How I'm Using Technology to DM

I run tabletop RPGs, mostly Savage Worlds these days, and I recently taught a room full of Linux nerds how to play Dungeon World at SouthEast LinuxFest. Somewhere along the way I have accumulated a little pile of tech that I lean on to run games, and since I am the kind of person who cannot leave a workflow un-tinkered-with, I have opinions about what actually helps and what is just shiny. This is not a product review. It is more of a confession about how a sysadmin brain approaches sitting behind the GM screen.

The first thing to understand is that running a game is an improvisation problem. Your players will always, always do the thing you did not prep for. They will ignore the dungeon you built and go talk to the blacksmith you invented as a throwaway line. Most of my tech is aimed squarely at that problem: not replacing prep, but making me faster and calmer when the players go off the map.

Notes are the foundation, and this is the least glamorous and most important part. Right now everything lives in large but searchable Google Docs, one big document per campaign that I can search in a couple of seconds. NPCs, locations, plot threads I dangled and have not resolved, the name of that blacksmith the party has now adopted as a mascot. It is not elegant, but it is fast, and fast is what matters at the table. Eventually I want all of this to live in something purpose-built like Foundry VTT, where the notes, the maps, and the dice all sit in one place instead of scattered across a doc and a tab and a die tray. That is the direction I am slowly moving. For now, the humble searchable doc does the job. The reason any of this matters is continuity. Players remember everything, especially the small stuff you said offhand three sessions ago, and nothing builds their trust in the world faster than you remembering it too. A searchable notes system means I look like I planned things I absolutely made up on the spot, because I wrote down the improvisation the moment it happened and it is now canon I can search.

Digital references are the next layer, and they are mostly about removing friction. Savage Worlds is not a heavy system, but there are still moments where I need a rule fast, and flipping through a physical book while five people wait is a momentum killer. Having the rules searchable on a tablet means I get the answer in seconds and we keep moving. The goal is always to keep the table’s attention on the story and not on me thumbing through an index. Anything that shortens the gap between a question and an answer at the table is worth having.

Now the part people actually want to hear about, which is AI. I use it, and I use it in a specific way that I think matters. I do not use it to run the game or to make decisions at the table. That is my job and honestly the fun part, so why would I outsource it. The mental model I have landed on is that I am the editor-in-chief and the AI is the staff writer. The creative direction is mine. I write the outline, I decide the key scenes, and I set the moments I want the session to hit. That is the part that actually is the game, and I would not hand it to anything. Then I take that skeleton and hand it to AI, Claude these days, and let it build the thing out into a functional session guide I can actually run from. It fleshes out the room descriptions, drafts the NPC dialogue in the voice I sketched, and turns my bullet list of beats into something with connective tissue. The ideas and the structure are mine. The build-out is the part I delegate, and it saves me hours every session.

The other place AI helps is when a player blindsides me. Say the party decides to found a merchant guild I never planned for, and suddenly I need the bones of how that works in my world, right now, while everyone is looking at me. A quick prompt gives me enough of a skeleton to improvise something that feels intentional. I am still the one making it fit the world and the tone, but I am doing it with a scaffold instead of from a dead stop. For a pulpy, cinematic system like Savage Worlds, where the goal is to keep the action moving and worry about the details later, that speed is a real gift.

The honest caveat is that all of this can become a trap. It is very easy to spend an hour fiddling with the perfect notes setup or generating content you will never use, and call it prep when it is really just procrastination with extra steps. The tech is only worth it when it makes the actual game at the actual table better. The moment it becomes the hobby instead of the tool for the hobby, I try to catch myself and close the laptop. The game happens between people rolling dice around a table. Everything on my screen is just there to help me show up ready and stay out of the way of the story.