Skip to main content

PEI WANG

Followers

59

Following

26

Co-author

1

Public Views

Uploads

Papers by PEI WANG

Toward a unified artificial intelligence

AAAI Fall Symposium on Achieving Human-Level …, 2004

To integrate existing AI techniques into a consistent system, an intelligent core is needed, whic... more To integrate existing AI techniques into a consistent system, an intelligent core is needed, which is general and flexible, and can use the other techniques as tools to solve concrete problems. Such a system, NARS, is introduced. It is a general-purpose reasoning system developed to be adaptive and capable of working with insufficient knowledge and resources. Compared to traditional reasoning system, NARS is different in all major components (language, semantics, inference rules, memory structure, and control mechanism).

A Model of Unified Perception and Cognition

Frontiers in Artificial Intelligence

This article discusses an approach to add perception functionality to a general-purpose intellige... more This article discusses an approach to add perception functionality to a general-purpose intelligent system, NARS. Differently from other AI approaches toward perception, our design is based on the following major opinions: (1) Perception primarily depends on the perceiver, and subjective experience is only partially and gradually transformed into objective (intersubjective) descriptions of the environment; (2) Perception is basically a process initiated by the perceiver itself to achieve its goals, and passive receiving of signals only plays a supplementary role; (3) Perception is fundamentally unified with cognition, and the difference between them is mostly quantitative, not qualitative. The directly relevant aspects of NARS are described to show the implications of these opinions in system design, and they are compared with the other approaches. Based on the research results of cognitive science, it is argued that the Narsian approach better fits the need of perception in Artific...

Goal Generation and Management in NARS

Artificial General Intelligence, 2022

AGI systems should be able to pursue their many goals autonomously while operating in realistic e... more AGI systems should be able to pursue their many goals autonomously while operating in realistic environments which are complex, dynamic, and often novel. This paper discusses the theory and mechanisms for goal generation and management in Non-Axiomatic Reasoning System (NARS). NARS works to accomplish its goals by performing executable actions while integrating feedback from its experience to build subjective, but useful, predictive and meaningful models. The system&amp;#39;s ever-changing knowledge allows it to adaptively derive new goals from its existing goals. Derived goals not only serve to accomplish their parent goals but also represent independent motivation. The system determines how and when to pursue its many goals based on priority, context, and knowledge acquired from its experience and reasoning capabilities.

Perception from an AGI Perspective

Artificial General Intelligence

This paper argues that according to the relevant discoveries of cognitive science, in AGI systems... more This paper argues that according to the relevant discoveries of cognitive science, in AGI systems perception should be subjective, active, and unified with other processes. This treatment of perception is fundamentally different from the mainstream approaches in computer vision and machine learning, where perception is taken to be objective, passive, and modular. The conceptual design of perception in the AGI system NARS is introduced, where the three features are realized altogether. Some preliminary testing cases are used to show the features of this novel approach.

What Should AGI Learn From AI & CogSci?

Abstract. While the fields of artificial intelligence (AI) and cognitive science (CogSci) both or... more Abstract. While the fields of artificial intelligence (AI) and cognitive science (CogSci) both originated from a deep interest in the same phe-nomenon – intelligence – and both setting themselves high aims in their early days, each has since greatly narrowed its focus, and all but aban-doned their core subject for a more limited version of the phenomenon. The many non-obvious causes for this change over the decades are per-haps understandable, but they have significantly reduced the potential of both fields to impact our understanding of the fundamentals of intelli-gence – in the wild and in the laboratory. This position paper argues that researchers in the field of artificial general intelligence (AGI) should care-fully posit their research objectives and methodology to avoid repeating the same mistakes. 1 The Big Picture of Intelligence and Cognition

Different Conceptions of Learning: Function Approximation vs. Self-Organization

Artificial General Intelligence, 2016

This paper compares two understandings of "learning" in the context of AGI research: algorithmic ... more This paper compares two understandings of "learning" in the context of AGI research: algorithmic learning that approximates an input/output function according to given instances, and inferential learning that organizes various aspects of the system according to experience. The former is how "learning" is often interpreted in the machine learning community, while the latter is exemplified by the AGI system NARS. This paper describes the learning mechanism of NARS, and contrasts it with canonical machine learning algorithms. It is concluded that inferential learning is arguably more fundamental for AGI systems.

An Architecture for Real-Time Reasoning and Learning

This paper compares the various conceptions of “real-time” in the context of AI, as different way... more This paper compares the various conceptions of “real-time” in the context of AI, as different ways of taking the processing time into consideration when problems are solved. An architecture of real-time reasoning and learning is introduced, which is one aspect of the AGI system NARS. The basic idea is to form problem-solving processes flexibly and dynamically at run time by using inference rules as building blocks and incrementally self-organizing the system’s beliefs and skills, under the restriction of time requirements of the tasks. NARS is designed under the Assumption of Insufficient Knowledge and Resources, which leads to an inherent ability to deal with varying situations in a timely manner.

OpenNARS Demonstration of Autonomous Learning and Decision-Making

NARS is a general-purpose reasoning system that is designed to be adaptive and capable of working... more NARS is a general-purpose reasoning system that is designed to be adaptive and capable of working with insufficient knowledge and resources. This, together with its ability to deal with uncertainty, qualifies it to be used as a core component of autonomous systems. Such systems often face novel situations they were not prepared for. Hence, a certain capability to adapt and improvise can be beneficial. NARS is capable of both, autonomous learning, and decision-making without requiring supervision. The purpose of this demo is to show these capabilities using our latest NARS implementation, OpenNARS v1.6.5. The examples we introduce involve unsupervised categorization and recognition of patterns (see [8]) in sensory channels, knowledge acquisition and procedural learning based on a stream of experience, and goal-oriented behaviours based on learned knowledge. Also, aspects of the implications of autonomous decisionmaking under the Assumption of Insufficient Knowledge and Resources (AIK...

Self-awareness and Self-control in NARS

This paper describes the self-awareness and self-control mechanisms of a general-purpose intellig... more This paper describes the self-awareness and self-control mechanisms of a general-purpose intelligent system, NARS. The system perceives its internal environment basically in the same way as how it perceives its external environment, though the sensors involved are completely different. NARS uses a “self” concept to organize its relevant beliefs, tasks, and operations. The concept has an innate core, though its content and structure are mostly acquired gradually from the system’s experience. The “self” concept and its ingredients play important roles in the control of the system.

On the validity of Dempster-Shafer Theory

2012 15th International Conference on Information Fusion, 2012

We challenge the validity of Dempster-Shafer Theory by using an emblematic example to show that D... more We challenge the validity of Dempster-Shafer Theory by using an emblematic example to show that DS rule produces counter-intuitive result. Further analysis reveals that the result comes from a understanding of evidence pooling which goes against the common expectation of this process. Although DS theory has attracted some interest of the scientific community working in information fusion and artificial intelligence, its validity to solve practical problems is problematic, because it is not applicable to evidences combination in general, but only to a certain type situations which still need to be clearly identified.

Toward a Logic of Everyday Reasoning

Springer Series in Cognitive and Neural Systems, 2019

Logic should return its focus to valid reasoning in real-world situations. Since classical logic ... more Logic should return its focus to valid reasoning in real-world situations. Since classical logic only covers valid reasoning in a highly idealized situation, there is a demand for a new logic for everyday reasoning that is based on more realistic assumptions, while still keeps the general, formal, and normative nature of logic. NAL (Non-Axiomatic Logic) is built for this purpose, which is based on the assumption that the reasoner has insufficient knowledge and resources with respect to the reasoning tasks to be carried out. In this situation, the notion of validity has to be re-established, and the grammar rules and inference rules of the logic need to be designed accordingly. Consequently, NAL has features very different from classical logic and other non-classical logics, and it provides a coherent solution to many problems in logic, artificial intelligence, and cognitive science.

Cumulative Learning

Artificial General Intelligence, 2019

An important feature of human learning is the ability to continuously accept new information and ... more An important feature of human learning is the ability to continuously accept new information and unify it with existing knowledge, a process that proceeds largely automatically and without catastrophic side-effects. A generally intelligent machine (AGI) should be able to learn a wide range of tasks in a variety of environments. Knowledge acquisition in partially-known and dynamic task-environments cannot happen all-at-once, and AGI-aspiring systems must thus be capable of cumulative learning: efficiently making use of existing knowledge while learning new things, increasing the scope of ability and knowledge incrementally-without catastrophic forgetting or damaging existing skills. Many aspects of such learning have been addressed in artificial intelligence (AI) research, but relatively few examples of cumulative learning have been demonstrated to date and no generally accepted explicit definition exists of this category of learning. Here we provide a general definition of cumulative learning and describe how it relates to other concepts frequently used in the AI literature.

Generalized Diagnostics with the Non-Axiomatic Reasoning System (NARS)

Artificial General Intelligence, 2019

Symbolic reasoning systems have leveraged propositional logic frameworks to build diagnostics too... more Symbolic reasoning systems have leveraged propositional logic frameworks to build diagnostics tools capable of describing complex artifacts, while also allowing for a controlled and efficacious search over failure modes. These diagnostic systems represent a complex and varied context in which to explore general intelligence. This paper explores the application of a different reasoning system to such frameworks, specifically, the Non-Axiomatic Reasoning System. It shows how statements can be built describing an artifact, and that NARS is capable of diagnosing abnormal states within examples of said artifact.

Agents in dynamic contexts, a system for learning plans

Proceedings of the 35th Annual ACM Symposium on Applied Computing, 2020

Scroll to the title and select a Blue link to open a paper. After viewing the paper, use the book... more

Special Issue “On Defining Artificial Intelligence”—Commentaries and Author’s Response

Journal of Artificial General Intelligence, 2020

Pei Wang's paper titled "On Defining Artificial Intelligence" was published in a special issue of... more Pei Wang's paper titled "On Defining Artificial Intelligence" was published in a special issue of the Journal of Artificial General Intelligence (JAGI) in December of last year (Wang, 2019). Wang has been at the forefront of AGI research for over two decades. His non-axiomatic approach to reasoning has stood as a singular example of what may lie beyond narrow AI, garnering interest from NASA and Cisco, among others. We consider his article one of the strongest attempts, since the beginning of the field, to address the long-standing lack of consensus for how to define the field and topic of artificial intelligence (AI). In the recent AGISI survey on defining intelligence (Monett and Lewis, 2018), Pei Wang's definition, The essence of intelligence is the principle of adapting to the environment while working with insufficient knowledge and resources. Accordingly, an intelligent system should rely on finite processing capacity, work in real time, open to unexpected tasks, and learn from experience. This working definition interprets "intelligence" as a form of "relative rationality" (Wang, 2008), 1. Most striking in these numbers is the glaring absence of female authors. A common reason among female academics for rejecting our invitation to contribute was overcommitment. As a community, we may want to think of new, different ways of engaging the full spectrum of AI practitioners if we value inclusion as an essential constituent of a healthy scientific growth. Self determination and willingness to participate are also essential. This is an open access article licensed under the Creative Commons BY-NC-ND License.

On Defining Artificial Intelligence

Journal of Artificial General Intelligence, 2019

This article systematically analyzes the problem of defining “artificial intelligence.” It starts... more This article systematically analyzes the problem of defining “artificial intelligence.” It starts by pointing out that a definition influences the path of the research, then establishes four criteria of a good working definition of a notion: being similar to its common usage, drawing a sharp boundary, leading to fruitful research, and as simple as possible. According to these criteria, the representative definitions in the field are analyzed. A new definition is proposed, according to it intelligence means “adaptation with insufficient knowledge and resources.” The implications of this definition are discussed, and it is compared with the other definitions. It is claimed that this definition sheds light on the solution of many existing problems and sets a sound foundation for the field.

Self in NARS, an AGI System

Frontiers in Robotics and AI, 2018

This article describes and discusses the self-related mechanisms of a general-purpose intelligent... more This article describes and discusses the self-related mechanisms of a general-purpose intelligent system, NARS. This system is designed to be adaptive and to work with insufficient knowledge and resources. The system's various cognitive functions are uniformly carried out by a central reasoning-learning process following a "non-axiomatic" logic. This logic captures the regularities of human empirical reasoning, where all beliefs are revisable according to evidence, and the meaning of concepts are grounded in the system's experience. NARS perceives its internal environment basically in the same way as how it perceives its external environment although the sensors involved are completely different. Consequently, its self-knowledge is mostly acquired and constructive, while being incomplete and subjective. Similarly, self-control in NARS is realized using mental operations, which supplement and adjust the automatic inference control routine. It is argued that a general-purpose intelligent system needs the notion of a "self," and the related knowledge and functions are developed gradually according to the system's experience. Such a mechanism has been implemented in NARS in a preliminary form.

Functionalist Emotion Model in NARS

Artificial General Intelligence, 2018

Emotions play a crucial role in different cognitive functions, such as action selection and decis... more Emotions play a crucial role in different cognitive functions, such as action selection and decision-making processes. This paper describes a new appraisal model for the emotion mechanism of NARS, an AGI system. Different from the previous appraisal model where emotions are triggered by the specific context, the new appraisal evaluates the relations between the system and its goals, based on a new set of criteria, including desirability, belief, and anticipation. Our work focuses on the functions of emotions and how emotional reactions could help NARS to improve its various cognitive capacities.

Conceptions of Artificial Intelligence and Singularity

Information, 2018

In the current discussions about "artificial intelligence" (AI) and "singularity", both labels ar... more In the current discussions about "artificial intelligence" (AI) and "singularity", both labels are used with several very different senses, and the confusion among these senses is the root of many disagreements. Similarly, although "artificial general intelligence" (AGI) has become a widely used term in the related discussions, many people are not really familiar with this research, including its aim and status. We analyze these notions, and introduce the results of our own AGI research. Our main conclusions are that: (1) it is possible to build a computer system that follows the same laws of thought and shows similar properties as the human mind, but, since such an AGI will have neither a human body nor human experience, it will not behave exactly like a human, nor will it be "smarter than a human" on all tasks; and (2) since the development of an AGI requires a reasonably good understanding of the general mechanism of intelligence, the system's behaviors will still be understandable and predictable in principle. Therefore, the success of AGI will not necessarily lead to a singularity beyond which the future becomes completely incomprehensible and uncontrollable.

The OpenNARS Implementation of the Non-Axiomatic Reasoning System

Lecture Notes in Computer Science, 2016