Abstract
A simulation design is introduced for using artificial intelligence (AI)-augmented developmental instruction for improving contemplative practices in the face of complexity. The focus of the design is on developmental instruction in the use of contemplative cognitive and moral reasoning practices to improve self-awareness and means to better discern and address complicated and complex challenges. Developmental instruction is supported by an AI-augmented tutoring aid provided during a simulated mission to Mars within an immersive 3D-world. A set of tasks are introduced during the Mars simulation to assist with the development of contemplative practices. The AI-augmented tutoring aid assists participants in the simulation to better understand and apply contemplative practices involving cognitive and moral reasoning meta-thinking suitable for each task. High levels of fidelity, involving visual, auditory, and interactive model-based reasoning (MBR) tools help to embed the senses, thoughts, and actions of participants in the feeling of traveling and being on Mars.
TopIntroduction
Since the second half of the twentieth century, physics has gradually lost the concision and simplicity of its classical theories. Modern theoretical models have become more and more complex, vague, and uncertain. Experimental verifications has become more difficult as well. (The Three-Body Problem, Cixin Liu, 2014)
In Cixin Liu’s science fiction novel, The Three-Body Problem, he introduces to readers several big questions related to complexity facing science and the developmental journey among key scientists in the novel about how their self-awareness grows in discerning and addressing complexity. Liu’s novel, winner of the Hugo Award for Best Novel, interweaves hard science with speculative fiction. Indeed, the title of the novel itself is a plot device for the unfolding story making use of the hard science that there is no solution to the general three-body problem in quantum mechanics. In general terms, the motion of three bodies in relation to each other is generally non-repeating, except in special cases. A complete solution for a particular three-body problem would provide the positions for three particles for all time (across time), given three initial positions and initial velocities. Unfortunately, no closed-form solution can be definitively offered for such a problem. This phenomenon is inherent to the chaotic nature of the time evolution of three-body or “n-body” (more than 2-body) systems. On a larger scale, for example, the three-body problem represented by the Moon, the Earth, and the Sun, takes a restrictive form whereby periodic calculated approximations are used to make predictions of trajectories of orbiting objects. In the case of space flight, a four-body problem (spacecraft, Earth, Sun, Moon), for example, can shift approximately to a three-body problem when gravity becomes negligible as the spacecraft travels further from a large mass and reverts back to a four-body problem when approaching an additional large mass such as another planet (e.g., Jupiter) and a moon (e.g., Europa) in addition to the Sun. In this example, the periodic calculated approximations require a computer to continuously analyze the calculated solutions.
The three-body problem helps to highlight the nature of some challenges encountered with evolving complex human-activity systems wherein no known solutions exist and the time evolution of system features is generally non-repeating and in flux (irreducible complexity). Such situations call for regular sense-making, approximations, and emergent strategies for solutions that cannot be tested in advance and will need to be adapted over time. But, not all challenges encountered by people with human-activity systems take the speculative form of a three-body problem. Rather, there are two-body problem forms (e.g., in classical mechanics; for example, a satellite orbiting a planet wherein the center of mass is always constant) involving the means to obtain a correct solution even though the system possesses complicated features (e.g., calculating the central force upon the satellite in orbit over time; uncertainty can be reduced given enough time and discovery of key features for a closed-form solution). Thus, some people can expect to encounter challenges representing forms of two-body problems (complicated) and three-body problems (complex). Generally speaking, post-secondary education can prepare most people with the means to initially address complicated challenges and offer the means for developing skills to improve over time. Unfortunately, however, most people are not well prepared to address challenges involving multidimensional aspects of irreducible complexity. This explains the need for developmental instruction for improving contemplative practices in the face of complexity. The following sections in this chapter explore in greater depth the nature of complexity, why there is a need to help people learn about differences between complicated and complex challenges, and describes a simulation design making use of developmental instruction to assist discernment and improving upon practices for addressing both challenges.
TopComplexity
Complexity is an evolving scientific theory addressing properties and behavioral phenomena of systems not well understood, explained, or predictable by means of conventional analysis of its’ parts or subsystems (Gleick, 1988). Complexity can be inherent to both structure and behavior of the system. And, interestingly, a complex system can have a complicated structure that is ultimately understandable but display behavior that cannot be regularly predicted across size and time scales (see Figure 1).
Key Terms in this Chapter
Augmented Cognition: Augmented cognition is an interdisciplinary field of scholarship addressing how cognitive processes can be supported directly with computer assistance or is influenced by human-computer interactions (to include robot assistance and interaction). Generally, with the application of AI, human cognition can be enhanced or augmented for improving metacognitive skills, reasoning, problem solving, and decision making. Applications offering augmented cognition have seen the most progress when applied using intelligent systems in domain-specific contexts and applications supporting interaction between machines and humans to improve performance of cognitively demanding tasks.
Immersive 3D-World Simulation: Immersive 3D worlds provide a way to experience a physics-based virtual reality (VR) modeled to support land, sea, air, and space environments as might be found in physical reality. Viewing applications are used to access an immersive 3D world and provide the means to establish a sense of presence over time and support interaction with 3D models and with other people, represented in the form avatars, in context of the environment. Interaction can take the form of text, voice, sharing content, and manipulation of models generating effects and responses. An immersive 3D-world can also allow users to create and place content in the environment. The environment size of an immersive 3D world can vary considerably from a few virtual acres to thousands of explorable acres. And, 3D worlds can be linked together across computer servers on a scale spanning the globe. Visits to 3D worlds can also be supported using hyper grid technology whereby a person’s avatar can travel easily from one 3D world to another. Increasingly, educators are making use of immersive 3D worlds to offer the means to embed simulations in context of VR real-world representations of problems and challenges to strengthen engagement, interaction, and meaningfulness of learning activities making use of model-based reasoning (MBR) and AI-augmented tutoring assistance.
Developmental Instruction: With developmental instruction, consideration is given toward offering instructional practices for helping a person transition to higher levels of cognitive and moral reasoning in alignment with expected developmental growth and demonstrated capabilities. Developmental growth to higher levels can be facilitated by instructional practices to introduce calculated incongruities encountered with the use of existing levels of cognitive and moral reasoning for facing challenges requiring higher levels of reasoning. Awareness of incongruities can help a person to perceive the need for growth and consequently be more likely to engage with developmental effort.
Complexity: Complexity is an evolving scientific theory addressing properties and behavioral phenomena of systems not well understood, explained, or predictable by means of conventional analysis of its’ parts or subsystems. Complex systems involve many components that are dynamically interacting across size and time scales. The components are self-organizing and form hierarchies and over time reveal emergent properties and or features that cannot be simply inferred from the behavior of the components making up the complex system.
Artificial Intelligence: There are multiple definitions for artificial intelligence. A common definition associates artificial intelligence to the means of a machine or software to imitate how humans reason, discover meaning, generalize, learn from past experiences, and behave. Some AI experts predict, however, the continued evolution of AI neural networks will introduce processing of information and adaptive neural networks that are very likely to become less like an imitation of human intelligence to the point of distinct forms of machine intelligence.
Contemplative Practices: The discernment of a situation or challenge for appropriate action makes use of contemplative practices involving holistic and deep patterns of thought (mindsets) associated with sense making and determining courses of action. Contemplative practices are supported by internal mental processes involved with metacognition associated with thought processes and awareness of one’s own cognitive strengths and weaknesses, knowledge about how to interpret the nature or type of challenge encountered, sense-making of new information, employment and switching of mindsets suitable to a challenge, with means to ultimately judge performance.
Model-Based Reasoning: The history of model-based reasoning (MBR) started with the creation of inferences in artificial intelligence (AI) to represent and model systems and their expected behavior in the physical world. Reasoning processes, making use of the inferences, involved comparison of the model system with observed data to gauge the degree of efficacy and accuracy of employed AI tools offered in intelligent decision- and performance-support systems. Over the years, MBR techniques employed in AI tools have been expanded for use in learning simulations to model a physical system, represent expertise usage of the system, with means to employ tutoring assistance for reasoning activities by learners to acquire or deepen understanding of the system and usage on the basis of how well the fit is between the learner and expert understanding levels and usage of the system. MBR, employed within immersive 3D simulations, can be designed to support usage of zone-of-proximal development (ZPD) to help guide the level and timing of tutoring offered to a learner on the basis of gaps revealed regarding system-model understanding and usage in context of a challenge.