Part 21 / 21

You Don't Get to Choose My Tools

October 8, 2026

I spent somewhere between five and twelve hours preparing a tabletop campaign.

That is an absurdly imprecise number, but I was not running a stopwatch. I was building a game for people I wanted to spend time with. I was thinking about how it would play, what choices might matter, what a group could discover, where the tension would come from, and how to keep the thing moving when real players inevitably did something more interesting than what I had planned.

I used AI to help me do it.

When I suggested the campaign, someone objected to AI being involved.

And I keep turning that moment over in my head, because I think it exposed a distinction that we are very bad at making.

You can decline the invitation

I want to establish something before we go any further: nobody owes me a seat at my table.

If someone doesn't like AI-assisted work, they can decline to play. They can decide that the environmental cost, labor questions, training-data concerns, or their own personal boundary makes the experience unappealing. Those are real issues, and an invitation is not a subpoena.

What bothered me was something adjacent to that entirely reasonable right.

The objection wasn't about a scene, a rule, a tone problem, an encounter, or a piece of prose someone had actually read. It was about the tool I used while making the thing. The work had been sorted into the unacceptable bin before anyone could encounter the work itself.

I know the familiar reply: tools are not neutral. Fair enough. A choice of tool can have ethical consequences. But that doesn't settle the more interesting question of whether a person who wants to enjoy your work also gets to prescribe the process by which you're allowed to make it.

To me, that's a very strange claim.

I love photography, but not if you use Photoshop

Imagine telling a photographer that you love their photos, but you'll only look at them if they promise never to use Photoshop.

Not that you dislike a particular edit. Not that the retouching made a face look strange. Just that the existence of Photoshop somewhere in the workflow makes the finished photograph unacceptable.

I can understand preferring film. I can understand admiring an unedited photograph. I can even understand an exhibition requiring a particular process. The strange part is making your preferred process a requirement for somebody else's private practice, and treating the requirement as if it were simply aesthetic criticism.

The analogy is not exact. AI systems have different training, labor, and environmental questions than photo-editing software. Those questions deserve serious argument, not a wave of the hand.

But the distinction holds: evaluating a work is not the same thing as controlling its maker's tools.

If the work is bad, tell me it is bad. If it misleads you about how it was made, tell me that. If it crosses a boundary you've set for what you personally consume, decline it.

Just don't pretend you have evaluated a creative work you haven't seen.

The invisible hours

There is another assumption hiding inside these conversations: if AI was involved, the human contribution must have been negligible.

That's not how I experienced making this campaign.

I chose to make it. I was the one trying to give it a shape. I made decisions about how players would move through the experience, what would be worth preparing, and what should remain loose enough to breathe at the table. AI assisted the process, but it didn't supply a reason for the campaign to exist, or a relationship with the people I wanted to share it with.

Those hours count.

They would still count if I had done every piece by hand. They would still count if I had bought a published adventure, remixed a rulebook, used a random-table generator, borrowed a dungeon map, or asked another GM to help untangle a plot hole.

Game masters have always made things with other people's tools. We don't generally treat the use of a monster manual as proof that the GM didn't build an experience.

A campaign is especially strange to judge by its production method, because a tabletop story is not even a finished object when the players arrive. It is a structure for making a story together. I can prepare revelations and consequences; I cannot prepare the precise sequence of things that people will decide to do.

That part only exists when we play.

What didn't happen

Here is the part that stings: the game was supposed to be an invitation.

I was offering hours of preparation, imagination, and attention so that a group of people could have a good time. That isn't heroic, and I don't deserve applause for it. It's just what you do when you love making things with friends.

But it changes how it feels when the response isn't, "Does the game look fun?" It's, "Was AI anywhere in the making of this?"

It makes the labor disappear behind the label.

And what I find myself grieving isn't the right to use a tool. I have that right. I can keep writing, building, experimenting, and publishing. The loss is a possible evening with people, a story that would have changed under their hands, and a little social space where we could have made something together.

That was the point of making it.

So I'm opening the notebook

The campaign was called Message in a Bottle. It was built for Savage Worlds. I have been thinking about publishing the GM-facing material as an archive of the adventure that might have been, with all the spoilers left in. I've started The Game You Didn't Play, an archival companion that will need the original scenario notes before it can honestly disclose the full story.

Not as revenge against a person who declined a game. Nobody deserves to be singled out for deciding what they want to participate in.

As a record of the work.

A GM's notebook is an unusually honest kind of artifact. It shows the seams. What was planned. What could change. Which surprises depended on the players, and which surprises only existed because the GM put them there.

And it makes visible a kind of labor that is easy to dismiss when you have decided in advance that an AI-assisted creation isn't really a creation.

I don't know whether publishing the notes will persuade anybody. I suspect it won't. People who have made up their minds about a process tend to be remarkably indifferent to the objects it produces.

But persuasion isn't the only reason to publish.

Sometimes the most useful answer to "you didn't make that" is to show your work.

The distinction I want to keep

I don't get to decide what you play, read, watch, buy, or invite into your life.

And you don't get to decide what tools I am permitted to use to make the things I offer.

Those two freedoms are not at war unless we insist they must be.

I can respect your no. You can respect my work enough not to erase it.

And somewhere in the middle, perhaps, we can stop having the entire conversation before the first die hits the table.


Process note: This essay was developed with AI assistance and human editorial direction. The linked campaign archive is being assembled from original notes; any planned plot details will be identified as plans, not events that actually occurred at a table.

Keep reading

Keep reading

Get updates from The Singularity Log by email

The Singularity Log currently has 21 installments. Get the next one in your inbox as it lands.

Updates when they’re ready. Free. Unsubscribe whenever.