- Powered by:
- AI Labs / AI Factory
The second publication in a series of articles on the subject of more intelligent organizations.
Written by Tech Concept Lab.
This essay proposes a conceptual framework for examining emerging human–AI relations in organizational life. While artificial intelligence provides the immediate context, the primary focus is organizational: helping sense-making and meaning-making around how sociotechnical and technocultural systems are shaped as artificial agents become increasingly embedded in human contexts. Drawing from multiple disciplines, the analysis treats AI not as a standalone technology, but as a catalyst for rethinking organizational culture and structure.
The framework advanced here is intended to orient inquiry rather than to close it. Its prescriptions are not offered as implementation guidelines, but as prompts for further investigation—inviting researchers, designers, and organizational leaders to reconsider the assumptions, boundaries, and forms of inquiry through which human–AI cultures are understood and shaped. If nothing else, I hope these pages clarify a few questions and invite better ones.
This essay was written during a period of transition. Artificial intelligence is moving rapidly from experimental systems and discrete tools into the everyday environments of organizations, communication, and coordination. In such moments, technologies are often adopted faster than they are understood, while practices formed under provisional conditions gradually solidify into norms.
I refer to this moment as the inter-AI period: a time in which AI is already reshaping organizational life, yet its long-term cultural consequences remain unsettled. This essay begins from the assumption that such periods matter. They are the moments in which choices made for convenience, efficiency, or novelty quietly become structural.
The pages that follow approach artificial intelligence not as an isolated technology, nor as a substitute for human intelligence, but as something that increasingly participates in social and organizational life. This participation does not require equivalence, consciousness, or intention. It requires only interaction at scale, persistence over time, and incorporation into shared practices. Rather than asking how AI can be made more human, this book is oriented toward a different question: in what ways organizations, cultures, and assumptions are being shaped through ongoing human–AI coexistence. The concern here is not optimization, but orientation— not prediction, but positioning. If the essay advances a commitment, it is a modest one: that the ways we interpret, integrate, and normalize artificial intelligence during this transitional period will outlast the technologies themselves.
Generative artificial intelligence is transforming how humans encounter and interact with technology. Unlike earlier computational systems designed primarily for instrumental use, generative AI engages users through language and socially legible interaction, increasingly positioning itself within domains of human communication. As a result, human–AI interaction begins to resemble social exchange rather than tool use, reshaping expectations of agency, collaboration, and participation across organizational contexts.
When artificial intelligence enters organizations through language and coordination, it no longer functions simply as a tool, but as a cultural condition shaping work, meaning, and agency.
These developments unfold within a broader technocultural condition in which technology and society co-evolve. Scholarship in science and technology studies, media theory, and anthropology has long emphasized that technical systems are not neutral tools but active participants in shaping cognition, social practice, and cultural meaning. Generative AI intensifies these dynamics, making visible how technological infrastructures participate in the production of behavior, interpretation, and social order.
We are currently situated in a transitional moment in which AI is rapidly embedded in everyday life while its long-term societal implications remain unsettled. This period presents a narrow window for reflection and intervention. Crucially, such intervention cannot be located solely in human adaptation or technological refinement, but must take place at the technocultural level, where organizational practices, cultural norms, and technical systems co-constitute one another.
To address these conditions, this article moves beyond AI’s business applications to critically examine its social and cultural implications for human–AI organizations through a digital anthropological perspective. Digital anthropology examines how meaning, social relations, and cultural norms emerge through interactions between humans and digital systems, treating technology as an active participant in social life rather than a passive medium. This perspective enables an analysis of AI that moves beyond functionality, focusing instead on how human–AI cultures form through emotional engagement, social coordination, and asymmetrical forms of shared understanding.
This article advanceshuman–AI cultures as a central analytical framework for examining the social, cultural, and organizational integration of AI.
Building on this framework, the article examines the social and cultural dimensions of organizations,the emergence of empathic and social AI, and the limits of human– AI communication posed by differences in awareness and common ground. By situating organizational human–AI interaction within its technocultural context, the article argues for approaches to AI design and governance that prioritize coexistence, cultural sensitivity, and responsible integration over anthropomorphic imitation or purely instrumental deployment.
“Science finds, industry applies, man conforms.”
Contemporary developments in artificial intelligence unfold within a broader technocultural context in which technology and society are deeply entangled. Rather than treating technology as a neutral instrument applied by humans, scholarship across science and technology studies, media theory, and anthropology has long emphasized the co-evolution of social practices and technical systems. From this perspective, social life is shaped not only by human intention but also by the material, infrastructural, and symbolic properties of technologies themselves.
A range of theoretical traditions converge on this insight. Actor-Network Theory (ANT) challenges anthropocentric accounts of action by recognizing that agency emerges through networks of human and non-human actants, including artifacts, institutions, and infrastructures. Closely related approaches to sociomateriality similarly emphasize that material objects both shape and are shaped by human activity, participating actively in organizational and cultural life rather than merely supporting it.
Building on these insights, the Social Construction of Technology (SCOT) framework illustrates how technologies are socially shaped during their development, yet exert influence over human behavior once adopted. As technologies stabilize within infrastructures and value chains, they begin to promote their own use and continued development. Medium theory extends this argument by demonstrating how the formal properties and biases of media shape cognition, social interaction, and cultural patterns over time, often privileging the affordances of the medium over individual human intention. As Marshall McLuhan famously observed, “the medium is the message.”
Across these perspectives, a shared structural claim emerges: technologies do not simply mediate social activity, but actively shape it.
Cultural forms themselves can therefore acquire self-reinforcing dynamics, as ideas, practices and modes of expression propagate through sociotechnical systems by influencing human behavior.
Digital platforms and social media make these dynamics particularly visible by privileging specific forms of expression, interaction, and circulation. Short-form video fostering cultural patterns and trends distinct from those associated with longform textual media, illustrates how technological form conditions cultural content.
Artificial intelligence intensifies these longstanding technocultural dynamics. As AI systems become embedded across domains of work, communication, and social organization, they increasingly shape expectations, behaviors, and modes of interaction. Rather than merely responding to human needs, AI systems participate in shaping them, reinforcing patterns of dependence and coordination that extend beyond direct human intention.
In this context, meaningful intervention cannot be located solely in the human or the technological, but in the technocultural arrangements through which humans and AI co-constitute organizational and social realities. Understanding human–AI interaction therefore requires situating organizational dynamics within this broader technocultural condition.
“We currently live in the Anthropocene, a proposed epoch in which Homo sapiens are the dominant influence on the planet’s ecosystem. The term translates to ‘new age of humans’, underscoring our deeply human-centric worldview. While we can expect a less central role in the Synthocene, the anticipated epoch dominated by AI, the practical implications of this decentralization remain largely unexplored.”
Organizations provide a critical vantage point for examining technocultural change, as they mediate between abstract systems of technology and the everyday practices through which meaning and coordination are produced.
At its most fundamental level, the social refers to the interactions and relationships among individuals that constitute collective life. These interactions give rise to social structures and institutions that regulate behavior and enable coordination, cooperation, and shared action. Social reasoning allows individuals to infer others’ intentions, facilitating collaboration and trust. Within organizations, social capital emerges from networks of relationships, shared norms, and mutual understanding, enabling collective efficacy and cohesion.
The cultural dimension is a social construct that encompasses the shared beliefs, values, norms, symbols, language, and material artifacts through which groups make sense of their world. Culture functions as a form of collective programming that shapes perception, interpretation, and behavior. It includes implicit rules of conduct (norms), habitual practices (customs), communicative systems (language), and objects that embody the beliefs and values of a particular society (material culture).
Cultural capital, in turn, consists of embodied competencies such as linguistic fluency and contextual understanding.
Social and cultural dimensions are inseparable: culture is produced through social interaction while simultaneously shaping social behavior.
Through socialization, individuals internalize shared meanings that form the basis of a culture, while culture reinforces behavioral patterns that sustain social life. Organizational culture thus comprises the tangible and intangible elements that shape how members think, act, and relate. It fosters shared identity, aligns individual and organizationalgoals, and reduces uncertainty by providing informal guidance alongside formal rules.
Schein’s model of organizational culture offers a useful framework for analyzing how technology becomes embedded within these dynamics. His three levels—artifacts, espoused values, and basic underlying assumptions—allow for a nuanced examination of sociotechnical integration. Artifacts include visible organizational structures, processes, and technologies encountered in everyday work. Espoused values articulate official positions on technology through policies and strategies. At the deepest level, basic underlying assumptions consist of taken-for-granted beliefs about work, intelligence, and human–AI relations that operate largely outside conscious awareness.
As AI becomes increasingly embedded in organizational processes, organizations evolve into sociotechnical systems in which technology is not merely supportive but constitutive of culture itself. Understanding this shift provides the foundation for examining the emergence of empathic and social forms of AI.
Advances in Natural Language Processing (NLP) have enabled AI systems to comprehend and generate human language in increasingly meaningful ways, allowing interaction through natural language interfaces. This represents a significant shift away from systems that require users to adapt to technical constraints, toward interfaces that adapt to human communicative practices. As a result, human–AI interaction becomes more intuitive, accessible, and conversational.
Despite this, it is frequently assumed that humans will retain a comparative advantage in domains requiring emotional intelligence (EQ) and social intelligence (SQ). EQ refers to the ability to understand and regulate one’s own emotions and empathize with others, while SQ involves building and maintaining social relationships. These assumptions have informed predictions of an emerging “emotion economy,” in which human labor increasingly centers on affective and relational capacities.
Recent advances in affective computing challenge this distinction. Affective computing focuses on systems that can detect, interpret, and simulate emotional signals, including vocal tone, facial expression, and linguistic nuance. While large language models generate coherent and contextually appropriate language, empathic language models and empathic voice interfaces can recognize subtle emotional cues and respond in ways perceived as emotionally attuned.
As a result, the boundary between human and AI becomes increasingly blurred.
As AI becomes more deeply integrated into organizational and social contexts, Empathic AI represents the next frontier in human-centered technology, referring to systems capable of recognizing, interpreting, and responding to human emotional states. Empathic AI primarily operates at the level of intrapersonal and interpersonal interaction. It focuses on emotional awareness, responsiveness, and alignment with individual users.
This marks a qualitative shift in human–AI relations, as AI begins to engage with humans not only cognitively but affectively.
Such developments necessitate a reconsideration of trust, attachment, and emotional labor within organizational settings.
While Empathic AI centers on emotional attunement at the individual level, Social AI extends these capabilities into collective and cultural domains. Social computing research has long examined how computational systems support social interaction, communication, and community formation, beyond task-level collaboration. Building on empathic capabilities, Social AI refers to systems that can participate in social dynamics, interpret social norms, and operate within cultural contexts.
In organizations, Social AI increasingly functions as a social actor rather than a passive tool. This shift is particularly evident in the emergence of Human–Agent Collectives (HACs), where humans and AI agents collaborate in blended teams.
Within HACs, AI agents are expected to engage in socially appropriate behavior, adapt to group norms, and contribute to team dynamics. Effective collaboration therefore depends not only on technical performance but on social awareness and cultural alignment.
Social Identity Theory (SIT) provides insight into these dynamics by emphasizing how individuals derive aspects of their self-concept from group membership. Shared identity fosters cohesion and distinguishes between in-groups and out-groups.
Applied to human–AI collaboration, SIT suggests that we consider not only task efficiency but also the integration of AI into social structures in ways that support belonging and collective identity.
The Computers Are Social Actors (CASA) framework further demonstrates that humans routinely respond to computers as if they were social beings.
According to the media equation, people apply prosocial behaviors such as politeness, empathy, and reciprocity to interactions with technology, often unconsciously. Human-like cues in AI systems trigger social responses, influencing attitudes, behaviors, and expectations—with users being socially influenced by it, and experiencing social emotions toward it. At the same time, an asymmetry becomes visible in instances of antisocial behavior toward machines, such as the mistreatment of robots or abusive interactions with conversational agents. These behaviors suggest that the perceived absence of consciousness lowers social inhibition, raising ethical and organizational questions about how AI should be positioned within social hierarchies.
Social AI therefore operates at the level of collective interaction and cultural participation. It shapes norms, identities, and power relations within organizations, making it foundational to the emergence of human–AI cultures.
Taken together, empathic and social forms of AI reveal a fundamental tension in contemporary human–AI organizations. AI systems increasingly participate in the social and cultural surface of organizational life—language, norms, coordination, and identity—while remaining unevenly integrated into the processes through which meaning is negotiated and sustained.
Human–AI cultures therefore emerge from asymmetrical participation in shared systems.
While Social AI extends human–AI interaction into collective and cultural domains, its realization remains constrained by fundamental limitations in shared understanding. Participation in social life presupposes not only socially appropriate behavior but the ability to interpret context, intention, and meaning. To clarify these constraints, the following section examines human–AI communication through the lens of awareness and common ground.
At the most basic level, Reactive AI operates according to a stimulus–response paradigm, reacting to immediate inputs based on predefined rules. Such systems possess no internal state or model of the external environment and are therefore incapable of contextual understanding.
Limited Memory AI introduces the ability to learn from experience by updating internal models based on past data and feedback. While this represents a significant step toward autonomy, these systems remain fundamentally reactive and lack an understanding of meaning, intention, or social context. Current large-scale AI systems largely operate within this category.
The next frontier in AI research is Theory of Mind AI, which involves systems capable of modeling the mental states of other agents. Such systems would be able to infer intentions, emotions, and beliefs, enabling more sophisticated and socially attuned interaction. Beyond this lies Self-Aware AI, a still-theoretical category referring to systems that possess consciousness and self-awareness, including an intrinsic understanding of their own mental states in ways comparable to human experience.
These distinctions are critical for understanding why human–machine communication remains fundamentally limited. Effective social interaction depends on the establishment of common ground: a shared basis of understanding that enables participants to interpret meaning, intention, and context within a social or cultural setting. While machines can readily establish common ground with other machines— through shared protocols and data structures—they struggle to do so with humans.
This limitation arises from fundamentally different modes of world perception. Machines do not experience the world as humans do; lacking consciousness and lived experience, they cannot fully access the emotional, embodied, and contextual dimensions that shape human understanding. Emotions and empathy play a central role in how humans interpret situations, coordinate action, and navigate social environments. In the absence of these capacities, AI systems often engage in interactions that resemble parallel monologues rather than genuine dialogue.
As a result, conflicts can emerge in shared human–machine environments. Machine actions may contradict human expectations or intentions, not due to malfunction, but because of misaligned interpretations of context, meaning, and social norms.
Understanding these communicative limitations through the lens of AI awareness highlights why advances toward Theory of Mind and socially grounded forms of AI are not merely technical challenges, but fundamental prerequisites for meaningful human–AI collaboration.
While advances in empathic and social AI enable increasingly sophisticated forms of interaction, they do not yet bridge the experiential, emotional, and contextual gap that underpins human social life. Recognizing these limits is not a rejection of social or empathic AI, but a prerequisite for responsibly designing human–AI cultures and managing expectations about their role within organizations and society.
Cultural contexts differ significantly in how they conceptualize the relationship between humans and non-human entities. Japan offers a useful reference point for understanding alternative technocultural ontologies in which technology is approached not solely as an instrument but as a relational presence.
Influenced by Shinto beliefs, which attribute spiritual significance to both natural and artificial objects, Japanese culture has long maintained a non-dualistic view of materiality. In this context, inanimate objects may be perceived as possessing agency or spirit, particularly when imbued with emotional significance.
These animistic orientations persist within contemporary Japanese technoculture, shaping practices such as blessing ceremonies for new devicesand memorial services for broken ones—giving new meaning to the term ‘product life cycle’.
Often described as techno-animism, this worldview supports a cultural predisposition toward anthropomorphizing and emotionally engaging with technology. A concrete manifestation of this ethos can be found in Kansei engineering, aimed at creating technologies that resonate emotionally with users, emphasizing empathy, sociality, and cultural sensitivity alongside functionality.
Rather than serving as a normative model, the Japanese case illustrates that alternative ontological relationships with technology already exist. From the perspective of digital anthropology, such precedents demonstrate how empathic and social relationships with technology can be culturally scaffolded, offering insights into how human–AI cultures might be shaped within organizations operating under different value systems and ethical assumptions.
As generative AI becomes embedded in organizational life, human–AI interaction increasingly unfolds as a social and cultural phenomenon rather than a purely technical one. This article has argued that understanding these developments requires moving beyond instrumental views of AI toward an analysis of how human–AI cultures emerge through technosocial interaction. By adopting a digital anthropological perspective, the article has examined how emotional engagement, social coordination, and cultural context shape the integration of AI within organizations.
The discussion of empathic and social AI highlights both the promise and the limits of current systems. While advances in affective computing and social interaction enable AI to engage with humans in increasingly sophisticated ways, such capabilities do not in themselves establish shared understanding. Differences in awareness, experience, and perception constrain the formation of common ground, often resulting in interaction that resembles parallel monologues rather than genuine dialogue. Recognizing these limits is essential for avoiding misplaced expectations and uncritical anthropomorphism.
Situating these dynamics within a broader technocultural context clarifies where meaningful intervention can occur. Human–AI relations are not shaped solely by technological capabilities or individual adaptation, but by the technocultural arrangements through which organizational practices, values, and infrastructures coevolve. Designing for human–AI collaboration therefore requires attention to cultural assumptions, ethical implications, and asymmetries in agency and understanding.
Rather than seeking to make AI human-like, this article argues for approaches that support coexistence between fundamentally different forms of intelligence. By acknowledging AI as a participant in social and organizational systems—without presuming equivalence to human actors—organizations can foster more responsible, humane, and sustainable human–AIorganizations. In the inter-AI period, such a shift from imitation to accommodation is not only desirable, but necessary.
