Composing Systems for Advancing Advancing

Modes of Conversation

Glanville was known to shift conferences toward conversation, discussion, and learning, as opposed to just listening to paper presentations. His concept of conversation might take a number of modes, embodying second-order cybernetic circular causal forms and including physical conversations with conceptual material as well as material processes, such as sharing and discussing an experiment; thinking through a process, like exploring ideas iteratively via pencil and paper; conversation with the self (possibly in combination with a physical iterative processes); conversations with others both organized and unorganized; and conversations with the constructed artifacts of others, such as Glanville’s constructed conversation with Pask addressing aspects of Pask’s (1961) book An Approach to Cybernetics (Glanville & Müller, 2007, pp. 13–27). In a sense, my conversation with Glanville in “Composing Composing” is a related approach to conversations after the fact. Many of the quotes I have chosen for “Composing Composing” came from Glanville’s writings and in particular from Cybernetics and Human Knowing articles. One can also have a conversation with interactive systems: relational databases, elements of AI, search engines in general, interactive artworks such as Recombinant Poetics (Seaman, 1999b, 2010; Seaman & Rössler, 2011), blogs, and listservs, etc. All of these forms of conversation can contribute to advancing advancing. I will loop back later in the article to my Insight Engine 2.0 project (Seaman, 2022a, 2022b), a self-organizing relational database and visualization system that explores a multi-perspective approach to potential insight and knowledge production (Seaman, 2014), incorporating new forms of AI functioning as collaborators along with a set of potential international individuals and human collaborators. This system of systems is still a work in progress.

Conversations—What is the Question? What is a Question? and What is a Context?

In the famous text Man-Computer Symbiosis, J. C. R. Licklider (1960) states:

Here human thought and intuition can potentially find new relationalities between disparate areas of research via new computational potentials working toward the creation of higher-order systems. It has been many years since the Licklider text. Huge technological potentials have opened up, especially in terms of intelligent natural language systems, analysis and focused creation across multiple disciplines, and the development of scientific experiments brought about through AI exploration. Researchers and system designers can also project future potentials. Perhaps we can create a computational system that deeply understands embodied notions of context and might also be introspective about that context. There is perhaps much to work on in relation to this conversational question, where the computational system is in conversation with itself.

In Philosophical Investigations, Wittgenstein (1958, para. 24) asked a very special pithy question—What is a question? Notice above that Licklider asks what is the question? Also, in this book, Wittgenstein (1958, para. 43) discussed “the meaning of a word is its use in the language.” Here, I ask a different but related question: What is a context? There is something very special about the human ability to quickly and effectively discern context as well as observe contexts from a meta-level of introspection. Here contexts define the use and meaning in terms of differing forms of language, as well as through human interaction with the environment and each other—through both embodied and textual means.

Sometimes this difference in terms of learning through embodied interaction with environments is not fully appreciated given the success of textually derived AI systems comprised of large language models. Yet there are now many people interested in exploring the enacted, embodied, embedded, extended paradigm as a means for understanding aspects of cognition (Ward & Stapleton, 2012). Glanville was interested in how conversations and language communities help to articulate notions of shared meaning, albeit unique for each individual. People should become more aware of the fact, if they haven’t already, that humans are embodied beings that in part use senses to build up accretive contextual knowledge over time. Knowledge is never fixed but enfolds new knowledge in an ongoing manner. Texts are of course a large part of learning, but what does this say about AIs that learn from text only? How does a computer’s sense of context differ from that of the human, which in part understands via an embodied, multi-sensed interaction with the world? Alternately, what happens to context when a human programs a computer? Do they still draw on embodied historical learning? I would say yes. What happens when a computer programs a computer? Do computational systems have a very different sense of context or no meta-level (introspective cognition of context) understanding at all? To what degree does human knowledge production draw on embodied interactions that enfold multi-modal sense understandings of the world with textual understandings (Pfeifer & Bongard, 2007)? One of my goals for neosentience production is to develop new intelligent systems that draw on an embodied sense of understanding context through sensing systems that are mixed with language-based AI—large language models. There are some new examples of AI that include elements of multimodal sensing (Mollick, 2023).

Glanville was interested in how mindsets help to determine ongoing understandings of context production. He wanted to point to how this difference might be better articulated in terms of new conversational contexts. Glanville (2012) states:

I will return to this concept in terms of the definition of intelligence, later in this document.

Variety

Along with conversation, Glanville was very interested in increasing variety. He was enamored of the writings of Ross Ashby on this subject and explored the increasing of variety as one approach to enhancing creativity in a general sense. In “A (Cybernetic) Musing: Variety and Creativity,” Glanville (2009) states:

In particular, Glanville was interested in expanding ideas of design in relation to cybernetics and expanding cybernetics in relation to ideas of design. Interestingly, part of Glanville’s contrarian modus operandi explored the temporary loss of control as a way of increasing variety. This was unlike the historical positioning of cybernetics in terms of “steersmanship” as articulated in Nobert Wiener’s (1948/1961) seminal book Cybernetics: Or, Control and Communication in the Animal and the Machine. Glanville (2009) states:

This concept, exploring unmanageability, can become actuated in different contexts. The “unmanageability” that Glanville points at is normally encountered in certain stages of learning and growth processes, especially learning that takes people outside of their comfort zone of knowledge production in relation to crossing the boundaries of differing intellectual domains. Negotiating unmanageability is not something to be shunned. To my mind, it is perhaps the price one pays for making some of the greatest advances in knowledge production. For many people this means doing double duty—studying both their discipline but also obtaining a strong working knowledge of new disciplines. Central to this concept is becoming aware of computational tools that potentially help articulate ecologies of relationality between these differing disciplines to promote intelligent conversations across disciplinary domains. This might be achieved by exploring new forms of AI (Hern, 2023). We hope to also explore and test open-source large language models Llama 3.1 and Mistral-NeMo in terms of building new forms or relationality between disciplines—new forms of conversation. One goal is to explore open source focused large language model systems in our search engine/visualization system Insight Engine 2.0.

To advance advancing one may need to explore new forms of computation to help elucidate different forms of relationality, especially relationality that embodies concepts related to biomimetics and bio-abstraction, to human/machine relationality. Glanville believed that unmanageability could increase variety, and thus promote instances of potential growth, particularly in relation to aspects of iterative design processes. In terms of the steersmanship and navigating the circular causal wheel of second order cybernetics related to conversations with the self and others, Insight Engine 2.0 seeks to include conversations with an intelligent computational system of systems by diverse human interactants with different levels of science and branches of intellectual exchange. Glanville pointed to a form of learning that explored allowing oneself to enter into unmanageable conversations that might initially start outside of one’s current abilities. Such strategies loop back to the beginnings of cybernetics in terms of exploring conversations between humans representing differing disciplines, yet perhaps shines a light on potential new forms of conversation, like conversations with an AI. Norbert Wiener talks about this kind of cross-disciplinary talk in Cybernetics. Speaking about defining early relations between computation and thinking, Wiener (1948/1961) stated “Everywhere we met with a sympathetic hearing, and the vocabulary of the engineers soon became contaminated with the terms of the neurophysiologist and the psychologist” (p. 15).

How can this disciplinary boundary crossing best be facilitated? In each case participants may potentially need to learn to apply aspects of knowledge production across differing domains, domains that are often historically isolated. This kind of disciplinary boundary crossing can in part be achieved by defining relevant ecologies of relations and functional biomimetic abstractions. Historically, this often meant taking initiative related to learning about a different discipline, understanding its vocabulary and gaining a grasp of certain of the central concepts relevant to that other discipline to help define and build this relevant ecology of relationalities. It may also be highly significant to include a team member or members that have deep knowledge of both or multiple domains, working on complex research problems.

One of the most important early texts discussing what has come to be called boundary objects is Sorting Things Out: Classification and Its Consequences (Bowker & Star, 2000). Here, a process is developed for exploring differing disciplinary perspectives in an articulate way. In the text they discuss the concept of boundary objects, defining them as representational forms that can be shared between different communities. In a related text, Star and Griesemer (1989) state: “Boundary objects are objects which are both plastic enough to adapt to local needs and the constraints of the several parties employing them, yet robust enough to maintain a common identity across sites” (p. 393)

Glanville (2009) states the following in “A (Cybernetic) Musing: Design and Cybernetics”:

Thus, designing systems that help model, design, and build new pluralistic understandings of contexts by including conversations exploring many different disciplinary perspectives is central in helping us articulate relevant networks of biofunctionality and its abstraction through biomimetics.

Unmanageability leading to growth in knowledge enables humans to fall outside of their comfort zones to increase variety, and often to take in knowledge derived from other personae, which might potentially also fall outside of one’s current mind-set. These kinds of conversations can help define systems that point in the right direction related to solving certain sets of complex problems. Undertaking this aspect of unmanageability is a circular accretive process. In such contexts the unknown becomes the known via process and iterative change brought about through conversations with the self and others, with papers and books, with systems and new computational methodologies to advance study and research. As discussed above, the other with whom one converses might increasingly include AI collaborators, working in concert with huge relational databases and eventually through differing means of visualization, sonification, and hapticization. Such interdisciplinary and transdisciplinary undertakings may also mean creating special systems that help define context-relevant jargon, so that computational linguistic processes enabling the articulate bridging of research domains can be as accurate as possible. Glanville (2012) states:

One of my goals is to define and design new conceptual approaches to draw on, enabling better access to certain articulate ecologies of relational concepts that embody aspects of temporary unmanageability. Another approach is to form a team of individuals to tackle together, through conversation, aspects of complexity. In discussing Glanville’s change from observer to composer above, Glanville provided a very broad definition of design. In one instance, Glanville discussed how design related to science in terms of composing. How does one come to design or better compose the ideas that form an experiment or an approach to a problem set? This points to the difference between the scientific method which seeks to keep the observer at bay, and the second-order cybernetic set of process that go on in the scientist via the conversations they have with themselves about how to design experiments. This puts design in an extremely important light, in terms of knowledge building processes. One designs science, and in particular scientific experiments and new approaches to knowledge production, as well as new approaches to highly complex linguistic and human-sense related embodied computational systems.

With AI, such approaches can touch on the nature of the uncanny, as machines exhibit what seems like human intelligence and/or addresses the abstraction of human intelligence. Of course, there is an ethics and responsibility attached to the design and use of such new systems. Many people fear the nature of unmanageability when it comes to conversing with machine intelligence.

Unpacking Human Sentience

One very long-term question that interests me is to what degree is it possible to unpack human sentience from multiple perspectives? I am here collapsing sentience and sapience as is often done by computer researchers. How can this process be valuable to knowledge production in general, especially in terms of defining new computational tools and approaches to knowledge production? Researchers can potentially create self-organizing computational systems to help in unpacking and elucidating this extremely complex network of enfolded processes which contribute to sentience/sapience arising. If scientists, researchers, and lay people never reach a full understanding via such systems, I feel that it is still important to chip away at this highly complex problem set over the span of a human lifetime, and beyond.

I have sought to enfold a series of differing foci though the development of generative systems, empowering disciplinary and transdisciplinary cross-pollination. In particular, neosentience (Seaman & Rössler, 2011) is a potential new form of AI that is inspired by mind/brain/body/environment, enfolding shared language use and self-observation, as well as employing meta-level systems.

I am thus seeking to explore composing systems for advancing transdisciplinary insight through conversation with the self and the knowledge of others, enfolding the focused employment of intelligent machine-based systems in this distributed network of conversations. The ecology of goals includes composing systems for advancing transdisciplinary insights (outlined in the next chapter: The Insight Engine 2.0: Advancing Advancing—Toward a Neosentient Model), and, in particular, authoring systems that point toward composing higher order intelligent systems.

Rössler and I defined a unique approach to articulating neosentience. In our book, Neosentience|The Benevolence Engine, we discuss a set of neosentient pragmatic benchmarks (“it” in the list below being a potentially neosentient entity):

• It can learn.

• It can intelligently navigate.

• It can interact via natural language.

• It can generate simulations of behavior (it can “think” about potential behaviors) before acting in physical space.

• It can function in a creative manner.

• It can come to have a deep situated knowledge of context through multimodal sensing.

• It can display mirror competence.

• It can function in a benevolent manner.

Synthetic emotions would become operative within the system. The neosentient entity is potentially brought up (brought to life) in a social and cultural sphere of reciprocal inter and intra-actions contributing to language and knowledge acquisition. This is achieved through embodied relations to the environment, self, and others.

Composing Systems for Advancing Advancing

Given the interest in holistic thinking as being central to second-order cybernetics and systems thinking, a question arises when seeking to articulate models that employ aspects of highly complex biological processes. Ludwig von Bertalanffy (1968), in his book General Systems Theory, discusses wholeness:

The human body is an extremely complex system of systems. In “A (Cybernetic) Musing: In the Animal and the Machine,” Glanville made some interesting observations about the relationality between the animal and the machine—pointing at the beginnings of cybernetics and then articulating how those ideas changed over time. Glanville (2009) states:

Neosentience research is driven in part by the latter approach, thinking of the interactive system as being organism-like and adaptive. Roy Ascott, an artist, discussed this in his text “Behaviourist Art and the Cybernetic Vision” (Ascott, 1966). I consider neosentience research, explored via the Insight Engine 2.0, to be the enfolding of conceptual art, design, science, the humanities, and cybernetics.

Advancing Advancing as Related to Neosentience Research

I am deeply interested in biomimetics and bio-abstraction as they are applied to new forms of robotics and computation. The goal is to best design systems enabling the examination of multiple of the above topics to help diverse interactants explore this ever-expanding ecology of relationalities. I am using a game engine for the front end of this interaction, as well as a website. Both are still under development. It came to me that if I substitute the interfacing of complex processes for “parts” when defining new models of highly complex environments, and seek to define how each process interfaces with the others in order to enable emergence, that perhaps I could potentially transcend aspects of the problem of using machine-oriented parts as a means to address biomimetics based on human biological systems. I thought that this might be productive, drawing on both metaphors and analogies related to the animal and the machine. Previously, I have referred to the body as “a not yet fully entailed ultra-complex bio-machine” (Seaman, 2013, p. 63). I will provide a short definition here: The human could be considered to be an ultracomplex bio-machine made up of an ecology of relations that relate to processes and interfaces that interact with other processes and interfaces within an autopoietic system.

One likes to circle back to Gordon Pask now and again, whose definition of cybernetics seems quite quotable here: “Cybernetics is the science or the art of manipulating defensible metaphors; showing how they may be constructed and what can be inferred as a result of their existence” (Pask, 1975a, p. 13). In particular, “the idea that the machine can be treated through the metaphor of the animal” (Glanville, 2009, p. 108) becomes highly relevant in neosentience research. Yet, I am still enfolding metaphors related to a very new kind of machine—the bio-machine.

Concerning cognition, from the thought of von Foerster, a multiplicity of processes are enfolded in contributing to interactions and conversations with the world. Foerster (1981) begins the introduction to Observing Systems with a recursive loop (Figure 1).

Figure 1

This diagram points to the overarching recursive computational complexity of human steering (read cognition, sensing, emergence, and human volition) and also points to the internal loops that contribute to the cybernetic steersmanship brought about through human holisticness.

Here, von Foerster simplifies cognition into one overarching recursive approach to computation in the brain/mind. In relation to this, von Foerster provided a definition for what he called a non-trivial machine. Sheena Calvert (2019), in a section of her text titled “The Trivial Machine vs. The Non-Trivial Machine” discussed the difference between these two varieties of machine:

Here, instead of “wheels and cogs” we can discuss processes that become interfunctional through differing forms of interface in the body, and their biomimetic abstraction. Calvert (2019) continues:

Where Calvert was discussing aspects of human language production as well as new forms of machine-oriented language use, we can also discuss the non-trivial machine in another higher-order manner. It is interesting to note that Søren Brier (2008, p. 340) in his book Cybersemiotics, points to von Foerster discussing humans as non-trivial machines. Brier quotes von Foerster: “Let us only make note of the fact that non-trivial machines are dependent on their own history, cannot be analytically determined, and are unpredictable” (von Foerster and Poerksen 2002, p. 58).

The idea of studying the mind as related to a set of physical processes at operation in the brain and the body was suggested very early on by von Neumann (1955):

The paradox of biomimetics relates to the following concept—we must better come to know the workings of the body in order to holistically abstract those complex functionalities into new forms of intelligent systems. The big push behind the interest in neosentience research for me is in designing an intelligent system that is closely aligned with the study of human sentience and sapience, and our own personal enactment of that sentience/sapience as explored through self-examination as well as across the understandings of a plethora of differing related research domains. Also, as I have explored intelligent systems over time, in each case these have helped me better come to understand my own interests in sentience—in observing and experiencing the world, in exploring creativity, in observing knowledge production and creating meta-level systems exploring meaning production.

Another form of Complementarity

I came across an old but very important article recently that seems to suggest a new set of perspectives related to von Neumann’s psycho-physical parallelism. In 1979, H. H. Pattee wrote a paper entitled “The Complementarity Principle and the Origin of Macromolecular Information.” Pattee’s (1979) conclusion points to an interesting set of ideas related to the articulation of such complex bio-systems:

A relevant book discussing approaches to mathematical biology is Integral Biomathics: Tracing The Road to Reality (Simeonov, Smith, & Ehresmann, 2012). In the chapter entitled “Processes and Problems That May Define the New BioMathematics Field,” Robert S. Root-Bernstein (2012) provides the following:

My text in the same volume is entitled “The Engine of Engines – Toward a Computational Ecology.” In it, I discuss the following:

As mentioned above, I think of the body as an ultracomplex bio-machine that has not as yet been fully articulated in terms of the complexity of its relational entailment structures (Seaman, 2013, p. 63).

Unmanageability in Neosentience Research

My approach was initially to ask myself what is it that I do when I try to solve a complex problem that includes one or more domains that are currently outside of my knowledge base? How have others historically explored such problems? In using the Insight Engine 2.0 system interactants might ask themselves what aspects of sentience/sapience production have they as individuals been most interested in? Are interactants interested in approaching these questions though poetics, philosophy, history, the study of cognition, ethics, etc.? What would you as an individual define as the overarching categories to research, and how might these change over time through interaction with the system, reflecting on both human sentience/sapience and the potentials of neosentience? I apologize for the breadth of topics in this text but, in a sense, I want to point out how many individuals might contribute and/or explore the system in their own way, or suggest new categories for relevant inclusion. So, as you are reading this article, take a moment to reflect on what your perspectives related to a contribution might be, and how this branch of research into human sentience/sapience might be put in conversation with others. You can explore the existing categories of research in the list below entitled Insight Engine 2.0 Research Categories—Neosentience.

One can focus on developing a team of collaborators to work on difficult problems, breaking up aspects of the potential solution in logical ways. One can also ask the more difficult question—how and where in the body does emergence take place contributing to human sentience/sapience? This question may be an impossible one to fully answer. One can also develop multiple ways to promote conversation across disciplinary domains. This also suggests the potential of opening up new transdisciplinary approaches to complex problems. This kind of conversation across varying foci explored by differing individuals was central to Glanville’s sense of the design process writ large.

The Insight Engine 2.0 seeks to draw on my long history as a media and AI researcher designing new forms of interfaces, generative environments exploring emergence and meta-meaning (Seaman, 1999b, 2010), as well as new qualities of interactivity, and to expand this research via a strong interdisciplinary/transdisciplinary collaboration that bridges neuroscience, computer science, mathematics, as well as the arts and humanities at Duke, as well as through international collaborations, still to be determined.

This research seeks to work toward the computational authorship of a tool to empower insight production—disciplinary, interdisciplinary, crossdisciplinary, and transdisciplinary individual and team-based research, and to potentially enable bisociational processes as discussed by Koestler in The Act of Creation, bringing disparate conceptual potentials into relevant juxtapositions through focused interaction. Koestler states (1964):

Koestler (1964) continues:

Extending this is my concept of poly-association, which, like employing bisociation, enables intelligent database searches that bring multiple concepts from quite different disciplinary contexts into juxtaposition through AI vector data textual analysis.

The Neosentience book (Seaman & Rössler, 2011) has hundreds of short micro-chapters. In the book I discuss my concept of recombinant informatics (influenced by Koestler’s bisociation). Take the ideas in any two micro-chapters and build a cogent bridge between them to define a new area of research. Here the creative thought of the individual exploring generated contexts through bisociation and/or poly-association comes into play.

The goal of my current work is to author a new interactive computational system of systems which will aid in knowledge production through pointed concept juxtaposition that is user-driven, as well as informed via conversation with AI. Additionally, the goal is to define systems that help us articulate meaningful connections and relational patterns between these disciplines. In the past interplay between disciplines has often been restricted because of particular publishing hegemonies and the limits of historical conceptual arenas.

Again, the goal is to compose a set of creative approaches to intelligent generative systems—meta-meaning systems that explore human intelligence working in concert with artificial intelligence, to advance the fields of both human and artificial intelligence, in the creation of an even higher-order intelligent system of systems. The overarching goal of Insight Engine 2.0 is to enable an international community of researchers, as well as interested individuals, to explore extremely complex problem sets via accessing relevant data.

In the operation of Insight Engine 2.0, a transdisciplinary database and unique set of visualization modes enable researchers from differing disciplines and advanced exploratory research domains to peruse data which may not normally be found in the papers and books related to a specific disciplinary domain. New knowledge and insight potentially arise via the intelligent juxtaposition of one or more such domains. No single discipline of science, the humanities, or the arts can tackle such a difficult information-related problem set—the defining of a model for neosentience production.

A special transdisciplinary distributed international team of teams will potentially arise out of the use of the Insight Engine 2.0, and the variety of forms of conversation it enacts. The Insight Engine 2.0 information system is being designed to support researchers, to empower them to access relevant transdisciplinary data from the database and to contribute to the higher order goal over time of articulating a functional neosentient model as an extended approach to open source knowledge production, in part, accessible via the internet as well as the Unreal game engine. Such a model is to be informed from many intellectual perspectives and transdisciplinary conversations facilitated by the Insight Engine 2.0 system, including a list-serve (still to be set up) and future information-oriented gatherings and relevant links. I often call this kind of approach a multi-perspective approach to knowledge production and, in this case, to neosentience production.

An Ultracomplex Bio-Machine

Robert Rosen (1999) saw the limitations of the machine metaphor in relation to biological research, yet, later in his book, changed his mind in relation to future research. In Essays on Life Itself, Rosen (1999) points to a particular limit, and later discusses a thought about the future in terms of approaching complexity:

Given the time that has past, perhaps researchers can continue to seek to devise a set of ultra-complex simulable models that are interfaced with each other, given the potentials of new computational systems and the development of new conceptual approaches. Additionally, researchers can work toward building a system to articulate new approaches exploring complex mathematics and multi-value logic that seeks to explore new forms of modeling ultra-complex systems, enabling the development of new methodologies to work in tandem with sophisticated high-resolution technologies to explore and map entailment structures. Perhaps the computers and linked advanced scanning technologies we need to render such observational systems operable are still off in the future. The articulation of entailment structures is one of the richest goals of the neosentience project, as is their biomimetic abstraction. Later in Essays on Life Itself, Rosen (1999) questions his initial assumptions and suggests:

The Insight Engine 2.0 seeks in part to provide a technological platform to aid in the research of such ideas and concepts that elucidate future approaches to unpacking biological complexity.

Entailment Structures

Is writing code a craft? The last couple of years have been a very exciting time in terms of how AI code has altered the arts, humanities, and sciences through ChatGPT (Wolfram, 2023) and other OpenAI related systems, as well as a plethora of new AI-based applications (Biswal, 2024) and, in particular, applications exploring creative production in the arts (Zhou & Lee, 2024). The Insight Engine 2.0 project seeks to provide a technological platform to aid in the research of future ideas related to unpacking complexity, defining new forms of mathematics, and articulating methodologies to best map the entailment structures that enable sentience to arise, as well as to explore the development of new forms of code and machinic systems to re-embody the functionality of these processes and entailment structures via biomimetics and bio-abstraction (Seaman, 1999a, pp. 175–176). The full entailment (the study of relevant structures at operation in the body) and the later potential emulation of the body’s functionality, especially related to what I have called the creation of an ultra-complex entailment network in response to Rosen’s thoughts presented above, has to my knowledge not been achieved at this time. This is due in part to the distributed, highly complex nature of the mind/brain, as it falls in relation to the operation of the body, and the environment the body is nested within, where a small event in one part of the brain might not be picked up in terms of the resolution of a technological scan or other contemporary technological methodology as related to multi-modal sensing, yet this might shift elements of thought and in turn lived interaction.

This inability to precisely define highly detailed entailment structures in the brain and body, is also potentially due to data processing strategies. One goal of neosentience research is to track the development of new high-resolution still and time-based scanning technologies. These high-end technological systems will have to be developed to undertake scanning tasks that perhaps will successfully encompass the biological functionality of the entire body. It is exciting to note that a huge breakthrough has just been made here at the Duke Institute for Brain Sciences. Reporting on recent advancements, Dan Vahaba (2023) states that “the refined MRI provides an important new way to visualize the connectivity of the entire brain at record-breaking resolution” (para. 5).

Researchers must come to know the human body better from many different perspectives to elucidate its functional abstraction. This includes researching the potential social and cultural ramifications as well as the ethics surrounding such study. It is the interfacing and concomitant interfunctionality of all of the systems in the body leading to sentience/sapience production that that I and partners are seeking to research and better understand in the service of defining a model for neosentience production.

The Enacted, Embodied, Embedded, Extended Approach

Central to neosentience is research into embodiment and how it relates to knowledge production and language use. Ward and Stapleton (2012) state that “when we say that cognition depends on ‘our activity’, or upon ‘bearing relations to an environment’, the phrases within the quotation marks can be cashed out in many different ways” (p. 90). In terms of the enacted, embodied, embedded, extended approach, Ward and Stapleton suggest that “to be a cognizer, in the sense which interests the enactivist, is to manifest an appropriate degree of attunement to the objects, features, threats and opportunities present in the immediate environment” (p. 91). Ward and Stapleton (2012) continue: “The features of the environment to which the system is attuned are not inert and independent of the system, but dependent upon, and specified at least partly in terms of, the system’s activity and capacities” (p. 91).

One set of neosentient-related goals is to research the embodied, embedded, enactive, and extended approaches to understanding cognition in the human, and then seek to articulate the entailment structures that enable this set of dynamic interrelations to function. This understanding means extending current modalities of AI through multi-modal machine-based sensing and articulating a different set of modalities for defining context, and introspection about context, partly through embodied interactions and synthetic memories of pattern flows of robotic experiences. This embodied approach is quite different from current language driven models of AI, which at this moment, in terms of systems like ChatGPT, are highly successful. Yet, perhaps such language driven approaches still have a distance to travel given the qualities that are operative in terms of human sentience/sapience as exemplified by deep contextual awareness and its abstraction in thought. This in part also plays out via the meta-level of introspection.

The Insight Engine 2.0—Focusing on a Model for the Production of Neosentience

The pragmatic goal is to make the Insight Engine 2.0 function in such a way as to point to, and enfold, new research data across disciplinary boundaries by using advanced information processing, computational linguistics, a natural language API, new open source forms of AI like Llama and Mistral and additional forms of AI acting as Micropeers (AI collaborators to enable the development of ecologies of intelligent relationalities in terms of specific research questions and related prompts). Additionally, the development of new informational paradigms through bisociation (Koestler, 1964, pp. 35–36) and poly-association are explored. Such research is about promoting new forms of connection and juxtaposition that transcend traditional boundaries of knowledge production. This research explores mutually beneficial articulated relational structures between disciplines that may have not been historically defined or as yet codified. It is my goal as a computational media and AI theorist to make systems that can help play midwife to new thoughts related to the future of AI. I seek to author interactive generative systems, working with my team, that enables new forms of academic border crossing in the service of new knowledge, perhaps bringing ideas into juxtaposition that historically have been held at a distance due in part to the hegemonies of differing research publications.

I seek, with an advanced team, to create AI-driven insight-oriented systems that bring disparate aspects of knowledge production into play (into juxtaposition) where new fields of relational understanding need to be negotiated and discussed. Concepts and potentials will additionally be suggested by conversations with new forms of AI, as well as through individual search mechanisms. Thus, my take on this has been to research and author systems that promote the cogent generative relational bridging of disciplinary practices through new forms of creative code authorship, as well as through stimulated conversation between human interactants. This kind of system promotes differing varieties of conversation including conversations with the self after reading papers as well as exploring conversation that becomes illuminated by exploiting the natural language potentials of AI. Central is the exploration of nested high-end systems that enable elements of AI, large language models, and machine learning to become operative as part of this system of systems—Insight Engine 2.0.

There are many new interface-related areas to be explored. Imagine employing a form of gravity used in game engines to be repurposed to enable specific foci to be repositioned in virtual space via a repurposed physics engine, where relevant ideas might become attracted to each other given appropriate meta-data, interactively reconfiguring a spatial textual environment. Additionally, conversations with material systems could be facilitated that the Insight Engine links to—i.e., exploring models for new forms of biomimetic computation. This might empower the navigation of a biological model juxtaposed with a code-based model enabling the perusal of elements of human functionality as related to the computer code that is mimicking it artificially. Such a system may also enable the exploration of new linked databases, external to the system, mapping the body and its functionalities. The researcher/interactant could potentially explore new forms of data generated by scanning that, once made operative in the system, could be navigated and at any point could be queried for related relational information. An individual could navigate to exact locations in the body and the brain, and access relevant linked papers or other differing scale models related to that aspect of human biofunctionality. One might also have conversations with relevant historical artifacts, exploring quotes from researchers and relevant papers driven by individuals’ interaction and navigation.

Call this a set of advanced conversations surrounding the perusal of ideas empowered via user-driven navigations of this self-organizing transdisciplinary search engine and visualization system. This Insight Engine 2.0’s functionality will in part incorporate moments of chosen chance processes as part of a menu system. Central is ongoing navigation leading to focused human and computationally articulated relationality, with new ideas driven by pointed queries via open-source large language model prompts as well as simple human queries of the relational database, and playful explorations of flying through 3D models and data; ideas that relevantly cross non-traditional disciplinary domains using the natural language AI and the navigable computational environment to help connect and articulate relationalities between domains; this includes the navigation of 3D worlds; ideas where visualization and in the future sonification and haptics-oriented interfaces are potentially explored enabling new means to examine datasets and related information annotations. Such a system seeks to provide a novel sense of interaction in terms of where to navigate in a spatial ecology of knowledge-production-approaches. These interfaces literally present a new set of perspectives that are providing an alternative to many traditional search methodologies. Yet, traditional approaches are also part of the menu system. One can move back and forth as needed from 3D models and visualizations (presented in 2d or 3d) to specific forms of datasets and to purely textual articulations and searches accessed in a more traditionally viewed dashboard.

Research Areas That Will Populate the Relational Database and Visualization Interfaces

Although Glanville in conversation with Seaman did not like certain forms of lists, as discussed above, I have explored a very broad set of initial research areas to be included in the database. In particular, this summer (2024) the list became highly articulated. As you read this very long list of foci, please consider writing to me with your ideas about any improvements. These different research areas forming the overarching database, explored through many different interface forms, include the following (although new research areas will continuously be added as they are defined):