Abstract
We consider the problem of creating general-purpose Minecraft agents capable of solving a wide range of goals in a complex environment as a testbed for studying hybrid neural-symbolic architectures for Artificial General Intelligence (AGI). We analyze the desirable behavior of such agents and sketch out an architecture for it. We implement a prototype of the agent, which is capable of achieving various goals in the Minecraft world and to perform exploration, and discuss the utility of more advanced AGI components to be developed and integrated into the agent in future.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Perez-Liebana, D., et al: The multi-agent reinforcement learning in MalmÖ (MARLÖ) competition. arXiv:1901.08129 (2019)
Guss, W.H., et al.: The MineRL 2019 competition on sample efficient reinforcement learning using human priors. arXiv:1904.10079 (2019)
Shah, R., et al: The MineRL BASALT competition on learning from human feedback. arXiv:2107.01969 (2021)
Goertzel, B.: Toward a formal characterization of real-world general intelligence. In: Proceedings of 3D Conference on Artificial General Intelligence, pp. 74–79. Atlantis Press (2010)
Zhu, M., et al.: MapGo: model-assisted policy optimization for goal-oriented tasks. arXiv:2105.06350 (2021)
Legg, S., Hutter, M.: Universal intelligence: a definition of machine intelligence. Minds Mach. 17(4), 391–444 (2007). https://doi.org/10.1007/s11023-007-9079-x
Badia, A.P., et al.: Agent57: outperforming the atari human benchmark. arXiv:2003.13350 (2020)
Hutter, M.: Universal Artificial Intelligence: Sequential Decisions Based on Algorithmic Probability. Springer, Heidelberg (2005). https://doi.org/10.1007/b138233
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Potapov, A., Belikov, A., Scherbakov, O., Bogdanov, V. (2023). General-Purpose Minecraft Agents and Hybrid AGI. In: Goertzel, B., Iklé, M., Potapov, A., Ponomaryov, D. (eds) Artificial General Intelligence. AGI 2022. Lecture Notes in Computer Science(), vol 13539. Springer, Cham. https://doi.org/10.1007/978-3-031-19907-3_8
Download citation
DOI: https://doi.org/10.1007/978-3-031-19907-3_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-19906-6
Online ISBN: 978-3-031-19907-3
eBook Packages: Computer ScienceComputer Science (R0)