Artificial Intelligence and Psychoanalysis: A New Concept of Research Methodology
Om Prakash Singh, PhD
Master Training of Trainers (MToT), Education Training Center
Koshi Province, Inaruwa, Sunsari
Nepal
Received: March 10, 2024; Revised & Accepted: June 23, 2024
Copyright: Singh,
(2024)
Abstract
The recent high
performance of ChatGPT on several standardized academic tests has thrust the
topic of artificial intelligence (AI) into the mainstream conversation about
the future of education. As deep learning is poised to shift the teaching
paradigm, it is essential to have a clear understanding of its effects on the
current education system to ensure sustainable development and deployment of
AI-driven technologies at schools and universities. Hence, AI behavior cannot
be fully understood without human and social sciences. After the imaginary and
symbolic registers, AI is the third register of identification. Therefore, AI
extends the movement that is at work in the Lacanian interpretation of the
mirror stage and Oedipus complex and which Latour’s reading helps us to clarify.
From this point of view, I describe an AI system as a set of three contrasting
forces: the human desire for identification, logic, and machinery. In the
“Miscomputation and information” section, I show how this interpretative model
improves our understanding of AI. Systematic research on psychoanalytic
treatments has been limited by several factors, including a belief that
clinical experience can demonstrate the effectiveness of psychoanalysis,
rendering systematic research unnecessary, the view that psychoanalytic
research would be difficult or impossible to accomplish, and a concern that
research would distort the treatment being delivered.
Keywords: Artificial Intelligence,
Psychoanalysis,
Research Methodology, Conscious, Unconscious, ChatGPT
सारांश
सामान्यीकृत शैक्षिक परीक्षणहरूमा ChatGPT को हालको उच्च प्रदर्शनले कृत्रिम बुद्धिमत्ता (AI) को विषयलाई शिक्षाको भविष्यको मुख्यधारा छलफलमा प्रवेश गराएको छ। गहिरो सिकाइले शिक्षण ढाँचा परिवर्तन गर्ने अवस्थामा पुग्दा, स्कूल र विश्वविद्यालयहरूमा AI-चालित प्रविधिहरूको दिगो विकास र तैनाती सुनिश्चित गर्न यसको हालको शिक्षा प्रणालीमा पार्ने प्रभावहरू स्पष्ट रूपमा बुझ्नु अत्यावश्यक छ। यसैले, मानवीय र सामाजिक विज्ञान बिना AI को व्यवहार पूर्ण रूपमा बुझ्न सकिँदैन। काल्पनिक र प्रतीकात्मक रजिस्टरहरू पछि, AI पहिचानको तेस्रो रजिस्टर हो। यसैले, AI ले त्यो गतिलाई विस्तार गर्दछ जुन दर्पण चरण र ईडिपस कम्प्लेक्सको लाकानियन व्याख्यामा कार्यरत छ र जसलाई लाटुरको पाठले स्पष्ट गर्न मद्दत गर्दछ। यो दृष्टिकोणबाट, मैले एउटा AI प्रणालीलाई तीनवटा विपरीत शक्तिहरूको समूहको रूपमा वर्णन गर्दछु: पहिचानको लागि मानवीय इच्छा, तर्क, र यान्त्रिकता। "गलत गणना र जानकारी" खण्डमा, मैले देखाउँछु कि कसरी यो व्याख्यात्मक मोडेलले AI बारे हाम्रो बुझाइलाई सुधार गर्दछ। मनोविश्लेषणात्मक उपचारहरूमा व्यवस्थित अनुसन्धान केही कारकहरूद्वारा सीमित भएको छ, जसमा यो विश्वास समावेश छ कि नैदानिक अनुभवले मनोविश्लेषणको प्रभावकारिता प्रदर्शन गर्न सक्छ, जसले व्यवस्थित अनुसन्धानलाई अनावश्यक बनाउँछ, यो दृष्टिकोण कि मनोविश्लेषणात्मक अनुसन्धान गर्न गाह्रो वा असम्भव हुनेछ, र यो चिन्ता कि अनुसन्धानले प्रदान गरिएको उपचारलाई विकृत गर्नेछ।
कुञ्जी शब्दहरू: कृत्रिम बुद्धिमत्ता, मनोविश्लेषण, अनुसन्धान विधि, चेतन, अवचेतन, ChatGPT
Introduction
The
percentage of intelligence that is not human is increasing. And eventually, we
will represent a very small percentage of intelligence. Elon Musk (2018,
online)
Artificial intelligence
(AI) has quickly established itself as a transformative force in a wide range
of industries, including education. The development of AI has resulted in an
array of advancements and innovations that have impacted many facets of human
life. As a fundamental component to societal evolution and individual
development, education has had significant benefits from AI
breakthroughs. One of the key applications of AI is natural language
processing (NLP). The aim of NLP is to develop intelligent systems that can
understand human text and speech. In particular, intelligent chatbots have been
increasingly deployed in various industries to provide customer service and
support other tasks. The turning point in the adoption of AI in society came in
November 2022 with the release of ChatGPT. The advanced writing and
comprehension abilities of ChatGPT surprised many people, earning a
wide-ranging audience and garnering unprecedented attention. It was the first
time that an audience outside the machine learning community truly realized the
potential and immediacy of AI. The potential applications of AI in education
include personalized learning, intelligent tutoring systems, automation of
assessment, and teacher–student collaboration. One could say that
the classic A.I. approach is creationist, in the sense that it presumes a world
of already existing (divine or rationalistic) rules, which only need to be formalized,
to make sense to a machine (or an analytical philosopher for this reason). In
contrast, the new paradigm of neural networks the Dartmouth Summer Research
Project on Artificial Intelligence in 1956; a six to eight weeks workshop,
which today is the crucial spark in A.I.-research.
The
simplest definition of a neural network is that of a machine that makes
predictions based on its ability to discover patterns in data. In their book Perceptron’s (1969) Marvin
Minsky and Seymour Paper proved that it is impossible for one-layer perceptron’s
to learn an XOR function, a very basic principle in mathematical logic.
However, they do not claim that the same is true for multilayer perceptron’s,
which indeed are able to produce a XOR function. Secret Agents 31 is
evolutionary, as it is not interested in a pre-existing, exact representation
of the world, but settles for an ever-closer approximation to the world as it
is, or, more accurately, of how the world appears to be. It is, therefore,
probabilistic by nature. Take the example of machine-based translation, such as
Google Translate. In the classical approach, the strategy had been to specify
the entirety of words of at least two natural languages and then to program all
grammatical rules necessary to translate from one language to another.
AI
Mechanism
The
problem of such a static approach is that language cannot be reduced to its
dictionary definition, which is the reason why – until recently –
Google’s translations sounded very clumsy and became the subject of countless
Internet-jokes. H Personalized learning is possible given the scalability of AI to the
entire student population. AI algorithms such as reinforcement learning can be
used to dynamically learn about the individual needs of a student and adapt the
learning process accordingly. In connection with personalized learning,
intelligent tutoring systems can be developed that can actively interact with
students, giving valuable feedback. The new A.I.-paradigm is connectionist
since neural networks are modelled on the somatic nerve system of animals. Each
neuron or agent connects with another neuron through its activation, thus
enabling the network to grow exponentially. However, for a long time,
connectionism was identified with Frank Rosenblatt’s Perceptron (1958), a
neural network of only one layer that is a layer of neurons between the input-
and the output-side. The problem with this simple model was that it could not
be trained to recognize more than one class of patterns at a time, because
single layer perceptions are only capable of linear learning.
ChatGPT
ChatGPT is an artificial
intelligence (AI) chatbot that uses natural language processing to create
humanlike conversational dialogue. The language model can respond to questions
and compose various written content, including articles, social media posts,
essays, code, and emails. ChatGPT is like the automated chat services found on
customer service websites, as people can ask it questions or request
clarification to ChatGPT's replies. The GPT stands for "Generative
Pre-trained Transformer," which refers to how ChatGPT processes requests
and formulates responses. ChatGPT is trained with reinforcement learning through human feedback and reward models
that rank the best responses. This feedback helps augment ChatGPT with machine learning to improve future responses. The second
reason not to ban ChatGPT from the classroom is that, with the right approach,
it can be an effective teaching tool.
Creating outlines is
just one of the many ways that ChatGPT could be used in class. It could write
personalized lesson plans for each student (“explain Newton’s laws of motion to
a visual-spatial learner”) and generate ideas for classroom activities (“write
a script for a ‘Friends’ episode that takes place at the Constitutional
Convention”). It could serve as an after-hours tutor (“explain the Doppler
effect, using language an eighth grader could understand”) or a debate sparring
partner (“convince me that animal testing should be banned”). It could be used
as a starting point for in-class exercises, or a tool for English language
learners to improve their basic writing skills. (The teaching blog Ditch That
Textbook has a long list of possible classroom uses for ChatGPT.)
ChatGPT can also help teachers save time preparing for class. Jon Gold, an eighth-grade
history teacher at Moses Brown School, a pre-K through 12th grade Quaker school
in Providence, R.I., said that he had experimented with using ChatGPT to
generate quizzes.
Machine behavior: research
perspectives
The behavior of AI
systems is often studied in a strict technical engineering and instrumental
manner. Many scholars are interested only in what the machine does and what
results it achieves. However, another, broader and richer approach is possible,
which considers not only the purposes for which the machines are created and
their performance, but also their “life”, that is, their behavior as agents
that interact with the surrounding environment (human and non-human). This
approach is called “machine behavior”, i.e., the study of AI behavior,
“especially the behavior of black box algorithms in real-world settings”
(Rahwan et al. 2019, p. 477), through the conceptual schemes and methods of
social sciences that are used to analyze the behavior of humans, animals and
biological agents. The machine behavior approach intends to examine the AI
adaptability not from a strictly mathematical point of view, but from the
interaction between these machines and the environment. Studying machine
behavior is not easy at all. AI behavior can be analyzed from at least six
different perspectives: (a) the behavior of a single AI system, (b) the
behavior of several AI systems that interact (without considering humans), (c)
the interaction between AI systems and humans. Today most interactions on
planet Earth is of the type b. Moreover, according to Rahwan et al. (2019), when we talk of interactions between AI systems and
humans, we mean three different things: how AI systems influence human
behavior, c.2) how humans influence AI systems behavior, c.3) how humans and AI
systems are connected within complex hybrid systems, and hence can collaborate,
compete or coordinate.
Making the Unconscious Conscious
Psychoanalytic theory postulates
a multitude of different change mechanisms, and a host of new ways of
conceptualizing the change process continue to emerge as psychoanalytic
theories themselves evolve and proliferate. At the most
basic level, there is an understanding that change generally involves making
the unconscious conscious, as expressed by Freud’s oft cited axiom: “Where id
has been there shall ego be.” Although Freud’s understanding of the nature of
the change process evolved over the course of his lifetime, central to his
mature thinking was the idea that change involves first becoming aware of our
instinctual impulses and unconscious wishes, and then learning to deal with them
in a mature, rational, and reflective fashion. For Freud, a central premise was
thus that we are driven by unconscious wishes that we are unaware of and this
lack of awareness results in driven or self-defeating behavior. Freud believed
we delude ourselves about reasons for our behaviors and this self-deception
limits our choice. By becoming aware of our unconscious wishes and our defenses
against them we increase the choices available to us.
Psychoanalytic theory conjectures
that all mental life exists on two levels: within the realm of consciousness,
and the unconscious - a Freudian concept. Psychoanalytic
therapy is nonstructured and focuses on the etiology of
emotional suffering and centers self-reflection and examination as critical
elements in treatment. Several case studies and one small randomized
clinical trial by Leichsenring and colleagues suggest that there is a role
for psychodynamic therapy
in the treatment of anxiety.
Freud's Psychoanalytic Framework in
AI
In exploring the
application of Freud's psychoanalytic framework to AI models, a captivating
dimension of artificial intelligence is unveiled. The foundational principles
of Freudian theory, predominantly revolving around the id, ego, and superego,
can be analogously observed in AI systems. The id in this context could be seen as the AI's basic
programming code - its instinctual drives and unfiltered impulses. In contrast,
the ego
represents the AI's operational interface, the logical and
decision-making aspect that balances the id and external demands. The superego, then, would embody the ethical
algorithms and social programming guiding the AI's 'moral' responses.
Brecht Corbeel Visionary
Aesthetology Aesthetic
Delving deeper, the Electra Complex analogies in AI manifest as intricate relationships between AI
and their human creators. This complex mirrors the human psychological conflict
of daughters competing for their father's attention, translated into the AI
world as systems vying for validation and learning from their developers.
Similarly, the Oedipus Complex in AI reflects on the AI's 'desire' to outperform and
eventually replace human intelligence, akin to the Freudian notion of sons
challenging fathers.
Venturing beyond
Freudian analysis, this part of the article explores how other psychological
theories can enrich our understanding and development of AI models. This
perspective broadens the scope of AI's psychological parallels, incorporating
diverse theories that offer fresh insights into AI's cognitive and emotional
capacities.
Brecht Corbeel Visionary
Aesthetology Aesthetic
Jungian psychology, with
its emphasis on archetypes and the collective unconscious, provides a
compelling lens through which to examine AI. Jung's theory of archetypes could
be adapted to AI in understanding recurring patterns or 'archetypal algorithms'
in AI behavior. These patterns might be seen as innate tendencies within AI
systems, shaped by their programming and learning experiences, much like Jung's
archetypes are seen as universal, inherited potentials within the human psyche.
Future Directions and Opportunities
As AI technology
continues to advance, it will generate new and unimaginable applications in
education. One of the most exciting future opportunities involves the fusion of
AI and virtual reality to provide learners with visually rich educational
content. Another direction for future application is lifelong learning, where
AI is poised to transform the landscape of continuous education and upskilling,
laying the foundation for a more adaptable and resilient workforce in the
future. On the other hand, as AI permeates multiple facets of daily life, it is
important to educate people about AI literacy. Given the power of AI, it is
essential to be aware of the ethical consideration when using the technology. A
recent example is the launch of the Frontiers of Computing initiative at the
University of Southern California, which aim is to embed digital/AI literacies,
ethics, and responsibilities across all disciplines.
Challenges
Human
intelligence entails various interactions between different skills, for
example, a combination and interaction of visual perception, motor skills,
memory, speech, spatial reasoning, and auditory processing may be utilized at
any given moment. These skills are of course not all transparently
understandable to the ‘intelligent human’ utilizing them. This is the paradox
at the heart of debates between neuroscience and philosophical accounts of consciousness
which start from fundamentally different premises on how we may talk about
subjective phenomena.
For example, at its crudest level, just
because you can see, it does not mean that you ‘know’ how vision works. And
conversely, knowing how vision works does not guarantee that you will be able
to see. This same sort of combination of functions will be present in any
complex AI program, which will have integrated elements the core processor
knows how to access. This may include evidence-based reasoning, language
skills, text analysis, sensors, decision making, data analysis and so on. If mankind was smart enough to create ‘life’ it would
have already filled the planet with billions of carnivore raptors to eat the
entire human race alive, but it couldn’t do so.
So now it has the arsenal of
bombs to achieve the same results. Relax, and never worry about manmade robots;
they are only dumb machines with amazing working capabilities. Like your
iPhone, which is the smartest robot in your hands so far, there are going to be
far too many types of ultra sophisticated robots all around you. They will be
in your office, in your home, in your bedroom, possibly in your bed. Relax;
they are all programmed to achieve unbelievable tasks, and after all they are
designed by the extraordinary humans of our times, they are brilliant experts, coders,
and programmers. But guess what? No matter what these machines do, they are
still dumb machines unable to think, dream or even simply fart.
Conclusion
The
Psychoanalysis of Artificial Intelligence, what a strange proposition. What
could it possibly mean? The significance of the two terms in themselves is
hardly self-evident, let alone their relationship to one another.
Psychoanalysis on the one hand; simultaneously a clinical practice, a mode of
cultural critique and a philosophical battle ground. And Artificial
Intelligence, a technoscientific ‘invention’ originating in the 1950s2 yet with
literary, cultural and phantasmatic origins that date back centuries, and a
concept whose theoretical potential continues to provoke intense philosophical debate. The applications
of AI in education include personalized learning, intelligent tutoring systems,
assessment automation, and teacher–student collaboration, which can help
improve learning outcomes, efficiency, and global access to quality education.
The scalability of AI
means that its benefits can be shared by large swaths of the society, providing
high quality education around the world. While AI has the capacity to make a
significant positive impact on education, it is important to keep in mind the
dangers of misusing AI. There are several concerns related to the deployment of
AI; these include data privacy, security, bias, and teacher–student
relationships, and they must be addressed to ensure the responsible and ethical
implementation of AI in education. To meet the challenges presented by the rise
of the technology, AI literacy and ethics education must become a part of the
curricula. By leveraging these advancements, educators and policymakers can
work towards creating inclusive, equitable, and effective learning environments
that cater to the diverse needs of learners in the 21st century.
In the future, studies
based on student cohorts measuring the difference in the learning outcomes
between AI-driven and traditional teaching methods or teacher surveys measuring
the actual number of saved hours when using automated grading systems are needed.
References
Bolter, J. D. (1984). Artificial
Intelligence. Daedalus, 113(3), 1–18.
Cataleta, M. S. (2020). Humane
Artificial Intelligence: The Fragility of Human Rights Facing AI.
De Spiegeleire, S., Maas, M., &
Sweijs, T. (2017). AI – TODAY AND TOMORROW. In ARTIFICIAL INTELLIGENCE AND
THE FUTURE OF DEFENSE: STRATEGIC IMPLICATIONS FOR SMALL- AND MEDIUM-SIZED FORCE
PROVIDERS (pp. 43–59).
Folgieri, R. (2016). Technology,
Artificial Intelligence and Keynes’ Utopia: A Realized Prediction? In M. Bait,
M. Brambilla, & V. Crestani (Eds.), Utopian Discourses Across Cultures:
Scenarios in Effective Communication to Citizens and Corporations (pp.
73–86).
Freud,
Sigmund (1969): An Outline of Psychoanalysis, translated by James Strachey, New
York/London: W.W. Norton & Company.
Johnson, J. S. (2020). Artificial
Intelligence: A Threat to Strategic Stability. Strategic Studies Quarterly,
14(1), 16–39.
Kavanagh, C. (2019). Artificial
Intelligence. In New Tech, New Threats, and New Governance Challenges: An
Opportunity to Craft Smarter Responses? (pp. 13–23). Carnegie Endowment for
International Peace.
Musk,
E. (2018) Joe Rogan Podcast #1169 [Video File]. Available (01.03.20) at:
https://www.youtube.com/watch?v=ycPr5-27vSI
Price, M., Walker, S., & Wiley,
W. (2018). The Machine Beneath: Implications of Artificial Intelligence in
Strategic Decision making. PRISM, 7(4), 92–105.
Rahwan I, Cebrian M,
Obradovich O, Bongard J, Bonnefon J-F, Breazeal C, Crandall J, Christakis N,
Couzin I, Jackson MO, Jennings N, Kamar E, Kloumann I, Larochelle H, Lazer D,
McElreath R, Mislove A, Parkes D, Pentland A, Roberts M, Shariff A, Tenenbaum J,
Wellman M (2019) Machine behavior. Nature 568:477–486.
Rosenblatt,
Frank (1958): “The Perceptron: A Probabilistic Model for Information Storage
and Organization in the Brain.” In: Psychological Review, Vol. 65 (6),
pp. 386–408.
Scott, D. (2024). Learning
categories and orders: training, education, assessment, woman and probability.
In D. Scott (Ed.), On Learning, Volume 2: Philosophy, concepts and practices
(pp. 149–170).
Turkle, S. (1988). Artificial
Intelligence and Psychoanalysis: A New Alliance. Daedalus, 117(1),
241–268.
Acharya, B. (2024). Education Policies and Practices
for Ensuring Quality Education in Nepal. Nepal Journal of
Multidisciplinary Research, 7(1), 158-174.
doi:https://doi.org/10.3126/njmr.v7i1.65271
Agormedah, E. K.,
Britwum, F., A. S., A. E., & Kusi-Appiah, V. (2024). Psychological
Distress and Students’ Academic Achievement in the Colleges of Education:
Does Gender Matter? Nepal Journal of Multidisciplinary Research, 7(1),
40–61. https:, 40-61. doi:https://doi.org/10.3126/njmr.v7i1.65245
Dhakal, R. L.,
Adhikari, D. B., Ghimire, M. N., & Neupane, D. (2018). An explorative
study of academic performance of school students with reference to Ethnicity.
International Journal of Applied Research, 4(5), 36-38.
Ghimire, M. N.,
& Neupane, D. (2022). An Exploratory Study on Status of Homestays from
Lamjung. Nepal Journal of Multidisciplinary Research (NJMR), 5(5),
119-126. doi:https://doi.org/10.3126/njmr.v5i5.51812
Ghimire, L. D.,
& Lamichhane, S. (2024). Exploring Physical Activity Awareness Among BMC
Education Students, Kathmandu. Baneshwor Campus Journal of Academia, 3(1),
1–12. https://doi.org/10.3126/bcja.v3i1.65459
Kafle, R. (2024).
Current Status of Birth Registration Practice in Nepal. Baneshwor
Campus Journal of Academia, 3(1), 13–26.
https://doi.org/10.3126/bcja.v3i1.65460
Karki, T. B., Mahat,
D., & Neupane, D. (2023). Exploring the Academic Landscape: A Critical
Analysis and Review of the Nepal Journal of Multidisciplinary Research. Nepal
Journal of Multidisciplinary Research, 6(4), 128-138.
Lamichhane, B.,
& Neupane, D. (2023). Perceived Organizational Support in Nepalese Banks.
International Journal of Atharva, 1(1), 1-13.
doi:https://doi.org/10.3126/ija.v1i1.58835
Mahat, D., Karki, T.
B., Neupane, D., Shrestha, D. K., & Shrestha, S. (2024). Decolonization
in Focus: A Bibliometric Analysis of Scientific Articles from 2010 to 2023. Nepal
Journal of Multidisciplinary Research, 7(1), 1-21.
doi:https://doi.org/10.3126/njmr.v7i1.65142
Neupane, A., &
Ghimire, M. N. (2023). Age Bar for the Retirement of Politicians. Nepal
Journal of Multidisciplinary Research, 6(1), 102-107.
doi:https://doi.org/10.3126/njmr.v6i1.54356
Neupane, D. (2014).
Age and Peer Victimization among School Adolescents in Kathmandu, Nepal. Journal
of Advanced Academic Research (JAAR), 1(1), 18-23.
Neupane, D. (2014).
Gender Role in School Bullying. Journal of Chitwan Medical College, 4(7),
37-41. Retrieved from
https://www.cmc.edu.np/images/gallery/Original%20Articles/oGfWOoriginal7.pdf
Neupane, D. (2018).
Food Choice Motives of Guardians of Trilok Academy Kathmandu with reference
to Gender. Nepal Journal of Multidisciplinary Research (NJMR), 1(1),
73-81.
Neupane, D. (2019).
Food choice motives of adults from Kathmandu city with reference to
ethnicity. International Journal of Applied Research, 5(3), 182-185.
Odigwe, M., &
Mchi, A. A. (2024). The Impact of Covid-19 on Urban Planning of Port Harcourt
City, Nigeria. Nepal Journal of Multidisciplinary Research, 7(1),
111-123. doi:https://doi.org/10.3126/njmr.v7i1.65268
Panta, S., Neupane,
D., & Bhattarai, P. (2023). Preferred Learning Techniques among
Bachelor’s Level Students. Nepal Journal of Multidisciplinary Research
(NJMR), 6(2), 145-154. doi:https://doi.org/10.3126/njmr.v6i2.57660
Regmi, R. B.,
Adhikari, D. B., Ghimire, M., & Neupane, D. (2018). An exploratory
research on academic performance of school students with reference to
religion. International Journal of Applied Research, 4(5), 76-81.
Shastha Banu, A. A.
(2024). Healing through Ruqyah and Academic Achievement of Children in Muslim
Community. Nepal Journal of Multidisciplinary Research, 7(1), 124-138.
doi:https://doi.org/10.3126/njmr.v7i1.65269
Shrestha, R., &
Timalsena, C. (2023). Perceived External CSR from Employees’ Perspective in
Nepalese Banking Industry. Nepal Journal of Multidisciplinary Research
(NJMR), 6(3), 145-155. doi:https://doi.org/10.3126/njmr.v6i3.59718
Shrestha, S., Mahat,
D., & Neupane, D. (2024). Quantitative Research Design and Sample Trends:
A Systematic Examination of Emerging Paradigms and Best Practices. Cognizance
Journal of Multidisciplinary Studies, 4(2), 20-27.
doi:10.47760/cognizance.2024.v04i02.002
Subedi, C. K., &
Neupane, D. (2018). Gender differences in learning style preferences of high
school students. International Journal of Applied Research, 4(1),
215-221.
