Imperfect

ai-friendly if communities

I have some thoughts on the exchange between Harris and Stressed Friend in IF Seal: I used an LLM to make IF and now I feel stressed and conflicted about it!.

AI code isn't foolproof, nor is it always problematic. LLMs can be wrong often, so augment them where correctness matters. Employ other tools, people, and verification. These and more resources can help you understand code or get over the finish line when you feel overwhelmed. Practice all of the above toward a complete performance.

To me, being demonized for using tools like AI is a reason to find a welcoming community over enduring even an ounce of stress. Confusion and lack of belonging, arising from community vitriol instead of dysfunctional code, isn't worth the time. You aren't alone in your journey. Find solace in people creating like experiences. Which interactive fiction communities are amenable to generative AI use such that they can better serve your creativity, health, and livelihood?

I'd rather see virtual spaces created than prevented if they can be healthily maintained. Starting small, tempering expectations, and shipping as early as possible tie into that. However, for those 95% done with a project that hasn't shipped, I suggest shipping it, then closing the loop. This sets a solid foundation that can be built upon with future projects.

Finish, publish, then market the games you started designing. These steps help you understand how to navigate and improve upon each phase a developed game requires. Take whatever profits you can from Patreon or otherwise, then figure out a sustainable path forward. Your fullest creative expression could be cheaper and quicker than you expect.

Publish a blog chronicling the development, launch, and aftermath of your game. Learn how to explain your process whether it involves AI or not. Doing so means you can do it that much better next time. Blogging and other communication methods could also help you find a community receptive to your game or helping out with it. You don't have to settle for locals of opposite polarity. This is the Internet after all.

Everyone starts somewhere. Many products fail to succeed yet teach a thousand lessons. Impart those into your next prototype and wonder just how much higher and further your next game can reach.

As someone finding both interactive fiction and AI intriguing enough to comment on, I'll end with some questions on their intersection:

Want to reach out? Connect with me however you prefer: