Situated Artificial Intelligence: Fundamentals and Applications for Integrating Intelligence With Action

By Fouad Sabry

What Is Situated Artificial Intelligence

The term "situated" is used in the fields of artificial intelligence and cognitive science to refer to an agent that is immersed in an environment. It is usual practice to use the term "situated" to refer to robots; however, some researchers contend that software agents can also be situated if the following conditions are met: they must exist in a dynamic environment, which they can control or alter through their activities, and which they must be able to feel or perceive.

How You Will Benefit

(I) Insights, and validations about the following topics:

Chapter 1: Situated

Chapter 2: Situated cognition

Chapter 3: Situated robotics

Chapter 4: Situated approach (artificial intelligence)

Chapter 5: Intelligent agent

Chapter 6: Embodied cognition

Chapter 7: Virtual intelligence

Chapter 8: Smart object

Chapter 9: Computer-supported collaborative learning

Chapter 10: Technoself studies

(II) Answering the public top questions about situated artificial intelligence.

(III) Real world examples for the usage of situated artificial intelligence in many fields.

Who This Book Is For

Professionals, undergraduate and graduate students, enthusiasts, hobbyists, and those who want to go beyond basic knowledge or information for any kind of situated artificial intelligence.

What Is Artificial Intelligence Series

The artificial intelligence book series provides comprehensive coverage in over 200 topics. Each ebook covers a specific Artificial Intelligence topic in depth, written by experts in the field. The series aims to give readers a thorough understanding of the concepts, techniques, history and applications of artificial intelligence. Topics covered include machine learning, deep learning, neural networks, computer vision, natural language processing, robotics, ethics and more. The ebooks are written for professionals, students, and anyone interested in learning about the latest developments in this rapidly advancing field.
The artificial intelligence book series provides an in-depth yet accessible exploration, from the fundamental concepts to the state-of-the-art research. With over 200 volumes, readers gain a thorough grounding in all aspects of Artificial Intelligence. The ebooks are designed to build knowledge systematically, with later volumes building on the foundations laid by earlier ones. This comprehensive series is an indispensable resource for anyone seeking to develop expertise in artificial intelligence.

• Language: English
• Publisher: One Billion Knowledgeable
• Release date: Jun 22, 2023

Book preview

Chapter 1: Situated

The term situated in artificial intelligence and cognitive science describes an agent that is integrated into its surroundings. Robots are typically referred regarded as being placed, however some academics contend that software agents can also be situated if:

They live in an environment that is dynamic (quickly changing), which

via their behaviors, they can influence or transform, and which

they are able to detect or discern.

Examples might include web-based agents that can update data or start online actions (like purchases), or virtual reality bots that live in and alter virtual environments like Second Life.

Generally speaking, being located is regarded as a component of being embodied, although it is important to analyze each viewpoint separately. According to the situated approach, intelligent behavior is influenced by the environment and how an agent interacts with it. The embodiment of an agent determines the type of these interactions.

{End Chapter 1}

Chapter 2: Situated cognition

According to the notion of situated cognition, knowing and action go hand in hand.

Situativity theorists advocate a model of learning and knowledge that emphasizes on-the-spot reasoning over the storing and retrieval of conceptual information. In essence, cognition and surroundings are inseparable. Instead, knowledge is already present and indivisible from activity, context, people, culture, and language. Since knowledge is determined by both the agent and the context, learning is therefore understood as a person's ability to function in a variety of contexts with increasing effectiveness rather than as a collection of knowledge.

While situated cognition emerged in educational psychology in the latter half of the 20th century, theories from anthropology, philosophy, critical discourse analysis, and sociolinguistics—Jean Lave and Etienne Wenger (1991), Martin Heidegger (1968), and Bakhtin (1981)—rejected the idea of truly objective knowledge and the tenets of Kantian empiricism.

The situational action research done by Lucy Suchman at Xerox Labs

The concept of affordances was first developed by James J. Gibson as a component of a relational theory of perception. Gibson concentrated on the affordances of actual items like doorknobs and chairs and proposed that these affordances were directly sensed by a person as opposed to being mediated by mental representations like mental models. It is significant to note that advocates of a more category-based model of perception have long disputed Gibson's claim that direct perception is an unmediated process of observing, perceiving, and recording particular properties from the environment.

This ecological psychology emphasis on agent-situation interactions was congruent with the situated cognition program of scholars like James G.

The movie Greeno, 1998), who found Gibson's apparent rejection of experimental psychology's factorization assumptions to be admirable.

The situated cognition perspective focused on perception-action instead of memory and retrieval…A perceiving/acting agent is coupled with a developing/adapting environment and what matters is how the two interact.

Additionally, according to Greeno (1994), affordances are preconditions for activity even though they do not directly influence behavior, they raise the probability that a specific activity or behavior will take place.

Later, Shaw, Turvey, and Mace (as described by Greeno, 1994) coined the term effectivities, which refers to an agent's skills that govern what the agent can accomplish and, as a result, the interactions that can occur. The effectivities and affordances, which acted in the moment, together, co-determined perception and behavior. Therefore, based on his effectivities, the agent directly viewed and interacted with the world, deciding what affordances could be taken up. This viewpoint is in line with Norman's (1988) perceived affordances hypothesis, which places an emphasis on the agent's perception of an object's utility rather than the thing itself.

The link between affordances and mental representations as presented from a more cognitivist perspective is an intriguing subject. While Glenberg & Robertson (1999) proposed that affordances constitute the foundation of mental models, Greeno (1998) contends that attunements to affordances are preferable than constructs like schemata and mental models.

Situated cognition is heavily influenced by the work of Gibson (1986) in the area of visual perception. According to Gibson, the brain does not translate inputs from the eye into symbolic representation when it comes to visual perception. Instead, the observer notices and absorbs the vast amount of data present in the environment. In particular, an agent detects affordances through identifying variants—what changes—and invariants—what does not change—across many contexts. The agent and the affordances of the environment jointly determine an agent's perceptions of invariants given a particular intention (or purposeful set), which are then built upon over time.

Perspectives on situated cognition and ecological psychology place an emphasis on perception and argue that memory has a much smaller function in learning. Instead, emphasis is placed on constantly fine-tuning perceptions and behaviors across contexts based on the environment's affordances and the agent's interaction with it (Greeno, 1994). Representations are formed and interpreted in activity rather than being stored and compared to prior information (Clancey, 1990).

Situated cognition views memory as an interaction with the outside environment, constrained by significant circumstances, that moves an agent toward a predetermined objective (intention). As a result, effectivities and affordances, which function in the moment, jointly govern perception and action. In order to determine what affordances can be picked up, the agent must directly experience and interact with the world. He does not merely recollect stored symbolic representations.

Situativity theorists reimagine knowledge as an action or verb rather than as an object, thing, or term. Agents demonstrate their knowledge through action by engaging with and becoming a part of various communities.

Since knowledge is anchored in action and cannot be separated from a person's personal, societal, and historical objectives

People engage with texts, not only read or produce them, and frequently, these interactions involve other people in different socio-cultural situations. The function of situated cognition in language learning activities is crucial since language is frequently the starting point for measuring and tracking learning gains in comprehension, subject-matter knowledge, and tool use both within and outside of the classroom. The use of tools, technology, and speech for full involvement typically determines membership and engagement in social and cultural organizations. It is important to consider how different social and cultural groups use and interact with these texts while teaching or studying languages or literacy.

Legitimate peripheral participation (LPP), according to Jean Lave and Wenger (1991), offers a framework to explain how individuals (or newcomers) integrate into a community of learners. Lave and Wenger's conception of situated cognition (also known as situated activity) placed a strong emphasis on legitimate peripheral participation since it brought socio-cultural and historical realizations