Project Thyia: A Forever Gameplayer

Abstract

The space of Artificial Intelligence entities is dominated by conversational bots. Some of them fit in our pockets and we take them everywhere we go, or allow them to be a part of human homes. Siri, Alexa, they are recognised as present in our world. But a lot of games research is restricted to existing in the separate realm of software. We enter different worlds when playing games, but those worlds cease to exist once we quit. Similarly, AI game-players are run once on a game (or maybe for longer periods of time, in the case of learning algorithms which need some, still limited, period for training), and they cease to exist once the game ends. But what if they didn't? What if there existed artificial game-players that continuously played games, learned from their experiences and kept getting better? What if they interacted with the real world and us, humans: livestreaming games, chatting with viewers, accepting suggestions for strategies or games to play, forming opinions on popular game titles? In this paper, we introduce the vision behind a new project called Thyia, which focuses around creating a present, continuous, 'always-on', interactive game-player.

Deep Reinforcement Learning: A State-of-the-Art Walkthrough

“…Given the super-human performance on the aforementioned environments, and more specifically, video games, one cannot overlook the onset of an era where competitiveness in gaming shifts from being human-controlled to bot-controlled, a direct consequence of which, for example, is Go champion Lee Se-dol quitting the game (BBC News, 2019). What follows, reasonably of course, is the rise of new challenges in Deep RL and AI in general, which aim at building new foundations and protocols for AI, taking the corresponding countermeasures, and/or fusing the realm of AI with our own reality (Gaina et al, 2019). Figure 1: Heatmap visualization of performance comparison between state-of-the-art modelfree algorithms in Atari games trained using the no-op regime, with respect to human performance taken from the work of Mnih et al (2015).…”

Rolling Horizon Evolutionary Algorithms for General Video Game Playing

“…The algorithm has been previously successfully employed in a variety of noisy optimisation problems, such as tuning game parameters [14] as well as AI game-player parameters [15], [11], [16]. A highly adaptive system which can optimise its parameters and structure so as to achieve best performance in various games could easily feed into a generic life-long learning system such as that presented in [17].…”

Adaptive General Search Framework for Games and Beyond

“…Self-adaptive agents were also identified as an important research challenge during the 2019 Dagstuhl seminar on Artificial and Computational Intelligence in Games [20]. Moreover, the need to look for more advanced artificial game players has also been highlighted by Gaina et al [21], who envisioned the development of an agent that goes beyond classical GGP agents. They propose to develop an "always on" game-playing agent, that learns from experience, continuously getting better and interacting with the world.…”