The Cognitive Imitation Games

What We Do:

The aim of the Cognitive Imitation Games is to design a comprehensive framework to evaluate the level and generality of Artificial Intelligence (AI). In the cognitive imitation games, AI need to perform just like or outperform a normal human being in different cognitive tasks, while at the same time, the model is cognitively and biologically inspired by human brains at multiple levels of details.

The Cognitive Imitation Games are not designed to evaluate narrow AI. Instead, they are designed to evaluate the generality of AI towards Artificial General Intelligence, and ultimately Superintelligence. Hence, frameworks in Cognitive Psychology on the multiple perspectives of intelligence (e.g. Multiple Intelligence Theory, and Triarchic Theory of Intelligence) are considered to be source of inspirations.

Cognitive Imitation Games are not only about evaluations of how general and intelligent the AIs are, but also emphasize on how the AIs achieve their intelligence and generality.

What We Do Not Do:

We do not want to completely evaluate a simulation of the whole brain at multiple scales (which could be considered as the main goal of computational neuroscience). We only simulate the cognitive functions and computational mechanisms with essential inspirations from human intelligence.

We do not simulate negative emotions, such as envy, hate, and hostility, from human being. Since bringing these emotions and cognitive functions to AI are dangerous, even very possible to create existential risks to humanity.

A Sketch of the Cognitive Imitation Games

Note: The Current Sketch of the Cognitive Imitation Games is mainly based on Multiple Intelligence Theory, and Triarchic Theory of Intelligence of human, with extensions from the perspective of AI. We are in the process of sorting out several hundred cognitive functions to complete this sketch and provide evaluations over different perspectives and levels of intelligence of AI.

A Framework Towards Biologically-inspired Artificial General Intelligence

The way AIs achieve their intelligence and generality is even more important than how general and intelligent the AIs are. The upper picture set a initial framework to achieve biologically-inspired Artificial General Intelligence.

The Cognitive Imitation Game Initiative

The Initiative is composed of scholars from different field of studies. Besides Artificial Intelligence, we are also from Cognitive Psychology, Social Psychology, Cognitive Neurosciences, Philosophy, Neurosciences, Physics, Computer Science, etc. You are very welcome to join the initiative.

Yi Zeng. Institute of Automation, Chinese Academy of Sciences, China.
Kenji Doya. Okinawa Institute of Science & Technology, Japan.
Zhe Sun. RIKEN Center for Advanced Photonics, Japan.
Shimon Edelman. Cornell University, USA.
Piero Scaruffi. Stanford University, USA.
Ian Max Andolina. Institute of Neuroscience, Chinese Academy of Sciences, China.
Yanjie Su. Peking University, China.
Zhi Li. Zhejiang University, China.
Yi Jiang. Institute of Psychology, Chinese Academy of Sciences, China.
Yulin Qin. Shanghai Jiaotong University, China.
Bailu Si. Beijing Normal University, China.

Related Work from the Team Members

Hui Feng, Yi Zeng, and Enmeng Lu. Brain-Inspired Affective Empathy Computational Model and Its Application on Altruistic Rescue Task, Frontiers in Computational Neuroscience, 2022.

Zhuoya Zhao, Enmeng Lu, Feifei Zhao, Yi Zeng, and Yuxuan Zhao. A Brain-Inspired Theory of Mind Spiking Neural Network for Reducing Safety Risks of Other Agents. Frontiers in Neuroscience, 2022.

Yi Zeng, Yuxuan Zhao, Tielin Zhang, Dongcheng Zhao, Feifei Zhao, and Enmeng Lu. A Brain-Inspired Model of Theory of Mind. Frontiers in Neurorobotics, 2020.

Qian Liang, Yi Zeng, Bo Xu. Temporal-Sequential Learning with a Brain-Inspired Spiking Neural Network and Its Application to Musical Memory. Frontiers in Computational Neuroscience, 2020.

Feifei Zhao, Qingqun Kong, Yi Zeng, Bo Xu. A Brain-inspired Visual Fear Responses Model for UAV Emergent Obstacle Dodging. IEEE Transactions on Cognitive and Developmental Systems, 12(1): 124-132, 2020.

Yi Zeng, Yuxuan Zhao, Jun Bai, Bo Xu. Towards Robot Self-consciousness (II): Brain-inspired Robot Bodily Self Model for Self-Recognition. Cognitive Computation, 10(2): 307-320, Springer, 2018.

Yi Zeng, Guixiang Wang, Bo Xu. A Basal Ganglia Network Centric Reinforcement Learning Model and Its Application in Unmanned Aerial Vehicle. IEEE Transactions on Cognitive and Developmental Systems, 10(2): 290-303, 2018.

Shimon Edelman. Computing the Mind. Oxford University Press, 2008.

Aman Agarwal, Shimon Edelman, Functionally effective conscious AI without suffering, Journal of Artificial Intelligence and Consciousness 7:39-50, 2020.

KenjiDoya, TadahiroTaniguchi. Toward evolutionary and developmental intelligence. Current Opinion in Behavioral Sciences, 29, 91-96, 2019.

Stefan Elfwing, Eiji Uchibe, Kenji Doya, Henrik I Christensen. Darwinian embodied evolution of the learning ability for survival. Adaptive Behavior, 2011.