https://www.dagstuhl.de/17471

November 19 – 24 , 2017, Dagstuhl Seminar 17471

Artificial and Computational Intelligence in Games: AI-Driven Game Design

Organizers

Elisabeth André (Universität Augsburg, DE)
Michael Cook (University of London, GB)
Mike Preuß (Universität Münster, DE)
Pieter Spronck (Tilburg University, NL)

For support, please contact

Dagstuhl Service Team

Documents

Dagstuhl Report, Volume 7, Issue 11 Dagstuhl Report
Aims & Scope
List of Participants
Shared Documents
Dagstuhl's Impact: Documents available
Dagstuhl Seminar Wiki

(Use seminar number and access code to log in)

Summary

The video game industry has been developing rapidly in the past decade. Whereas ten years ago video games were almost exclusively aimed at entertainment, nowadays they are used in a variety of places in everyday life. All kinds of organizations now use video games for simulation and training. Educational institutes use video games to enrich and replace parts of courses. Governmental and health care agencies use video games to educate people and stimulate them to lead more productive lives. On top of that, the entertainment-focused games industry continues to grow and is a major industry both culturally and financially.

Two parallel developments can be observed in the games industry. On the one hand, the high-profile entertainment games ("triple-A games") see a steady increase of time and financial resources invested in their development, to keep up with technological advances and to be able to compete in a tough market. On the other hand, the number of smaller, special-purpose games in development (including so-called "serious games") increases dramatically too, in particular in research, training, and education. Moreover, as the pervasiveness of video games increases, so does the the number of people involved in creating them. The job of creating games is no longer limited to specialist programmers and artists. Instead, those who need to use the games become heavily involved in their creation.

The serious-games domain poses additional challenges to game development beyond all the challenges already posed by games for entertainment, namely the need for a strong relation with the "real world". Serious games often have a purpose in training, which entails that the game worlds must be a realistic depiction of the actual environment in which user functions, in particular where "behaviors" are concerned. The big data revolution means that huge quantities of data about the real world are becoming available along with the means of processing them, which may offer the possibility to automatically construct games on the basis of such data. This is a particularly enticing notion, given the financial constraints for constructing serious games, which means that professional content producers might not be available.

Furthermore, new computer games are expected to much better incorporate the different needs of a wide variety of customers, to provide more alternative modes, solution paths, incentives, emotional states, and difficulty levels. Game design, and especially balancing, must take this into account. However, this increases the complexity of design and production considerably, such that AI-based tools that can assist the human developer or even partly automatize processes are more desired than ever.

Summarizing, we note the following four trends in modern game development:

  1. Technological advances have lead to an increased challenge in developing modern video games, even for expert game developers
  2. There is an increased need for non-experts to be able to design and develop games
  3. There is an increased need for realism in the virtual world behaviors, in particular in the area of serious games
  4. A greater variety of players and a better availability of data about players leads to the need for more variable and better customizable games, which require a more complex development process.

A solution for each of these issues can be found in the application of artificial-intelligence (AI) techniques to drive the design and development of games. From the perspective of AI-driven game design, AI supports or even takes over the role of the human game developer in creating particular elements of a game, and even complete new games.

While the game industry tends to use a small selection of well-known algorithms to generate elements of game worlds (in particular where graphics and animation are concerned), the use of AI to create new environments, new behaviors of virtual characters, new narratives, new game rules, or even new gameplay mechanisms is at present limited to a very small number of researchers. We see it, however, as the main direction in which innovation in game design and development can be found.

AI-driven game design sees applications in the design of virtual worlds, virtual characters, narratives, and game mechanics. Moreover, it can be used to assist in the human design process, and to adapt games automatically after publication. Finally, it can support the automated analysis of generated game elements. Each of these topics is a research domain by itself, which requires an interdisciplinary approach which borrows from computer science, psychology, cognitive science, and even the creative arts. A common ground is found in artificial intelligence techniques, in particular machine learning.

For this seminar, we brought together computer scientists and creative experts with the common goals of gaining a deeper understanding of various aspects of games, and of further improving games, in particular by using AI-techniques used to generate games and game elements. The goal was to look beyond what is currently possible and in use, and take steps towards the future of AI-driven game development. Besides theoretical discussions, part of the seminar was spent on trying to achieve first practical results.

Reports on the discussions and results achieved are found on the following pages. All in all, the organizers and participants deemed the seminar a great success, and are eager to continue into some of the directions that were focused on during the week at Schloss Dagstuhl.

License
  Creative Commons BY 3.0 Unported license
  Elisabeth André, Michael Cook, Mike Preuß, and Pieter Spronck

Dagstuhl Seminar Series

Classification

  • Artificial Intelligence / Robotics
  • Modelling / Simulation
  • Multimedia

Keywords

  • Multi-agent systems
  • Game design
  • Serious games
  • Entertainment modeling
  • Dynamical systems

Book exhibition

Books from the participants of the current Seminar 

Book exhibition in the library, ground floor, during the seminar week.

Documentation

In the series Dagstuhl Reports each Dagstuhl Seminar and Dagstuhl Perspectives Workshop is documented. The seminar organizers, in cooperation with the collector, prepare a report that includes contributions from the participants' talks together with a summary of the seminar.

 

Download overview leaflet (PDF).

Publications

Furthermore, a comprehensive peer-reviewed collection of research papers can be published in the series Dagstuhl Follow-Ups.

Dagstuhl's Impact

Please inform us when a publication was published as a result from your seminar. These publications are listed in the category Dagstuhl's Impact and are presented on a special shelf on the ground floor of the library.

NSF young researcher support