Hi guys! Long time, no post! (I’m really bad at the “every two months” thing aren’t I…)
So yesterday was my birthday (I’m 24 now wow!) and one of my favorite birthday presents was this news article. Although I’m not sure what my research has to do with business or markets, I don’t mind that they bounced the original story from here. A few months back, a few colleagues and I announced the beginning of a new competition for artificial intelligence called the Generative Design in Minecraft Competition (or GDMC for short). I made a cool website, coded up an easy-to-use framework, submitted a paper about the competition, created a discord channel, and just yesterday we opened up submissions.
IF YOU DECIDE TO SKIP THE REST OF THIS ARTICLE BUT WANT TO COMPETE, a few important things to know:
- Get code here
- Get rules/submit here
- June 30th: Submission Deadline
- July: Evaluation of Submissions
- August 7-10th: Results announced at FDG 2018
Now at this point, if you haven’t looked at any of these links, you’re probably wondering,
“Mike, what is this competition about?”
I’m so glad y’all ask questions.
What is GDMC?
The Generative Design in Minecraft Competition is designed to facilitate the exploration of new possibilities in the realm of 3-dimensional world generation and manipulation. Currently, the competition only has one track: Settlement Generation although we do plan to have more in the future.
The aim of this competition is to advance the state of the art of procedural generation of settlements and game content. The goal is to be able to produce Minecraft settlements of comparable quality to those produced by groups of human builders (see the section on Good Human Examples). In other words, developing an AI to build something that we consider functional AND pretty to look at.
Minecraft itself already features a settlement generator, and there are a few settlement generators provided by mods. But if we are being honest, the vanilla Minecraft settlement generator is terrible. Some of the generators we observe from modders are incredibly creative and build amazing cities and towns. We see chances for improvement towards human-like generation of settlements in the following areas:
- Adaptation to the Environment and Terrain
- Functionality from an Embodied Perspective
- Narrative Integration
- Maintaining Aesthetics
All of these are described below, but if you are looking for more guidance, please check out the website here.
What do you need to participate?
- A copy of Minecraft
- The framework, which is linked above
- A dream
- Memes for inspiration
Adaption to the Environment and Terrain
Real life human settlements are adapted towards their surrounding environment and terrain in multiple different ways. If there’s a river nearby, I might build a bridge across it. If the climate is cold, I would build insulation into my house. Cities and towns also might change the surrounding environment to their needs as well. Rather than building a road over a mountain, I might build a tunnel through it. Or I might dam a river to protect my city against a flood. In summary, settlements shape and in turn are shaped by their surroundings. As part of this challenge, we encourage participants to develop generators that produce different settlements for different kinds of maps, reflecting the available materials, surrounding terrain and biome conditions.
Functionality from an Embodied Perspective
Minecraft is not just a creativity tool, but also a game. A Minecraft settlement is not just an aesthetic artifact; it also needs to fulfill a functional role (i.e. people live here). Can I walk around the town? Are there doors to enter buildings? If buildings are multi-story, are there ways to go up and down floors? Are there parts of the city that are completely inaccessible to a player? What about food and water, how would I live and stay alive if I habited this place? Another challenge in this competition is to produce a generator that can ensure that the settlement provides a maximum of functionality and affordances for the player.
Human settlements are more than just collections of functional buildings. Each city tells a story about its history, who are the people that live there, and how these people see the world. When we look at human created Minecraft settlements, we can often see how certain human or imagined cultures are reflected in the created buildings. There are cities that resemble ancient Rome, mythical Elven forest outposts, and modern US cities. Often, the builds also reflect very clear narrative ideas, such as “this is a defensive mining outpost built in a harsh environment”, or “this is a capital city, built to impress foreign and domestic visitors alike”. The best settlements are entirely evocative, i.e. they manage to transport this story by looks alone. A third challenge we propose is to allow the generator to create this evocative narrative as it builds the settlement.
This is a hard one to describe for me (as I’m not an artist or designer, ask anyone..). I going to quote our paper about this instead, which you can find here:
Aesthetics are arguably subjective, yet architects and city planners usually follow a range of principles when it comes to designing a settlement or a house. While untrained humans manage to intuitively realize these principles with some success, it is difficult to design an algorithm with automated aesthetic judgment. Existing Minecraft mods that add structures to the world circumvent this problem by hand-designing appropriate templates, and then building a settlement out of those templates. This solution has several problems: it allows for little variation between the buildings, as they all need to be pre-designed, and it also allows for little adaptation of the buildings to the surrounding buildings, the underlying terrain, etc. Buildings are sometimes parameterized, and hence reflect some specific environmental settings, such as available materials or the climate they are in. This makes them more adaptive, but there seems to be trade-off between controlling the exact look of a structure, and making it adaptable towards uncontrolled environmental factors. The challenge in our case is to ensure that buildings still follow basic design principles while being adaptive, functional and evocative.
There is also the further problem that the overall settlement itself should follow certain aesthetic rules. How buildings are aligned and what sight lines exist can play an important role for the overall feel of the settlement. Finally, there is also the question of how well the particular aesthetic expression chosen aligns with the other challenges. For example, a settlement which has a narrative of being a foreboding fortress, and is designed with that functionality, should also have an aesthetic that reflects this.
That’s all I got..
I have a bunch of new work coming out for the Foundations of Digital Games (FDG) Conference 2018 so I will be probably blogging about that in the coming weeks. For now that’s all I have to write about. I hope some of you try out the competition! You have a month left and can ask all your questions in the discord channel! Good luck!