The Pathfinder: 2nd Edition AI Subreddit
Friday, July 11th, 2025

First, a note to the reader: this is a medium-form post. While the first two paragraphs might seem like a banal description of a subreddit, or perhaps even an *advertisement* to my glass-half-empty readers out there, I promise you more interesting commentary follows. I interleave a curated gallery of images with at least enough words to make the experience feel more like a relaxed yet intentional visit to a museum or a morning hour in the park, rather than a drive-by dopamine hit.
There’s a new subreddit on the block — r/PF2E_AI. The "PF2E" part of its name is short for Pathfinder Second Edition, a table-top roleplaying game ultimately dervied from Dungeons & Dragons. But the forum’s lax rules allow for much more than things purely to do with the TTRPG: anything in the realm of fantasy goes, as long as the media was at least partially generated by artificial intelligence.
Image and video models are the stuff of dreams for any fantasy fanatic. This forum’s top posts are even better. They comprise a curated stream of the best fantasy stuff that internet anons have dreamt up. The beautiful thing here is I can say that literally. The subreddit isn't limited to what people are capable of physically creating or digitally painting. It’s what people are capable of dreaming. The only limiting factor is their own creativity and the capabilities of today’s models. That’s really cool.
Here are some of my favorite posts from the subreddit, along with some general commentary about AI and its effects on the human psyche that I hope you find interesting:

It's interesting how quickly people have accepted that AI can produce work like this. The above image isn't extraordinary. But neither is it completely mundane. Depending on how you, individually you, rate limit the onslaught of information hitting your brain, this image will elicit different effects. The one true thing I can say here is: the less you rate limit, the more mundane things seem on average. Is that okay? It feels not okay. It makes me feel bad. It's like having cheat codes to a video game --- the easier it is to access the end result, the less satisfying it is to achieve. Does satisfaction require hard work? Does salvation? I don't know. But I think there's certainly some truth to the idea. In the words of techno-lizard overlords, the idea is "directionally correct."

Maybe it's just a life hack. Maybe we can increase enjoyment by just slowing down. Seeing this Necromancer Raising an Undead Dragon in 2014 would certainly get my goober juices flowing. If I were a little manic, I'd start making my sixth unfinished Oblivion mod. Or I'd Skype my friends to see what they were up to on this lovely Saturday morning. But here we are, in 2025, where you can create an infinite stream of fantasy dreams.

That generates in my brain another positive perspective on the situation. Back in the day, you'd fry your dopamine receptors, or whatever the proper wording is for the brain chemistry concept we are all too familiar with, by scrolling through the wonders that other people have created. Maybe now, we have it better. Now we can work for our dopamine overload. Now we can try to think of what we want to see before we see it. And we can take the time to type in the prompt and wait a minute for the image to be generated. Sure, the deleterious effects of scrolling have gotten worse. (How the hell is that even possible?) But our options to escape it have strictly increased. With some discipline, we can force ourselves to be the creator instead of the observer. And in being the creator, we break the death cycle content firehose. Was discipline always the answer?

I probably pricked some fingers with the last paragraph. You could always have been a creator rather than a consumer, even before AI. It's just that AI brings the barrier of creation to a new low. Many people who didn't quite have the energy or motiviation or skill to create are now able. I think this also speaks to some of the frustration surrounding AI models today. If we separate the net effect of content produced by AI into content you consume and content you create, we see a striking contrast: the consumption experience has arguably gotten worse. At least as of July 2025, there is more content (bad) and it's shittier on average (bad). But you, as a creator, have strictly more capability than before (good), and the barrier to getting started and coming to a finished product is also orders of magnitude lower, now (very good).
That is all. If you're into fantasy art, you should definitely hit the "Join" button on that subreddit.
Email Updates
You can be notified of new articles by subscribing for free. You can unsubscribe at that page, too.
We also publish an RSS feed.
About
This is a magazine covering the intersection of video games and AI, blemishes and all.
I publish a monthly retrospective and detailed articles about individual games and tools.
Have A Story For Me?
Did you publish a game or tool that leverages AI? I want to know. Please email me at reply@aigamesmag.com.
What I Think About Subscriptions
In today's day and age, no one is gonna pay for a subscription, and it's stupid to try to force one on people. Even the very successful tech news website, Ars Technica, tries to offer a subscription. Why? I'm not here to dissuade people from trying to create something of value. And I think high quality journalism is extremely valuable. But this form of monetization is clearly disconnected from the value provided. Most readers are drive-by web surfers. Even if they end up reading a quarter of your magazine's articles over time, they probably wouldn't pay for a subscription. I think publishers should choose between one of two options: (A) put everything behind a paywall or (B) put nothing behind a paywall and run simple advertisements. I chose to go with the latter option. I don't fault those who go with the former. But if you're running a hybrid strategy, just call the "subscription" what it is — a donation.
Front Page ✦ About ✦ RSS ✦ Email Updates ✦ © 2025 AI Games Magazine