Augmentative Psychology-1-What Are You Becoming Through Your AI Engagement?
Augmentative Psychology Series — Article 1 of 2
Every generation invents tools that change what people can do.
Artificial intelligence is different.
It is changing the conditions under which people become who they are.
Every interaction with an AI system accomplishes two things simultaneously. One task in the world, and one quiet shaping of the capacities the person brings to their next task. Sometimes that shaping deepens judgment, enriches discernment, and strengthens the person’s capacity for genuine professional engagement. Sometimes it erodes those same capacities through disuse. We have studied the task extensively. We barely have a framework for understanding and deliberately designing the shaping.
The questions that AI has brought to the forefront cannot be adequately addressed within existing conceptual boundaries. They require a developmental framework that takes the trajectory of human capacities as its primary unit of analysis and the intentional design of those trajectories as its practical concern. That framework forms the foundation of an emerging field. Augmentative Psychology.
What the field is for
Augmentative Psychology is the science of developmental relationships between human beings and the augmentative systems with which they continuously interact. It studies how those relationships shape the development of human capacities, and how they can be designed to strengthen rather than diminish human judgment, agency, and flourishing (Unnithan, 2025b)
This is not simply a theory about AI. It is a science of human development in an age of augmentation, and its purpose is practical as much as theoretical. It exists to help individuals, educators, professionals, and organisations answer questions that existing frameworks do not yet address. How can I strengthen my judgment while using AI rather than outsourcing it? How do I preserve my capacity for deep thinking, creativity, and discernment when AI handles more of my cognitive work? How should AI be integrated into education and work so that it develops rather than diminishes human capability?
The field rests on three intellectual commitments that together define its distinctive character.
The first is a developmental commitment. Human capacities are not fixed properties of individuals but change over time, forming trajectories that move toward greater judgment, agency, and flourishing, or away from them. This is what distinguishes Augmentative Psychology from fields that study AI’s effects on behaviour or experience in a given interaction. The person is not stable. They are on a trajectory, and that trajectory is the primary object of study.
The second is a relational commitment. Those developmental changes do not arise inside the person or inside the system, but through sustained coupling between them. Human capacities are emergent qualities of the developmental relationship, not properties that the person brings to it independently. This is why the relationship matters, where development actually happens.
The third is the normative commitment. If the developmental trajectories of human–AI engagement are shaped by identifiable conditions rather than determined by the technology itself, then those conditions become legitimate objects of intentional design. Augmentative Psychology therefore seeks not only to explain how human capacities change through sustained AI engagement, but also to identify the educational, organisational, technological, and policy conditions under which those capacities are more likely to be strengthened rather than diminished. In this sense, it is both a descriptive science of human development under augmentation and a design science concerned with cultivating human flourishing.
The evidence is already accumulating
This is not a philosophical speculation. Evidence is already accumulating from multiple independent research streams that sustained AI engagement is reshaping human cognitive architecture in ways that performance metrics alone cannot capture.
Research on cognitive offloading shows that when people know information is externally available, they are less likely to encode it internally, and that AI advances this effect beyond what earlier external storage technologies produced because it generates answers rather than merely storing them (Sparrow, Liu, and Wegner, 2011; Oakley, Sejnowski, and Venkataraman, 2025). Studies on AI-assisted education document that learners who receive AI-generated solutions consistently outperform peers on immediate assessments while showing weaker neural engagement, poorer memory consolidation, and more formulaic thinking over time (Lee et al., 2025). Research on cognitive offloading among student populations found a significant negative correlation between reliance on AI tools and critical thinking skills among 666 participants across multiple countries (Gerlich, 2024). Unnithan’s (2025a) research with high school students in Kerala, India also showed similar findings. Research on algorithmic management in the gig economy documents systematic deskilling, the progressive atrophy of spatial judgment, interpersonal discernment, and professional autonomy in workers whose every decision is governed by platform metrics (Alacovska, 2025; Kadolkar et al., 2025). Research on AI companionship shows emotional dependency patterns developing alongside declining tolerance for the complexity and unpredictability of genuine human relationships (Fang et al., 2025; Laestadius et al., 2024).
These are not isolated findings. They are different measurements of the same underlying dynamic. Sustained AI engagement is producing divergent developmental trajectories, and current frameworks provide only partial tools for naming, predicting, or intervening in those trajectories with precision.
Why psychology needs this framework
Psychology has historically studied human development through interactions with parents, peers, institutions, and culture. Yet increasingly, one of the most persistent developmental partners in modern life is neither human nor institution, but an intelligent technological system. These systems do not merely deliver information. They participate in perception, memory, judgment, attention, and decision-making in ways that continuously shape the person engaging with them.
Existing frameworks provide valuable insights into particular aspects of human-AI interaction, but developmental trajectory has yet to become the primary organising principle of the field. The gap has not gone unnoticed within the discipline itself.
In 2001, Krämer and Bente called for psychologists to enter the emerging field of human-technology interaction and contribute their expertise to advancing it. Twenty years later, reviewing what had actually happened, they found the opposite. Psychological knowledge and methods had been adopted by computer scientists, while psychologists themselves had largely stayed out. Their conclusion was pointed. While psychology is indispensable, they wrote, psychologists are not (Krämer and Bente, 2021, p. 255).
A word on why the field is called Augmentative Psychology, because the name is not accidental and the existing alternatives do not fit.
Human-AI psychology is too narrow. It ties the field to a specific technology rather than to the developmental question. As AI and other augmenting systems take forms we cannot yet anticipate, a field named for today’s technology would need to reinvent itself with every major technological turn. Cyberpsychology studies behaviour in digital environments. Human-computer interaction focuses on usability and interface design. Augmented Cognition, the closest neighbouring field, asks how systems should adapt to a user’s present cognitive state to optimise performance in the moment. AI psychology is ambiguous, covering everything from chatbot design to machine consciousness. None of these places the developmental consequences of sustained augmentation at the centre of inquiry. Augmentative Psychology was chosen because its central concern is neither technology nor interaction but the developmental trajectory of the human being within the augmenting relationship.
The term itself carries a precise intellectual lineage. In Augmentative and Alternative Communication, the clinical field that helps people with speech impairments (ASHA, 2004), augmentative specifically refers to technology that supplements existing human capacity rather than replaces it, in direct contrast to alternative, which refers to replacement. Douglas Engelbart’s foundational 1962 framework for augmenting human intellect positioned computers as partners that extend human reasoning rather than as substitutes for it (Engelbart, 1962). Across these traditions, augmentation carries a consistent developmental intuition that technology is valuable when it expands what people are capable of becoming.
Augmentative Psychology adopts this orientation and shifts the question. Not what does technology augment, but what does sustained augmentation do to the person over time. Whereas Augmented Cognition asks how systems should adapt to the user’s present cognitive state (Schmorrow and Kruse, 2004), Augmentative Psychology asks how repeated interaction with those systems changes the user’s future cognitive capacities. The field treats augmentation not as a property of technology but as a developmental relationship whose consequences unfold across time.
Underlying this is a specific ontological commitment. Augmentative Psychology rests on a relational ontology in which human capacities are not properties of isolated individuals nor outputs of technological systems, but emergent qualities of the developmental relationships between them. Grounded in the philosophical principle of dependent coarising, the field holds that the person and the augmenting system coconstitute each other through sustained engagement, and that what either becomes cannot be understood apart from the history of that relationship.
A 2025 peer-reviewed paper arguing formally for a new interdisciplinary field of AI psychology observes that existing frameworks including cyberpsychology and human-computer interaction capture only part of the resulting cognitive, emotional, developmental, and social dynamics, and that traditional psychological frameworks have largely viewed AI as a technology or medium rather than as a partner in social interaction (Joseph, 2025). The British Psychological Society President has stated that the question is no longer whether AI will shape psychological practice, but whether psychologists will have shaped it when it does (BPS, 2024). A senior APA researcher noted there has never been a better time for psychologists to identify key drivers of human flourishing in AI-augmented environments (APA Monitor, 2024).
What these calls share is the recognition of a gap. What they do not yet provide is a framework that takes developmental trajectory as its primary unit of analysis, specifies the mechanism through which AI engagement shapes that trajectory, and generates falsifiable propositions about what moderates the direction of change. Augmentative Psychology is the attempt to build that framework.
The central proposition
AI is no longer simply changing what humans can do. It is changing the conditions under which humans become who they are.
Get Dr. Vijayakumar Parameswaran Unnithan’s stories in your inbox
Join Medium for free to get updates from this writer.
This shifts the research question fundamentally. Rather than asking whether AI makes people more productive, Augmentative Psychology asks about the kinds of human beings are emerging through sustained participation in AI-augmented environments, and how those conditions can be intentionally designed to produce wiser, more discerning, more fully flourishing people.
Rather than treating AI as an external technology acting upon stable individuals, it conceptualises human and AI as engaged in continuous developmental coupling, a relational process in which every interaction is simultaneously an opportunity for capacity formation or capacity erosion.
This is not a claim that AI is harmful. It is a claim that the developmental consequences of AI engagement are real, are divergent depending on conditions, and are largely invisible to the performance-based measures that dominate current evaluation. The field’s promise is that those conditions are identifiable, designable, and worth getting right.
Four trajectories, not one
The most practically useful contribution of this framework is the identification of four qualitatively distinct developmental trajectories that sustained AI engagement may produce. Current discourse tends to treat augmentation as a binary. AI either helps or harms. The reality is more precise and more actionable.
Augmentation. Human capacities genuinely strengthen through AI engagement. The professional brings genuine curiosity to AI outputs, holds them provisionally, tests them against their own judgment, and uses the interaction to extend rather than replace their own reasoning. Capacities for discernment, attunement, and independent thought grow. This is the trajectory that good design can produce.
Compensation. AI fills temporary gaps without producing lasting developmental change. Genuine scaffolding in the educational psychology sense. Temporary support that is withdrawn as the person’s own capacity develops. The most benign trajectory, and the most unstable. It drifts toward augmentation if approached regeneratively, toward substitution if approached extractively.
Substitution. AI progressively replaces capacities rather than developing them. Outputs are produced through the AI’s capability rather than the person’s own developing capacity. Short-term efficiency is real. Long-term capacity narrowing is invisible at the performance level until the AI is withdrawn or the situation demands more than it can provide.
Atrophy. The person becomes progressively less able to perform the relevant capacities even when the AI is absent. Atrophy is analytically distinct from substitution. Substitution describes what the AI does instead of the person; atrophy describes what the person becomes as a result. A professional on the Substitution trajectory delegates to AI but retains the underlying capacity. A professional on the Atrophy trajectory has lost it. The diagnostic test is withdrawal. Remove the AI and measure performance. Substitution produces reduced efficiency; atrophy produces genuine incapacity. This makes atrophy the most consequential trajectory and the most empirically tractable. We can confirm it now, not only longitudinally, by testing capacity under AI-absent conditions.
These four trajectories are not successive stages of human development. They are alternative developmental pathways that emerge from different patterns of sustained human-AI engagement. Individuals, teams, organisations, and even professions may move between these pathways over time as the conditions of augmentation change. A clinician on a Substitution trajectory today may move toward Augmentation if the design of their AI tools and the recognition environment around them changes. A student on a Compensation trajectory may drift toward Atrophy if the scaffolding is never gradually withdrawn. The trajectories are tendencies under current conditions, not fixed destinations.
The critical question the framework asks of any AI deployment is about the kind of trajectory the people in this system are on. And what would it take to move them toward augmentation, if need be?
Coming next
These four trajectories raise the question that matters most practically. What actually determines which trajectory a person, a profession, or an organisation ends up on? The answer, and it is perhaps the framework’s most counterintuitive finding, is that the trajectory is not determined by the AI system itself. The same tool produces augmentation in one person and atrophy in another. What determines the direction, what can be deliberately designed, and what this means for governance, organisations, education, and the professions is the subject of the second article in this series.
The full theoretical architecture, including formal constructs, empirically testable propositions, a multi-level coupling framework, and a research programme spanning different domains of psychology, is developed in the forthcoming working paper titled Augmentative Psychology: Foundations of a Field (Unnithan, 2026).
References
Alacovska, A. (2025). Algorithmic paranoia: Gig workers’ affective experience of abusive algorithmic management. New Technology, Work and Employment, 40(1), 45–68.
APA Monitor. (2024). Technology and the future of psychological practice. Monitor on Psychology, 55(3).
ASHA. (2004). Augmentative and alternative communication: Knowledge and skills for service delivery. American Journal of Speech-Language Pathology, 13(2), 77–83.
BPS. (2024). President’s column: AI and the future of psychology. The Psychologist. British Psychological Society.
Engelbart, D. C. (1962). Augmenting human intellect: A conceptual framework. Stanford Research Institute. https://www.dougengelbart.org/content/view/138/
Fang, C. M., Liu, A. R., Danry, V., Lee, E., Chan, S. W. T., Pataranutaporn, P., and Agarwal, S. (2025). How AI and human behaviors shape psychosocial effects of extended chatbot use: A longitudinal randomised controlled study. arXiv preprint arXiv:2503.17473.
Gerlich, M. (2024). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. Societies, 14(1), 1–16.
Joseph, J. (2025). The algorithmic self: How AI is reshaping human identity, introspection, and agency. Frontiers in Psychology, 16. https://doi.org/10.3389/fpsyg.2025.1645795
Kadolkar, I., et al. (2025). Algorithmic management in the gig economy: A systematic review and research integration. Journal of Organizational Behavior, 46(2), 112–139.
Krämer, N., and Bente, G. (2021). Interactions with artificial entities reloaded: 20 years of research from a social psychological perspective. i-com, 20(3), 253–262. https://doi.org/10.1515/icom-2021-0032
Laestadius, L., Bishop, A., Gonzalez, M., Illenčík, D., and Campos-Castillo, C. (2024). Too human and not human enough: Mental health harms from emotional dependence on the social chatbot Replika. New Media and Society, 26(10), 5923–5941.
Oakley, B., Sejnowski, T., and Venkataraman, A. (2025). The memory paradox. Nature Human Behaviour, 9(1), 12–19.
Schmorrow, D. D., and Kruse, A. A. (2004). Augmented cognition. In W. S. Bainbridge (Ed.), Berkshire Encyclopedia of Human-Computer Interaction (pp. 54–59). Berkshire Publishing Group.
Sparrow, B., Liu, J., and Wegner, D. M. (2011). Google effects on memory: Cognitive consequences of having information at our fingertips. Science, 333(6043), 776–778.
Unnithan, V. P. (2025a) Am I thinking? CareerTotus Working Paper Series.
Unnithan, V P (2025b) Emerging Careers in Augmentative Psychology: Advancing a Human-Centred Artificial Intelligence Future. Medium
Unnithan, V. P. (2026). Augmentative psychology: Foundations of a field. CareerTotus Working Paper Series. Forthcoming.
