Indie Poetry Press

Blog

Aim Your Fire Where It Belongs

I received a DM from a concerned person who told me that a small indie press I support uses AI and had shut down comments after people started raising concerns.

I already knew this. I was the one who suggested they close the comments. Not because they had something to hide, but because sometimes silence is the only way to protect your work from a mob that isn't listening.

I know something about mobs. A couple of years ago I was cancelled myself, over an opinion about a toxic poetry goddess who a year after my pitchfork incident scammed vulnerable women out of thousands of dollars. I know what it is to watch a group of people grow angrier and angrier until the original thing you said has almost disappeared under the outrage about the fact that you said it. I know what a pile-on does to your nervous system. I know how fast a literary discussion turns into a punishment ritual.

I am tired of pretending that this is activism. Sometimes it is just cruelty with political vocabulary.

What interests me about the current arguments around AI is not whether there are legitimate things to be angry about. There absolutely are. Enormous things.

The environmental cost of the infrastructure being built around AI is appalling. Data centres require enormous amounts of electricity and water, and researchers are still struggling to get transparent, reliable data about the full environmental footprint, because the companies running these infrastructures do not disclose enough. One recent study estimated the carbon footprint of AI systems alone could have been somewhere between 32.6 and 79.7 million tonnes of CO2 in 2025, with a water footprint potentially reaching hundreds of billions of litres. The United Nations University has called this a matter of governance and environmental justice, because the benefits can be distributed globally while the burdens, electricity demand, water withdrawal, land use, mineral extraction, waste, are concentrated in specific communities.

There are serious questions about copyright too. Authors, journalists, musicians, photographers and visual artists are fighting some of the largest technology companies in the world over whether copyrighted work can be copied and used to train AI systems without permission or payment. Dozens upon dozens of lawsuits in the United States alone. Courts reaching different conclusions about different aspects of training and fair use. The legal question is still being fought.

Those are real fights.

Artists should be able to ask what happens to their work. Writers should be able to ask whether their books trained a commercial system. Communities should be able to ask why they are expected to absorb the environmental and infrastructural cost of somebody else's technological gold rush. Workers should be able to ask what happens when AI walks into their workplace. People should be able to demand regulation, transparency, consent, compensation, environmental accountability.

I am not arguing against any of that. I am arguing about where we aim the fire.

Because sometimes I look at the way the literary community talks about AI and I think we have completely lost the plot. We are screaming at the person who uses a tool while the people who own the infrastructure build an empire. I don't think that is an accident.

The machine is not the politics

AI is a tool. That does not mean it is an innocent one. It does not mean the technology was built in a politically neutral room, or that the companies behind it are neutral. They are not.

It has been developed inside capitalism, by extraordinarily powerful corporations, with enormous money, energy, data and infrastructure behind it. How it was built matters. Who owns it matters. What data trained it matters. Who profits, who bears the cost, matters.

But the fact that something can be used badly does not tell us what it can only be used for.

A camera can document police violence. It can also be turned into surveillance. A printing press can publish revolutionary literature. It can also print fascist propaganda. The internet can connect workers across borders. It can also become a machine for exploitation and misinformation. A telephone can organise a strike. It can also be how a manager reaches you on your day off to ask where the fuck you are.

The existence of harmful uses does not make the object morally contaminated. It makes the question of who controls it, and how it gets used, more important.

This is where the conversation gets interesting for people like us. Because AI is already being used by ordinary people to navigate systems that were never built for them.

A worker can use it to write emails to management that say what they actually mean without accidentally getting themselves fired. To understand legal arguments when they are trying to work out whether someone is screwing them over. To interrogate policy. To make sense of bureaucratic language. To find the holes in an argument. To prepare questions. To translate things. To organise thought when they are angry enough that their brain has become a bag of bees.

That is AI too.

And there is something politically interesting happening there. Because the people who already have institutional power have always had access to tools that ordinary workers don't. Managers have HR departments. Companies have lawyers. Institutions have policy writers. Governments have armies of civil servants. Corporations have consultants, analysts, researchers, communications teams.

And then someone who works on the floor is expected to walk into a meeting with a manager, completely alone, and somehow articulate the legal, contractual, procedural and ethical reasons why the thing being demanded of them is bullshit.

If a tool can help that person understand their rights, formulate their argument, translate the policy, prepare for the meeting and write the email afterwards, why would I automatically call that politically regressive?

Maybe the tenant should be able to understand the letter from the landlord. Maybe the care worker should be able to decipher the employment policy. Maybe the person who cannot afford a lawyer should at least be able to understand the legal language being thrown at them.

Maybe a group of workers can use these tools to compare contracts and find the problems they have in common. Maybe unions can translate materials into twenty languages, document workplace abuses, or find the one sentence in a policy that management hoped nobody would notice.

That possibility matters to me. Because technology does not have to stay in the hands of the people who built it.

Use their fucking tools

The technology companies would quite like us to believe the future is something they are going to give us. They build the infrastructure. They own the platforms. They control enormous amounts of data. They attract the investment. They employ the engineers. They tell governments what regulation is supposedly possible. They tell us what the future looks like.

And then we stand outside their castle arguing with each other about whether somebody in a poetry group used ChatGPT to help write an email.

No. Look up. The tool is here. It is not going away. We cannot uninvent it.

So the political question is not: How do we make the machine disappear?

The political question is: How do we get our hands on it?

How do workers use it. How do unions, disabled people, migrants and people who cannot afford lawyers use it. How do artists use it without surrendering their rights.

How do we make the infrastructure accountable to the communities living with its consequences. How do we force companies to disclose energy and water use. How do we build copyright systems that protect creators instead of handing corporations another loophole. How do we stop this becoming one more way capital takes power from labour.

And perhaps most importantly: how do we organise around it?

Because the technology is not the movement. People are.

The environmental question is not imaginary

There is a tendency in these conversations to hear criticism of AI's environmental impact and conclude that nobody should use AI. That is not the only possible conclusion. I think it is a dangerously incomplete one.

The infrastructure is already being built. Data centres are consuming electricity and water, at a scale researchers are still struggling to measure. Land is being used, energy systems reorganised, hardware manufactured and eventually discarded. Communities are being asked to accommodate all of it. And the people making those decisions are often not the people who will live with the consequences most directly.

That is a political problem. An environmental justice problem. An infrastructure problem. A labour problem. A question of who gets to decide what gets built, where, who pays for it, who profits from it.

And that is precisely why I don't want to reduce the conversation to whether some care worker somewhere used ChatGPT to write an email. Those are not remotely equivalent political questions.

If you are angry about the water, the electricity, the emissions, the land, be angry about the infrastructure. If you are angry about copyright, organise around copyright. If you are angry about workers being replaced or deskilled, organise workers. If you want transparency, regulation, creators paid, fight for those things directly.

But don't pretend that bullying a small independent poet is the same thing as challenging the corporations building these systems. It isn't.

And yes, copyright matters

This is particularly important to me because I am a writer and publisher. I make books. I publish other people's books. I care deeply about copyright.

The current AI copyright fights are not some imaginary culture-war invention. Authors, publishers, journalists, musicians and artists are suing AI companies over the use of copyrighted work in training systems, and courts are working through questions that have not been conclusively settled.

There are legitimate questions about consent. About licensing. About compensation. About whether creators should have the right to opt out. About what happens when a system reproduces or closely imitates someone's work. Those questions deserve serious political and legal fights.

But again, the answer isn't to pretend that everyone who touches an AI tool is personally responsible for the entire structure of the industry. Someone can be furious about what corporations do with artists' work and still use a language model to help them write an email. Those positions are not contradictory. I think they are perfectly compatible.

You can say: you do not get to take artists' work without consent. And also: people are going to use the technology that now exists to make their own lives a little easier. Those are different questions.

We have done this before

We have a funny habit of imagining that every new technology represents some unprecedented moral rupture. It usually doesn't.

The history of technology is full of tools whose uses became contested because the social structures around them were contested. Photography changed art. Radio changed politics. Television changed advertising. The internet changed publishing. Social media changed communication. Every one of them altered who could make things, who could distribute things, who could reach an audience. And every one created new forms of exploitation alongside new forms of possibility.

The camera did not invent surveillance. Capitalism did not become exploitative the day the internet arrived. And a language model did not invent the exploitation of writers.

The technology changes the terrain. It doesn't abolish politics.

That distinction is important, because if we decide the only politically acceptable response to new technology is rejection, we leave the terrain to the people who are perfectly happy to own it. I don't want to do that. I want the fucking terrain.

Even dynamite has a history

Alfred Nobel developed dynamite as a more practical, comparatively safer way of handling explosive blasting than nitroglycerin, particularly for construction and mining. It was subsequently used for military purposes too.

That history usually gets reduced to a neat little moral story: Nobel invented something for good, and humans turned it toward destruction. But the interesting thing is not whether dynamite was good or bad. It is that a technology can have radically different consequences depending on who controls it, why they use it, and what political and economic structures surround its use.

We don't look at a stick of dynamite and ask whether the object has a soul. We ask what someone is doing with it.

AI is obviously different in countless ways. Vastly more complicated. Its social consequences still unfolding. But the basic political question remains useful. Who has the tool. Who controls it. Who profits from it. Who gets hurt by the way it is deployed. And who gets to use it for something else.

Small presses are not Silicon Valley

This is the part that really gets me. I watch people in literary communities turn on tiny independent presses, poets and artists with very little money or institutional power, while the actual technology companies remain almost abstract.

A small press using an AI tool is not equivalent to a multinational corporation building data centres. A poet using AI to brainstorm is not equivalent to an AI company scraping millions of books. A care worker using ChatGPT to understand a workplace policy is not equivalent to a corporation using automation to eliminate an entire department.

The scale matters. The power relationship matters. The money matters. The ownership matters. The consequences matter.

And if your political analysis cannot tell the difference between those things, it isn't really political analysis. It's purity politics. And purity politics is incredibly convenient for capitalism, because while everybody is busy proving their neighbour is insufficiently pure, nobody is organising against the people who actually hold power.

Purity is a luxury

I grew up working class. My father worked in a factory. My mother was a housewife. We lived below the poverty line.

I have spent my adult life making things with whatever tools I could afford. Yes, some of those tools belong to corporations I don't particularly like. I have published through Amazon. I have used corporate platforms. I use technology made by enormous companies. I live inside capitalism. So do you.

There is no clean little corner outside it where I can build my independent poetry press using only ethically harvested paper, locally manufactured computers, independently hosted servers, perfectly compensated labour and a bicycle-powered printing press. I'd love that world. It doesn't exist. Pretending that it does doesn't make me more politically radical. It just makes it harder to talk honestly about power.

The question is not whether we are implicated. We are. The question is what we do with the things available to us.

I can use Amazon to distribute a book that challenges capitalism. I can use the internet to organise people against the companies that own the internet. I can use a corporate platform to publish something that makes somebody very powerful uncomfortable. And a worker can use an AI tool to help them write an email that lets them stand up to management without setting their own career on fire.

That is not liberation. But it is leverage. And leverage matters.

Aim your fire where it belongs

This is what I wish the literary community would do more often. Stop asking only: who used the tool. Start asking: who owns it. Who built it. Who profits. Whose work was used. Who consented, and who didn't. Who pays for the environmental damage, the electricity, the water. Who lives next to the infrastructure. Who gets the jobs, who loses them. Who has the power to change the rules, and who is making those decisions instead. And what we can do collectively to change that.

Because if you actually care about workers, artists, writers, the environment and democratic control of technology, there is an enormous amount of organising to be done.

You can use AI to write a union leaflet and translate it, analyse a contract, compare employment policies, or make bureaucratic language understandable to the people trapped in it.

Use their tools to organise against them. That is a much more interesting political proposition to me than yelling at an independent poet.

The machine is here

I don't want a literary culture that spends all its energy policing who is allowed to touch this. I want us to stop confusing individual moral purity with collective political power, because those are not the same thing.

A person using ChatGPT to write an email to their manager is not the enemy. A care worker using it to understand their employment rights is not the enemy. A migrant using it to translate a union document is not the enemy. A disabled person using it to navigate a bureaucratic system is not the enemy. A writer using it to organise their thoughts is not the enemy.

The question is what happens next. Does the tool make us more dependent on the people who own it, or can we use it to become more capable of challenging them?

That is the fight I am interested in. Not whether somebody is pure enough to use the machine. Whether ordinary people can get enough power to have a say in what the machine becomes.

And sometimes you just need to shut the comments

Which brings me back to that indie press.

Yes, they use AI. Yes, people are allowed to have serious ethical objections to that. Yes, they are allowed to decide they don't want to buy a book from a press that uses it. That is called having a choice.

But a pile-on is not a political programme. A mob descending on a small press is not climate policy, copyright reform, worker organising, environmental justice, collective bargaining, regulation, or a redistribution of power. Sometimes it is just a bunch of people yelling at someone who is easier to reach than the people they are actually angry with.

I have been on the receiving end of that. So when I told them to close the comments, I wasn't helping them hide. I was helping them protect themselves.

There are times when the most political thing you can do is refuse to perform outrage. There are times when you have to stop explaining yourself to people who aren't listening. There are times when you have to close the fucking comments, and go somewhere your anger can actually do some good. So I'm off to Amsterdam now in an Amnesty International T-shirt. I'm also going to keep writing. I'm going to keep publishing. I'm going to keep making books.

I know fewer people are sending me their poems these days. I understand why. If a stranger is going to hold your work before it exists as a book, you want to know it is safe with them.

So here is what I can actually promise. Your unpublished work is safe with me. I am not feeding your manuscript into anything. I read it the way I have always read submissions, slowly, with my own attention, because that is the only honest offer I have to give.

What I can't promise is what happens once a book is out in the world. Publishing has always meant letting go of a certain kind of control, and that was true before AI existed. It is just true in a different way now. Even a physical book bought secondhand can end up scanned for training data. Anthropic did exactly that with millions of print books it had legally purchased, and a court found the practice lawful. That has nothing to do with piracy, or with who published the book, or how. It is simply a feature of the world a published book now enters.

I don't say that to be alarming. I say it because I would rather be honest about the limits of what any of us can guarantee than pretend a small press can offer a safety that doesn't really exist anywhere right now.

What I can offer is respect while your work is with me, and honesty about the rest.

And I am going to keep asking the much more uncomfortable question: who has the power here? Because that is where I want my anger to go. Not sideways. Upwards. Towards the people who own the infrastructure. Towards the corporations making the decisions. Towards the politicians writing the rules. Towards the systems that make ordinary people fight each other over scraps while somebody else gets richer.

Meanwhile, you can keep telling your mates that Sunday Mornings at the River is an AI-enabling press. I'll keep telling mine that it's a press run by someone who is trying, honestly, to figure this out in public.

Rebecca Rijsdijk