(Belatedly, because I was in a place without good wifi for a few days after the conference, apologies:)
Thursday was the day of Lost Levels, when folk coalesce in the Yerba Buena Gardens for unconference-style talks about subjects dear to them. It’s a good place to find altgames conversations; and as it happens, Kate Compton had organized a Lost Levels section about AI use for altgames purposes, and even put together a zine introducing various tools. So that was pretty excellent, though I wasn’t able to attend all of it, as I also had lunch plans. It made for very different types of discussions from the utility scoring examples, combat and pathfinding solutions that tend to make up most of the main AI Summit.
Later that afternoon, we had a procedural text discussion in the park as well, sort of by accident. I’d put out feelers on Twitter looking/expecting to find maybe three or four people who wanted to have coffee and talk about textual procgen. In practice we had such a large crowd that we couldn’t fit into a coffee shop and had to return to the Gardens and effectively be a Lost Level of our own. (I’d feel bad about the unofficial nature of our participation there, except that the whole point of LL is to be unofficial, so hm.)
In any case, that turned into a set of really enjoyable conversations about what people are looking to do with procedural text, whether that’s representing a complex world model, or focusing on the poetic aspects of the language, or combining text and art in evocative ways. Devine Lu Linvega told us a bit about Paradise; the Sproggiwood/Caves of Qud creators talked a bit about using procedural text in a game that isn’t primarily text based; I talked some about Annals of the Parrigues and also about what I’d like to do with partially procedural dialogue in order to blend emotional aspects better into authored conversation. There were a bunch of other subgroups of the conversation that I didn’t overhear, since there were so many people, but it was great to meet so many people who are into the topic, and I hope to follow up on some of these things more later.
Friday morning Squinky spoke about designing discomfort, giving an overview of many of their own works as well as a survey of others in this general space, and talking about techniques such as realtime dialogue that doesn’t give the player time to contemplate their response (or in which pauses are noted); controls that are intentionally awkward in order to represent social discomforts.
Friday morning there was also a talk about idle games that sparked some interesting discussion, though I wasn’t able to be at that one. (The talk isn’t available yet, and I don’t know for sure that it will be; however, here’s a talk from 2015 that sets some of the stage.)
Friday afternoon, I spoke as part of the Rules of the Game panel, and argued for why we should be thinking about visualization (and visualization tools) early in the process of designing a new system; on the same panel, Liz England gave a great talk about documentation and specifically about knowing the target audience of your documentation so that you can focus on the kind of content and presentation that will work best for them. At the same time, Joe Humfrey was giving an introduction to ink, inkle studios’ newly open-sourced IF tool.
Later Friday afternoon, I also checked out the talk on lost greats from the TRS-80, including Atlantic Balloon Crossing and a few battle/resource-collecting games. I was struck by how much the pleasure of these lies in watching the numbers go up and down and engaging with a simulation as a piece of math. Perhaps this isn’t so far off from the appeal of idle games either.