• Corpus ID: 246035731

Combining Fast and Slow Thinking for Human-like and Efficient Navigation in Constrained Environments

  title={Combining Fast and Slow Thinking for Human-like and Efficient Navigation in Constrained Environments},
  author={M. B. Ganapini and Murray Campbell and F. Fabiano and L. Horesh and Jonathan Lenchner and Andrea Loreggia and Nicholas Mattei and Taher Rahgooy and Francesca Rossi and Biplav Srivastava and Brent Venable},
Current AI systems lack several important human capabilities, such as adaptability, generalizability, self-control, consistency, common sense, and causal reasoning. We believe that existing cognitive theories of human decision making, such as the thinking fast and slow theory, can provide insights on how to advance AI systems towards some of these capabilities. In this paper, we propose a general architecture that is based on fast/slow solvers and a metacognitive component. We then present… 

Figures from this paper



Thinking Fast and Slow in AI

It is hoped that the high-level description of the vision included in this paper can stimulate the AI research community to define, try and evaluate new methodologies, frameworks, and evaluation metrics, in the spirit of achieving a better understanding of both human and machine intelligence.

Interleaving Fast and Slow Decision Making

This work considers how to interleave the two styles of decision-making, i.e., how System 1 and System 2 should be used together, and proposes a novel and general framework which includes a new System 0 to oversee Systems 1 and 2.

Thinking Fast and Slow with Deep Learning and Tree Search

This paper presents Expert Iteration (ExIt), a novel reinforcement learning algorithm which decomposes the problem into separate planning and generalisation tasks, and shows that ExIt outperforms REINFORCE for training a neural network to play the board game Hex, and the final tree search agent, trained tabula rasa, defeats MoHex 1.0.

Meta-Reasoning: Monitoring and Control of Thinking and Reasoning

Metareasoning - Thinking about Thinking

This volume offers a simple model of reasoning about reason as a framework for its discussions, and considers metalevel control of computational activities, introspective monitoring, distributed metareasoning, and, putting all these aspects of metareasonsing together, models of the self.

Thinking fast and slow: Optimization decomposition across timescales

This paper seeks to provide a theoretical framework for how to design controllers that are decomposed across timescales in this way, and exhibits a design, named Multi-timescale Reflexive Predictive Control (MRPC), which maintains a pertimestep cost within a constant factor of the offline optimal in an adversarial setting.

Building Ethically Bounded AI

The notion of ethically-bounded AI is defined and motivated, two concrete examples are described, and some outstanding challenges are outlined.

The Consciousness Prior

A new prior is proposed for learning representations of high-level concepts of the kind the authors manipulate with language, inspired by cognitive neuroscience theories of consciousness, that makes it natural to map conscious states to natural language utterances or to express classical AI knowledge in a form similar to facts and rules.

Task complexity interacts with state-space uncertainty in the arbitration between model-based and model-free learning

It is shown that participants tended to increase model-based control in response to increasing task complexity, but resorted to model-free when both uncertainty and task complexity were high, suggesting that these two variables interact during the arbitration process.