Toward a Comprehensive Pedagogy

Abstract

As we discussed in part 1, humanity is facing many critical challenges, most of which are open-ended and complex systems, and they require broad input and experiences to address. We expect students and educators to be engaged in such societal issues and form multidisciplinary teams to address them. The increased complexity of technical methodologies and know-how expected of future engineers, challenges existing curriculum strategies. The deep learning needed to dig into a discipline, to understand and apply it, must be combined with a learning strategy to increase the ability of students to relate to, and connect with, other disciplines in a meaningful way. It is not a matter of substituting students’ learning of core technical competencies; it is a matter of creating synergy in the learning process so students will experience the inevitable interaction between technical and contextual learning.

1 Introduction

As we discussed in part 1, humanity is facing many critical challenges, most of which are open-ended and complex systems, and they require broad input and experiences to address. We expect students and educators to be engaged in such societal issues and form multidisciplinary teams to address them. The increased complexity of technical methodologies and know-how expected of future engineers, challenges existing curriculum strategies. The deep learning needed to dig into a discipline, to understand and apply it, must be combined with a learning strategy to increase the ability of students to relate to, and connect with, other disciplines in a meaningful way. It is not a matter of substituting students’ learning of core technical competencies; it is a matter of creating synergy in the learning process so students will experience the inevitable interaction between technical and contextual learning.

As noted in the previous parts, the system is expanded beyond relations between technical and material constructs to include relations between people and the discourses, practices, and institutions they carry into the particular sociotechnical system. Together with the increase in complexity of both technology systems and value chains comes an increased need for perspective shifts to gain a comprehensive understanding of the system. Perspective shifts include a call for inclusiveness in the learning process, and this shift is based on a pragmatic approach to learning. Systems call for interdisciplinary views which basically are learned by formation of sets of actions to potentially change the system relations.

Because these challenges are complex with many interacting elements and of broad impact, learning achieved from pure disciplinary courses is not sufficient to teach students who address such challenges. There are several educational institutions that offer interdisciplinary courses to address human challenges. Whereas interdisciplinary education is necessary for addressing human challenges, this is not sufficient. In this section, we are not addressing the content of such courses, nor are we proposing new content, rather we are stressing the importance of pedagogy and ways to understand the learning processes beyond interdisciplinary learning. Unique pedagogy is needed to develop the appropriate skills and bridge the gaps between traditional academic fields and enable what is required to address large-scale challenges.

We start by a discussion of several general pedagogies in which skill development can be achieved within the context of systems thinking and design. Such skills combine computational, visual, experimental, and aesthetic methods. In addition, we advocate for those human-related skills, such as teamwork and collaboration, communication, and critical thinking, to be part of the skill building.

We stress that the cornerstone for all these pedagogies is the need to address open-ended challenges where answers are not fully known to even the educators. Many students are not familiar with working on problems that may have more than one acceptable answer and may require deeper studies, and also, students might not know how to conduct investigations or imagine that a problem may have more than one answer. For many years, students are immersed in attending lectures, reading textbooks, and solving problem sets, but the real world requires different training. Shifting out of a paradigm that is centered on teachers downloading information and requiring students to memorize this information is not an easy task for the students or for the educators.

Nevertheless, there are several educational methods that engage students and create an internal motivation that leads them to be well engaged with the content. These concepts belong to the general philosophical direction that appeared more than 100 years ago, when John Dewey advocated for educational progressiveness where authoritarianism is abolished (Dewey, 1897), and the emphasis is placed on delivering knowledge within students’ interests and experiences.

Edward Lee Thorndike debated with John Dewey over this philosophical direction and methodologies, in the early part of the twentieth century (Goodenough, 1950; Tomlinson, 1997). The debate was intense, and different educators and psychologists took different sides and created modifications of Dewey’s theories. In some sense, Thorndike won the debate, and it took many years for educators to realize the importance of giving the learners the ability to choose their own paths.

Being part of a social system and learning through social interactions is another advancement. In the past 25 years, major shifts took place. Digital technologies made a significant drive toward enabling learners to choose their paths. MOOCs helped in the shift from traditional learning methods but might have contributed to the creation of silo learners. Teaching online pushed the boundaries, with more progress in creating learning and assessing methods.

The historical perspective has integrative value. We start with an overview that illustrates the path toward the state we reached, and then we discuss some of the progress that is taking place in the early part of the twenty-first century in different universities. At the end of this chapter, we point out that our understanding of how people learn has improved a lot and the different theories bring different aspects, some of which can be integrated to generate wisdom upon which educators can use to deliver the best learning.

We end the discussions of this chapter with a short analysis of artificial intelligence and its impact. We introduce AI as a cognition augmenter with influence on learning content and methodologies. We keep the learner’s agency and future in the center of this discussion and attempt to analyze AI relationship to society, ethics, and commercialization.

2 Learning Theories

Learning is a human practice. It is related to development, knowledge, and creating skills. Through learning, we adapt to changing environments and create a better life for ourselves and for others. Learning is complex as it involves several factors that are all related to humanity. In thinking about learning, we must introduce factors from our biology and psychology, the culture, and the environment and, more recently, technology and its implications.

All such factors influence learning in complex and interactive pathways. Factors related to human behavior, cognitive, emotions, political systems, culture, and prior experiences must be considered and integrated. Obviously, learning cannot be understood through a single theory, and several learning theories have evolved and many overlap. Over the years, different orientations were developed and emphasized different aspects of the above-mentioned factors. These learning frameworks bear importance to pedagogy, and we emphasize that none of them is always true and there is significant overlap among these theories. Learning theories can be traced into certain categories: behaviorism, cognitivism, and constructivism.

3 Behaviorism and Connectionism

Thorndike’s pioneering work (Thorndike, 1898) on comparative psychology led to the emergence of educational psychology and that had impact on behavior analysis and reinforcement theory. Today, some of these concepts have bearing on AI and machine learning methodologies. Thorndike’s law of effect established the basic framework for several empirical laws in behavior psychology and had a long-lasting impact on pedagogy (Thorndike, 1905). Through experiments on animals, like cats, Thorndike concluded that ‘responses that produce a satisfying effect in a particular situation become more likely to occur again in that situation, and responses that produce a discomforting effect become less likely to occur again in that situation’. (Gray, 2011, pp. 108–109).

This emphasized behavior conditioning and reinforcement of learning by repeating facts and drilling became known as the law of exercise. Thus, learning is directly related to the amount of repetition, or practice, i.e., the drill. In addition, Thorndike stated that learning is the result of associations forming between stimuli and responses, which was known as the theory of connectionism and was elaborated on by others like Pavlov (1927), Skinner (1938), and Hull (1935).

Although Dewey advocated that there are missing elements in the Thorndike model, the method of learning by memorizing became essential to pedagogy for many years and may have been encouraged by Bloom’s taxonomy (Bloom & Krathwohl, 1969; Dewey, 1998). Bloom, however, put problem solving at a higher-order skill and his taxonomy puts remembering as a foundation for creating and recognizing elements and patterns. The taxonomy considers understanding, applying, analyzing, and evaluating as the steps between remembering, toward creative actions.

An important modification of the taxonomy created a set of verbs and products to illustrate how the different levels operate (Anderson et al., 2001). In addition, a consideration of the knowledge dimension was added. Different levels of knowledge, starting with factual, conceptual, procedural, and meta-cognitive knowledge were considered (Anderson et al., 2001). This results in 24 learning objectives, with the simplest being remembering facts, and the highest being innovative creation, which considers meta-cognitive creative actions (Heer, 2012). Bloom’s taxonomy has been applied in formal curricula descriptions at many universities and taxonomies can give an understanding of progressing learning.

In general, behaviorism focuses on behavior patterns, and many parts of our educational system are based on these ideas, where we describe the patterns in lectures as input and require the students to replicate the patterns, e.g., through problem solving, as output, which we assess. The learner’s inner world is not taken into consideration. A particular instruction, such as memorizing, is one of these patterns.

Appropriate feedback loops, such as praising and grading, reinforce these patterns. This area has developed over many years, and in some sense, we cannot ignore behavior as part of how we learn, and the importance of memorizing and repetition is often overlooked in the critical literature. Some parts were added, with educators modifying Bloom’s taxonomy, and we found an additional space for independent human thinking and believing in the learner’s intellectual capacity was further emphasized (Heer, 2012).

4 Cognitive Constructivism

Cognitive constructivism introduces the capacity of the human brain as a central element in the learning process and approximates the human as a computational machine with a processing capability, storage, and retrieval mechanism. Thus, one must follow the information flow into memory and consider previous experiences and how they interact with new knowledge. Further, knowledge comprises of mental representations with images, mental constructs, imaginations, and other human constructs, physical and mental, all of which interact in a complex way with the new knowledge.

The learner is not passive but is actively participating in the development of the knowledge by engaging in their interpretations and analysis. Furthermore, the learner’s stage of cognitive development influences what new knowledge is assimilated and how it fits with older knowledge. Piaget introduced the principle of equilibration that states that all cognitive development progresses toward increasingly complex and stable levels of organization (Piaget & Elkind, 1967). Equilibration takes place through a process of adaption and assimilation of new information into existing cognitive structures. The accommodation of the new information forms the new cognitive structures. This process continues as more information is introduced, and the learner employs their previous knowledge and their engagements to assimilate further knowledge.

These concepts lead us to realizing that acquiring knowledge is not a uniform process. It is a personal activity that depends on many factors, including previous experiences, cultural background, and maturity. Repetition is less important and drilling information is, most of the time, irrelevant, but being able to recall information is critical. With today’s digital search engines and machine learning, remembering lost its prime space, and more emphasis must be placed on the learner’s ability to invest personal time and energy to assimilate and possibly modify what messages they are receiving (Perry, 1999). This is not to say that there are no items that we need to remember, but that the human brain can be less burden by many facts when they are accessible through digital methods. Obviously, a student or investigator must be able to recall information, and this lower-level cognition is still needed for analysis and reaching reasonable conclusions. Even if information is passively received, connections to prior knowledge will require personal involvement. Learning how to learn requires training, effort, and ability to interact well with different digital media.

5 Social Constructivism

A crucial role for learning is making sense of the world and the meaning of its events. Social constructivism emphasizes the collaborative process among people and their connections to culture and society. Thus, learning is considered as socially constructed, rather than innate, or passively absorbed. In fact, long ago, Dewey stated that ‘Learning is a social activity—it is something we do together, in interaction with each other, rather than an abstract concept’ (Dewey, 1986) Since then, these concepts were refined with elaborations and details. Not only is it well recognized that learning is a social process, and it is influenced by cultural factors, but significant insights were obtained about the process of learning and related mental processes. In addition, the dynamics of the influence of culture were studied. Cultural evolution in time and space, and with technology, showed that learners are not copies of a certain construct and, in fact, learners participate in the construct of the culture they belong to.

This complex dynamic interaction makes acquiring knowledge a process built on symbolic mental constructs that are created by internal symbolic mental processes (Fox, 2001). The complex assimilation process for new knowledge that is accommodated and integrated within existing knowledge and mental structures was examined. Several theories were constructed to explain this process. Here, we present some of these theories, and later, we connect them with new pedagogy that has been emerging by integrating digital technologies during the past ten years or so.

5.1 Learning Through Sensing

For Bruner, the purpose of education (Bruner, 1973, 1977) is not to impart knowledge, but to create autonomous learners, who become creators of innovations. In Bruner’s theories, thinking is based on physical actions and the use of mental images as well as the five senses. Bruner believed that social interactions are essential in creating the learning.

Bruner proposed that learning involves encoding physical action-based information and storing it in memory. For example, infants learn by doing, rather than by internal representation or thinking. This mode of learning through physical activities develops to more complex ones later, for example, riding a bike or skating. Information is stored as sensory images, i.e., they are visual ones. This is an important notion which became part of our teaching as we commonly insert diagrams and illustrations in presentations and inherently believe that an image is worth a thousand words. Bruner believed that this encoded information continues to evolve as we grow older and later it gets stored primarily as words and symbols (Bruner, 1978). Some of the symbolic information becomes structured as in the case of mathematical equations, maps, and musical notes. These symbolic images end up enabling learning other topics (Bruner, 1966).

Symbols are very powerful, and we use them as shorthand in many domains, such as road signs and guidelines in hospitals and buildings. One may drive this even further by considering that numbers and alphabets are nothing but symbols. It is interesting to note that written words in Far-Eastern languages are modified images and they represent more complete meaning, whereas the invention of the alphabet by the Phoenicians attempted to create the sound used to articulate the words. For example, the written word of listening in Chinese has three elements: the ear to hear, the eye to watch the expressions of the speaker, and the mind or the heart to decipher the meaning, see Fig. 5.1. The learning symbol is written in two ways: hands and farming tools with off-spring child in a house, and the other symbolizes a bird practicing flying under the sun. These symbols are added to create meaning. Similarly, there are many signs that are universally recognized, like a stop sign or the exclamation mark for paying attention.

Fig. 5.1

Chinese symbol for to listen

In addition, learning through images is well established and is facilitated by technology such as YouTube and other visual apps.

5.2 Learning Through Social Interactions

Bruner constructed learning within social interactions and through a combination of cognitive and physical processes. Earlier, it was Vygotsky who placed much emphasis on culture and community in creating higher-order mental functions like learning (Vygotsky et al., 1978). His pioneering work was obscured by the lack of translations and the politics of the USSR. For Vygotsky, the environment is critical in creating the learning. Where learners grow up will influence how they think and what they think about. Social forces, such as values and beliefs, are part of human development as well as learning (Mcleod, 2008).

Therefore, communities play a very important role in the process of making meaning, and learning is a matter of sharing and negotiating socially constituted knowledge. With the advent of digital technologies, we observe changes in the formation of communities. There are now virtual communities, and these might be quasi-stable. They end up being formed quickly and dissolve fast too. Nevertheless, learning through such interactions is influential and must be considered, albeit not all their contributions are positive.

Social interactions affecting learning were also discussed by Piaget, who believed that regardless of culture, people pass through universal development stages. Young individuals construct their knowledge through early stages and culture is an indirect force in learning. Piaget also emphasized peer-to-peer learning from which the learner obtains deeper knowledge. Similarly, Vygotsky believed that early development is critical and will require support from adults.

From the above discussion, the emphasis on creating knowledge through a construct influenced by culture and social interactions does not give enough emphasis to the learner agency and self-direction. Of course, not all people are able to be self-guided. But it can be encouraged through the environment and the learning processes.

By the mid-1970s, Ernest von Glasersfeld introduced additional notions to what and how knowledge is constructed. Von Glasersfeld believed that all knowledge is constructed and not perceived by our senses (Cardellini, 2006). Active participation is required to acquire knowledge. With that, construction theory evolved to make knowledge a mental construct by individuals. Knowledge resides within our minds, and it may not match any reality (Driscoll, 2000). People are constantly developing their individual mental models of the real world from their perceptions of that world, and they update their models by new experiences and information. With Arends stating that meaning comes from experiences (Arends, 1998), constructivism made experiential learning essential.

Experiences integrate knowledge as well as disseminate knowledge, and individuals learn by interacting with each other, and this learning is both cultural and personal. Learning is filtered by the individual interpretations and their personal values. As the diversity of knowledge is shared through social interactions, knowledge and meaning emerge. These steps are critical for creating common knowledge, which is a step toward creating culture.

6 Educators as Enablers

Creating the appropriate environment for reflections and collaborative discussions, which lead to knowledge creation, is an immense task and noble responsibility. Thus, the most important responsibility for educators is not to instruct but to create environments where learners actively interact and co-create knowledge and meaning. Another aspect of the educator responsibility is to enable learners to fill in their knowledge gaps. As students may not come with similar backgrounds, scaffolding becomes a necessary enabler to create a participatory environment. Brooks and Brooks (1999) wrote on such environments, and Honebein (1996) summarized some pedagogical goals of constructivist learning environments, as:

• (a) To provide experiences for knowledge construction processes where it is left up to the students to determine how they will learn.

• (b) To provide experiences in which educators provide alternative solutions and students appreciate the presence of multiple perspectives.

• (c) To create authentic tasks connected to realistic contexts.

• (d) To provide agency for the students and make them the center of the learning process.

• (e) To ensure collaborative effects and social experiences.

• (f) To encourage using different modes of representation including imagery.

• (g) To encourage meta-cognitive processes such as reflections.

In Table 5.1, we present contrasting views between traditional and constructivist classrooms. Clearly, the constructivist method puts the students as active learners at the center of the stage, and places emphasis on their interactions. The move from the traditional classroom, based on behaviorist thinking, to the constructivist classroom, is fundamental for developing learning environments for complex problem solving.

Table 5.1 Classroom orientations

6.1 Students Are the Focal Point

Over the years, there were several moves to change the classroom from a lecturing place to a learning space, with different tools and mechanisms. In addition, the emphasis has been to change the classroom from being a formal platform for educators to be at the center of the stage to download their knowledge, to having the students as the focal point, ready to construct their own knowledge. This is a major shift in pedagogy.

Introducing such learning environments has been practiced over many years by many institutions. In particular, as early as the beginning of the twentieth century, Marie Montessori developed a theory and practice for educating children in which the child develops natural interests and activities within a supportive environment, which engages a variety of materials, including mathematics, natural science, culture, and the arts (Montessori et al., 2017). The learners develop a sense of self-responsibility and follow and develop their innate interests.

This educational style operates within a psychological self-construction that occurs through an environment plus interactions. The assumption is that all learners follow an innate path of psychological development and act freely when they are within an environment prepared according to the Montessori model. In addition, in this model, the role of the instructor is not to teach but to guide and counsel the learners by letting them create their own learning pathway and provide support when needed.

The Montessori schools are a special place, but the concept is broad and covers several other models. All of these models have a commonality, which is unification of three elements: First, the instructor is an advisor and supporter and does not download knowledge to a group of listeners; second, the learners have agency and interact with the environment, in the broadest sense, and with their peers, and third, the learning is taking place in an appropriate space, where interactions are facilitated by a lack of barriers and the presence of appropriate tools.

7 Learning by Doing

Concepts such as learning by doing, hands-on-projects, experiential learning, cooperative learning, and project-based learning belong to the superset of the ideal, which Dewey aspired to institute. These concepts differ on the execution side but have similar foundations and the objective of problem solving and critical thinking (Parker & Thomsen, 2019). The models favor group work and developing social skills, and they propose understanding and actions as the goals of learning as opposed to rote knowledge.

For example, cooperative learning (Johnson et al., 2014) and active learning create excitement when students develop their collaborative skills, build their self-confidence, and learn to take risks (Brame, 2016). Students may perform experiments that elucidate their learning and develop new ideas that test some theory. Active learning is not limited to hardware, and there are big opportunities in engaging students with computer simulations and modeling. We developed such modules for fluid mechanics and thermal transfer, as well as robotics. In addition, with Python as a platform, data analysis is accessible and may bring out important social findings.

Different methods put different emphasis on social responsibility. Project-based learning is another constructivist method in which the students learn by exposure to different problems using experiential learning and discovery (Guerra et al., 2017). In this method, emphasis is on big challenges and open-ended questions. Other social skills are also learned through these methods including critical thinking and communications.

In one variation in engineering, the context included four phases for creating comprehensive outcomes. These included conceiving, designing, implementing, and operating (CDIO) (Crawley et al., 2007). Within this context, there is flexibility to achieve each of these phases using different techniques. In general, the work that the learner is performing needs to be consistent with the cognitive capacity of the learner. Having a group of learners of different backgrounds enables creating comprehensive outcomes.

In general, in these methodologies, the instructors provide the theme of the projects. These projects could vary in scope and complexity, which will be discussed in the next chapter. Some might be related to a human challenge and do not have a single known answer. Although the instructor(s) choose the topic, they leave it up to the students to choose their problem statements and work on them. In choosing topics, attention should be paid to keep these topics broad enough to be useful as examples for future activities.

8 Connectivism in Online Learning

The importance of the social content in learning became a central pedagogy theme and practiced in many learning environments. In addition, this element made it in a model on its own, called connectivism. Different aspects of connection were practiced and known before this name was adopted. For example, the concept of peer-to-peer learning has been practiced at classrooms and outside them. But the advancements in digital technologies gave a new flavor for connectivism. In fact, technology created a very strong pedagogical shift in democratizing education, creating new environments, and allowing for fast experiments and measurable assessments. It enabled students to make choices in what to study and when. Initially, the concept of online learning, such as MOOCS (Baturay, 2015), allowed students to learn on their own pace and choose content from a large menu. This became popular enough that some universities taught online only, and some like MIT (Abelson, 2008) and Harvard (Brown & Adler, 2008) adopted the approach.

The concept of online learning was not only to provide economy of scale and freedom of choice of what to learn but also to create social connections among learners. This concept was presented by Oliver (2000) and then by Siemens (2005) as well as Downes (2005), who emphasized the connective model and advocated that the Internet is the medium for connective learning, or what is also termed as e-learning.

Learning online provides unprecedented learning opportunities. The ease of connecting, as well as the diversity of participants, which includes cultural experiences, provides tremendous opportunities for innovative outcomes. If we were to consider that every learner is a node and every node is a learner, the network becomes infinite, and no classroom can match the scale of the digital networks. And, since each individual learner has a distinctive point of view, cultural background, knowledge, and values, opportunities for creativity are enormous.

Learning as networks is a new realm, and interactions among learners can be viewed as interactions among diverse nodes in a network as in systems, described in Part I. These interactions give rise to complexity and unexpected outcomes such as emergence and even chaotic results. Self-organization has been observed in networked learning, and the diversity of the participants encourages interdisciplinary education and enables broad creative ideas.

Clearly, e-learning is a powerful platform, and it enables reflections and fast communication through blogging, email, and chat rooms. Although this platform is devoid of issues related to scale, different policies and the culture of different institutions can pose some obstacles for collaborations. As an example, a collaborative and flexible concept was envisioned in 2016–2017 and attempted to construct a new model for collaborative education by offering different courses online from different universities. In this model, learners decide which course to take from a menu of courses offered by more than one institution, and then graduate with a degree from one of them. The concept was discussed among UTEC, Harvard SEAS, Philadelphia University, and OCAD, but differences in cultures among these institutions hindered the sustainability of the collaborations.

In addition to having people in nodes, machines can occupy the nodes allowing for a new dynamic of human–machine interactions to enable a new type of learning and creativity. Co-invention with machines will become an essential part of our learning and doing. In addition, some of the content will be abridged and more mathematical simulations will be readily available, leading to shorter time for design and manufacturing.

Learning on the Internet matured but, perhaps, the most profound change took place during the COVID-19 pandemic. Not being able to communicate face-to-face made teaching instructions move to online, communication. This forcing function created the needed impetus for a significant change toward online education, which integrated several of the progressive pedagogies outlined before.

9 A Third Dimension of Learning Theories

Learning theories have developed, and in our education systems, we find elements of all theories, also in the new digital platforms. But for student-centered learning, it is important to understand student motivation, and therefore, another dimension is needed, which is not present in the above theories, which mostly concern the content and the interaction. Student motivation is a whole field in itself, related to psychology and the affective domain.

Illeris (2015) has developed a landscape of learning theories (see Fig. 5.2) and has defined three dimensions of learning, related to content, incentive, and interaction in and with society. In relation to the previous, Illeris adds the affective dimension of learning, in terms of the learner’s inner motivation and incentives of learning.

Fig. 5.2

Dimensions of learning (Illeris, 2015)

This discussion is an attempt to map the dimensions of learning theories and each theory will add a bridge to the understanding of how students learn. But it also reminds us that we cannot only form teaching and learning based on the content dimension grounded in cognitive or social cognitive learning theories, as the emotional incentives are important parts of learners’ motivations and learning. In the same line of argumentation, interaction and collaboration are needed to combine societal challenges with academic content.

Furthermore, the different dimensions of learning offer different aspects of diversity and inclusivity. The content perspective relates to the individual and cognitive dimensions of learning. Piaget introduced the concept of schemas, which organize comparable impulses from the outside world. The dynamics in these schemas constitute learning in its cognitive sense through the before mentioned assimilations (adding to schemas) and accommodations (changing schemas). Divergent thinking is about moving across schemas and exploring new impulses, which we have not yet explored.

When rethinking engineering education, the schemas that we have traditionally used in our understanding of electrical engineering, software engineering, and so forth must be rethought as well. The schemas we have explored have to be expanded, and in order not to overwhelm our cognitive capacity, we must find ways of selecting the relevant spheres to gain new impulses, and we must find other people who can complement our schemas in new ways.

Inclusivity in the cognitive domain relates to what we perceive as valuable knowledge constructs and, thereby, valuable sites to be explored and important schemas to (re)visit. In this sense, engineering educators have an important role in selecting core content but also in providing flexibility for students to select content themselves to complement their personal learning path.

The incentive perspective relates to the individual and emotional dimensions of learning. Diversity then relates to exploring different sources of motivation and inspiration for our engineering work—it is a diversity of feelings. This emphasis on affective learning outcomes has lately drawn attention to the role of emotions in engineering (Lönngren et al., 2020). In this respect, engineering students must be aware of, and consider, both intrinsic and extrinsic sources of motivation for students, peer, and self-motivation.

From an educational point of view and despite a rather limited focus on affective learning outcomes, different theories have shown attention to intrinsic sources of motivation for student learning. Positive psychology, e.g., person-centered teaching (Rogers, 1969), has had a tremendous influence on educational models. However, such models also create challenges for academic staff in motivating and empowering students to work with motivational aspects of their learning.

Without doubt, the problem determines the choice of applying relevant theories. This is a pragmatic view on learning and knowledge construction. If there are motivational issues, there is a need for looking into the corner of incentives, and even if there is research evidence that active learning increases student motivation, there might still be issues which can be understood and developed further by more psychological theories than cognitive or social cognitive theories.

10 Artificial Intelligence as a Cognitive Augmenter

Artificial intelligence has been in a steady development and improvement. AI has reached a level that its impact on learning is obvious. AI is related to e-learning, but it is far more impactful. Thus, it is critical to include this digital modality as part of our analysis.

10.1 Advancements and Concerns

AI is a sophisticated technology that provides augmentation to cognitive capacity. It should be considered as a tool that provides opportunities for learners and educators, and yet, it may present distortions and misleading information. AI consists of computer applications with very broad capabilities and objectives. It covers several areas of science, engineering, and automated techniques, as well as machine learning and deep learning. In general, AI systems algorithmic models perform cognitive and perceptual functions of the world. Such considerations were previously reserved for human thinking, judgment, and reasoning.

Now machines can augment human thinking and model observations using supervised and unsupervised learning and reinforcements. So, should we rejoice when we know that machines acquire data and perform structured and unstructured analysis and they process information and then create knowledge and recommendations to reach particular goals? Clearly, there are several science and engineering domains that can readily take advantage of the powerful AI. But will this technology be sensible enough and stay under the watch dog of humanity? Or will it stray fast and far and add an uncolorable new challenges?

AI augmentations may exist in physical and virtual dimensions and do not necessarily require acquisition of data. AI has functions that include rule-based analysis, e.g., expert systems, artificial neural network including vector machines, Bayesian networks, and decision trees. Also, symbolic AI has advanced to an impressive capability that flourished with the advancement of fast hardware. In addition, symbolic AI-enabled inverse design with very important implications to improved experimental research.

It should not be surprising that AI is not always guided by human objectives and sometimes it appears that it operates autonomously. This capacity for autonomous cognition raises a key issue of whether AI can independently operate within ethics, e.g., Boddington (2017); Whittaker et al. (2018); Winfield and Jirotka (2018). Ethical considerations are critical part of engineering and design. Being responsible and innovative, people can utilize AI for creating provocations and learning exercises. But ethics is a human trait, and it is a complex topic that is related to culture and time, and AI should not be bounded by only legal norms and laws. Jobin and colleagues (Jobin et al., 2019) identified 84 published sets of ethical principles for AI, which include transparency, justice and fairness, non-maleficence, responsibility, and privacy.

In addition, UNESCO (2021) describes draft recommendations on the ethics of artificial intelligence. AI is advancing continuously and has significant power and ability to integrate and synthesize and must be constrained by a code of ethics like the one used for medical practices.

More recently, AI language models based on an advanced version of the generative pretrained transformer (GPT) became available with a human chat interface and designed to generate human-like text. GPT-3 and 4 are pretrained over a series of language models which have been trained with a very large dataset of textual information and can be applied to deal with specific language-related tasks. The machine learned rules of grammar and syntax, the meaning of words and how they are used in many varying contexts. This learning enabled it to become a good writer of different text, including problem-driven reasoning and analogy-driven reasoning. As an example, design-by-analogy is the projection of existing reference in a source domain to address a comparable challenge in the target domain. With such capabilities, GPT-4 can perform interesting tasks, but they are not error free. Nevertheless, ChatGPT-4 will provide significant services including searches and supporting Microsoft Office suite, providing programming codes, and customer support. However, such AI will take some time to perfect these tasks, and it should be monitored for validity and accuracy. Meanwhile, engineering and design can use this technique and create effective learning exercises.

The design of the ChatGPT has an interesting element of involving the human in the decision making. When a chat did not satisfy the interacting human, the AI learns and shifts in attempt to satisfy the human. In doing so, the machine perfects the context of the request. This enable the AI to learn better and faster and deliver a closer contextual information.

More advanced GPTs are underway, and they are expected to appear with improved human-like capabilities. Here, we encounter a philosophical point on whether we can talk about learning in the context of machines as learning requires consciousness and agency. Neither one of them is attained by the current technology (Rehak, 2021).

10.2 Artificial Intelligence in Education (AIED)

AIED is an emerging field with several aspects. AI connections to education can be grouped under four headings: ‘Learning about AI, Learning with AI, Using AI to learn about learning, and Preparing for AI’ (Holmes et al., 2019). AI, as an intelligent tool, has unique contributions to education and should be considered with an eye to its strengths and weaknesses (Miao & Holmes, 2021). Each aspect of AIED needs to be considered independently.

• (A) Learning about AI

  Learning the AI tools and techniques is very important and there could be a process that is highly mechanical and automated. AI can then be effective contributor to teaching AI technologies. For example, it can provide instructions about machine learning and natural language processing together with the statistics and coding. This may also serve people who are interested in developing and contributing to the creation of AI applications.

  There is a need to understand AI algorithms and how these algorithms find patterns and connections in the data and make this literacy accessible to learners and citizens of different backgrounds (Miao & Holmes, 2021). But this literacy should not be limited to the technical components. The human dimension must be an equal part of the learning. The impact of AI on human cognition and agency must be discussed at length (Holmes et al., 2019). For example, people must understand power and political motivations that are behind the adoption of automated decision making.

  It should be clear that learning with AI is not complete and might be even defective, if it was not supervised by humans who introduce human thinking in this process, including ethics and understanding the limitations of AI.

• (B) Learning with AI

  As AI intelligence improves, one might be tempted to use AI as a substitute teacher. However, such an effort should be studied very carefully as the lack of human interaction and sociability may have grave effects on the students’ mental health. Instructions from machines might be very different from the teaching and mentoring with a human touch. Young people are very impressionable, and it is possible that the absence of the human and the lack of the emotional content will adversely affect their learning. Young people need to have role models who provide motivations and excitement. Machines are void of the human elements and students may feel isolated causing them to drop from their educational paths.

• (C) Using AI to Learn About Learning

  Learning analytics and educational data mining are tools to gather data on how learners learn, their learning progression, and which learning designs are effective. The goals of such studies are to inform learners, teachers, and other stakeholders about students’ practices. But there are other negative side effects to AI gathering such information. Such information may affect admission policies, retention of students, and education planning. This overlapping but nonetheless distinct issues cannot be left to algorithmic thinking.

• (D) Preparing for the impact of AI

  This involves ensuring that citizens understand the risks of AI and are prepared for its possible impacts. Therefore, preparing for AI should be integrated within learning about AI. The impact of AI on human dimensions needs to be emphasized despite the tension with economic gains. The purpose of education becomes a crucial point to consider. Education cannot imply only to transmit knowledge that is selected without regard to the learners’ input and delivered in a process that does not focus on the learners. Transferred knowledge must help people to develop their individual potential, self-actualize, and promote understanding, tolerance, and sociability among all peoples.

As we advocated in this chapter and other chapters of this book, a shift to learner-centric learning where students have control over the learning processes, thereby maximizing their agency, is critical. Some thoughts about the role of AI in education embody a naive approach to teaching and learning. Most of the proposals consist of simple approach of providing direct instructions and informing of prespecified content, while adapting to the individual capacity, and then assessing the individual’s performance using AI-driven e-grading and proctoring. Such an approach has many issues. First, this style of advising negates the importance of agency and innovation. Second, adapting to the capacity of the learner by using AI evaluation methods has many pit faults. Some commercial companies that deal with education sought to use AI-driven tools to create personalized recommendations for students to pick from and create their educational content. Thus, the AI algorithms become responsible for the individual educational pathway, which might be personalized but might not be focused on the learner’s destination of interest. Third, e-proctoring was shown to fail in considering the human dimensions. This method was accused of ‘intrusion (Barrett et al., 2019; Hager et al., 2019), racial discrimination, preventing learners taking their exams and exacerbating mental health problems, while having little impact on cheating or attainment’ (Brown, 2020 and Conijn et al., 2022).

In short, there is tremendous need to carefully consider the impact of AI algorithms on the development of human cognition (Ilkka, 2018) and the education of a new generation. There is a tremendous commercial interest in AI and an incredible amount of investment in it, which approaches US$80 billion, of which US $2 billion were targeting education, one of which is championed by Google (Google, 2023). Among these excitements, voices must advocate the need for a significant pedagogical depth; and that unmanaged AI can pose significant risks to education (Holmes et al., 2019).

These concerns, however, should not discourage people from utilizing this technology, and while AI can provide insights and recommendations, it should not be allowed to make final decisions on its own. Humans must carefully evaluate the information provided by AI and consider all relevant factors before making conclusive decisions.