds. They were already there. What changed was her ability to participate.
Ask people who are working extensively with AI right now, and I do not mean asking ChatGPT to tidy an email. What keeps emerging is not simply automation. It is access.
A person without a software department can build software.
Someone without a research assistant can interrogate large bodies of information.
Someone without money for constant tutoring can have something resembling personalised tuition.
Someone without access to publishing professionals can learn structure, editing, marketing and production.
A tiny regional organisation can access analytical and administrative capabilities that once required several employees.
A person with a disability or language barrier can participate in work that was previously harder to access.
Someone geographically isolated from universities, professional networks and metropolitan institutions can participate in conversations that previously happened somewhere else.
This is not a minor side effect of the technology. It may be one of its defining effects.
The reason it is so easy to miss is that everyone is using AI differently. How it enters one person’s work and life may look nothing like how it enters another’s. A journalist, a novelist, a software developer, a student, a disabled person, a small business owner and a regional arts organisation can all be talking about “AI” while describing almost entirely different experiences.
That is why I think users need to be in the room when we talk about what this technology is becoming.
Two instincts can be true
O’Brien has developed a veteran instinct through a long career spent navigating extraordinarily difficult social and political environments. He sees a powerful technology, can easily imagine its next evolution, and has a well-documented history of watching weak governments fail to contain concentrated corporate power. From that vantage point, the danger is not difficult to see.
My instinct has always pulled in another direction. Technology can also give us mere mortals widespread access. It can increase participation and agency. Give people more knowledge, more capability and more ways into systems, and it becomes harder for gatekeepers to hold back growth and innovation.
I think both instincts can be right.
That is actually one of the central questions of the next decade:
Does AI become primarily a technology of concentration, or a technology of distribution?
Does it allow already powerful institutions to do more with fewer people?
Or does it allow millions of ordinary people to do things that previously required institutions?
Probably both. Public policy may substantially determine which side dominates.
Interestingly, O’Brien eventually arrives at something very close to this question of agency when he talks about community action. His conclusion is not that people are powerless. He points to grassroots political movements and argues that communities retain the capacity to change systems themselves.
The tool is not the author
Yet the arrival of AI has somehow revived an idealised image of authorship in which the “real writer” is romanticised, sitting alone at an Ernest Hemingway style Underwood vintage typewriter, personally generating every sentence without external assistance. Apparently, if the story did not arrive like that, send it back until the purists are satisfied.
That has never been the complete story of publishing.
We have already accepted enormous amounts of assistance around creative work:
Structural editors.
Copyeditors.
Researchers.
Proofreaders.
Translators.
Writing groups and, sometimes, ghost-writers.
Publishing teams that help transform manuscripts into finished books.
What happens when capability moves downwards, to us small folk? What I hear much less often in these discussions is the word participation. That is not a minor effect. It is potentially one of the defining effects of the technology. And it is largely absent when AI is discussed only as something being done to us.
The existence of assistance does not, by itself, answer the question of authorship. We still ask who made the decisions, who supplied the ideas, who understood the material, who accepted or rejected suggestions, who shaped the characters, who decided what mattered, and who takes responsibility for the finished work.
That is why Astley’s line matters so much to me. If writing has never been reducible to the physical instrument used to produce words, then we need a more serious conversation than simply asking whether a machine touched the page.
“AI can execute your vision, but it cannot give you one.”
We can start to describe three emerging groups:
Those who reject AI
Those who outsource too much of their thinking to it
And those who use it extensively while retaining the very capacities that make human judgment valuable, curiosity, creativity, taste and critical thought.
That distinction matters. Capability is not the same thing as authorship. Execution is not the same thing as judgment. A tool may help someone research, design, write, code or market at a level previously requiring several specialists, but it still requires a person to decide what should exist in the first place.
Human beings keep moving the boundary
O’Brien cautions against casually comparing AI with the printing press or the Industrial Revolution. His argument is that the speed is radically different, and that earlier machines did not threaten to reproduce intellectual capability in the way AI potentially does.
He is right about the speed. We can all feel it.
But I am less convinced that intellectual displacement itself is unprecedented.
The printing press automated work once performed by skilled scribes.
Photography transformed professions built around visual representation.
Recorded music transformed live performance.
Calculators automated mathematical labour.
Desktop publishing eliminated whole layers of typesetting and production work.
Search engines radically altered research.
Each technology removed some forms of expertise while simultaneously allowing vastly more people to participate. Artificial intelligence is broader and faster than those technologies, perhaps radically so. But the historical lesson cannot simply be that automation destroys human value. Human beings repeatedly move the boundary of what we consider valuable, skilled and worth doing.
Writers sacred identity
Why is it that AI seems to shift the foundations of literature more than any other industry?
WARNING A load-bearing ego is a structural feature commonly found in literary institutions. Removal may cause sudden gatekeeper collapse.
People judge a new tool by the labour it appears to remove, before understanding the new labour, judgement and possibilities it introduces. The particularly confronting thing for writers is that AI does not merely make the pen faster. It challenges their authority to decide who qualifies as a writer at all.
The barriers were part of the status system. Years of training, industry connections, editors, agents and publisher approval helped establish who could call themselves a writer. AI allows “mere mortals” to move around some of those barriers.
That is why Astley’s line is so interesting: “To be a writer has nothing to do with pen and paper.” If that is true, then authorship cannot ultimately reside in the physical act of producing sentences. It has to reside somewhere deeper, in perception, intention, imagination, judgement, selection, emotional understanding and responsibility for the completed work.
If the writer is not the pen, why does changing the pen threaten writers so profoundly?
Film has always been visibly collaborative. A director uses cameras, lenses, editors, composers, stunt teams, visual-effects artists and hundreds of technicians. Nobody imagines Christopher Nolan personally building the IMAX camera and developing every frame in a bathtub.
Music also has an accepted technical infrastructure: instruments, microphones, producers, synthesisers, sampling, Auto-Tune, mixing and mastering. There have been enormous fights over each innovation, but the public can still distinguish the artist from the equipment.
Writing carries this mythology:
One solitary mind. One blank page. One authentic voice. Everything came directly from the writer’s soul through the fingers.
So when AI enters that particular room, many writers do not experience it as a new production tool. They experience it as something trespassing inside the boundary of the self.
We need users in the room
This is what I find missing from many public discussions about AI. I am hearing a lot from top-tier authorities. The speakers are often excellent journalists, authors, academics, lawyers and policymakers. They bring decades of experience that should be respected. But the conversation often reaches the edge of the technology and stops.
We hear what AI might do to society.
What it might do to employment.
What corporations might do with it.
What governments should do about it.
All important questions.
But it is equally important to make room for people who are actually using these systems extensively, not merely building or selling them, but learning with them, creating with them, testing them, arguing with them and discovering where they fail.
Those people see another part of the picture. They see the hallucinations, the limitations, the maddening mistakes and the need for human judgement. But many also see something surprisingly hopeful.
Participation.
Regulation needs better nouns
Both futures can exist simultaneously. AI can help powerful organisations remove workers, reduce costs and consolidate control. It can also allow individuals to acquire abilities previously locked behind money, geography, education and institutional access.
This is where regulation becomes far more interesting than blanket prohibition, because the category itself is too broad.
We should be able to distinguish between:
Someone using AI to understand a difficult historical document and someone using it to manufacture thousands of fake political identities.
An author using AI during the development of a novel and automated surveillance.
Personalised education and autonomous weapons.
Helping a small organisation write a grant and industrial-scale disinformation.
Protecting creators from unauthorised commercial exploitation of their work and preventing creators themselves from experimenting with artificial intelligence.
Calling all of it simply “AI” obscures more than it explains.
It is entirely possible to support better licensing, attribution, provenance and compensation mechanisms while simultaneously embracing AI as a creative and practical tool. Protecting artists does not require preventing artists from experimenting.
Kerry O’Brien’s conclusion
O’Brien ends by rejecting defeatism and reminding his audience that communities still possess the capacity to change systems themselves. He points to political movements beginning around kitchen tables and citizens discovering that institutions are not immutable.
That is agency.
Will the purity panic pass? I think Artificial intelligence raises profound questions about power. But it may also give individuals access to an unprecedented amount of capability. The challenge is not simply stopping the future from happening. It is deciding who gets to participate in building it.
I largely accept O’Brien’s strongest arguments: concentrated power is dangerous, governments have been slow, misinformation is serious, and creators deserve consideration.
I would simply add the part he does not substantially explore:
What does AI look like from the bottom up rather than the top down?
The question is not whether AI will change society. It is who gets to participate in building what comes next.
"AI is only as clever as its user." - Emma Read
Comments