Abstract

This paper explores the transformative role of Artificial Intelligence (AI) in reshaping intellectual paradigms, highlighting a shift from traditional knowledge systems to dynamic, adaptive frameworks. While current AI technologies are anchored in historical data, their limitations in addressing future intellectual challenges necessitate a more innovative approach. The concept of "Alternative Intelligence" is introduced as a new paradigm that transcends traditional AI. This model fosters creativity, adaptability, and novel thought processes, enabling transformative intellectual synergies between machines and people. Using language education as a reference point, the study illustrates how tools like dynamic, AI-powered books can personalize learning, foster collaboration, and redefine traditional educational structures. The paper advocates for integrating human and machine capabilities into a synergistic intellectual ecology to address the widening gap between technological progress and human cognitive development. By embracing Alternative Intelligence, society can reimagine and redefine knowledge creation for a rapidly evolving future and attempt to reduce the gap between technological and individual development. Keywords: Artificial Intelligence, Alternative Intelligence, Intellectual transformation

I INTRODUCTION

This reflective paper explores the potential impacts of the exponential growth of Artificial Intelligence (AI) on the global intellectual landscape, using education as a point of reference — particularly language education, the author’s area of expertise. Education offers a unique perspective as it is central to human development, cultural transmission and the creation of shared intellectual frameworks. Unlike other sectors where the impact of AI tends to be limited, education influences fundamental skills such as critical thinking and creativity and impacts all human activities, hence its use as a reference point.

As AI becomes integrated into education systems, it will contribute not only to change teaching and learning systems, but also challenge traditional beliefs and approaches to knowledge dissemination, collaboration and even authorship of documents and resources. Thus, focusing on the relationship between education and AI can help us better understand AI’s potential to reshape and nurture the development and evolution of human intellect.

On November 30, 2022, life on Earth changed forever as ChatGPT was released at no cost to the general public by OpenAI. Within five days it had amassed a following of one million users, the fastest adoption of any product in the history of the planet. The introduction of AI, particularly tools like ChatGPT and other Large Language Models (LLMs) in general, marks a profound shift in how we approach life and education. While much attention has been given to the LLMs’ ability to create stunning “things”, whether texts or other resources, educational or otherwise, their broader transformative implications deserve closer attention. Take, for example, the transformation of the most common educational tool, the book. This will serve as the starting point for our discussion.

Traditional books made of pages, chapters, and fixed text, with a set starting point and ending point (either on paper or virtual) are static repositories of knowledge. They offer no adaptability, presenting content in set formats that either suit the reader or do not. They cannot be changed and cannot adapt to circumstances, although the reader may be able to adapt to them. Essentially, the book is immutable; what it is, it remains, regardless of the reader's actions.

In contrast, empowered by AI, we can now envision what we will call the "dynamic book." This transforms the traditional book concept into an adaptable, interactive learning tool tailored to individual needs, empowering both learners and educators. Starting conceptually with a prompt to create a digital version of an old-fashioned book complete with chapters and pages, this soon evolves into something much more dynamic, hence its name.

For the purpose of our discussion, here are some of the features a dynamic book might have. Dynamic books will enable personalized learning, where, for example, a learner of medical English could access vocabulary simplified to their proficiency level, such as CEFR B1[1] or A2, while advanced learners could engage with more complicated text at CEFR C1 level, providing seamless customization to learners’ abilities. Dynamic books could also offer a range of explanations, where concepts could be simplified or deepened according to the learner's understanding, and could adapt to individual cognitive levels. Additionally, interactive feedback could be included, where learners could reconfigure, content based on their preferences, fostering active engagement. They could also participate in dialogues, and (quasi-)authentic conversations directly with the materials presented.

Dynamic books could integrate multimodal engagement, allowing learners to listen to content, access concise summaries, or initiate discussions of varying length and complexity with the book itself to deepen comprehension. They could also provide language versatility with instant translations, enabling learners to navigate content in their preferred language.

Beyond these features, dynamic books could foster experiential learning through immersive technologies. For instance, history learners might explore ancient civilizations using virtual reality (VR), while language learners could practice communicating in realistic, tailored scenarios. If necessary, content regeneration would allow entire texts or books to be restructured to match specific interests, such as focusing a French textbook on environmental conservation.

Under these circumstances, a ‘book’ is no longer a ‘book’ but something more like an Interactive Learning Environment. It may have started out as something looking like and organized like a traditional book (though digitized) but quickly changed.

II CHALLENGES OF ADAPTABILITY

While the potential of dynamic books is transformative, they introduce complexities that challenge academic structures and create genuine obstacles in the real world. For instance, consider the issue of authorship ambiguity. Imagine a teacher, an AI system, and a group of learners co-creating a dynamic book. This can lead to serious disagreement regarding who should be recognized as the author. If a learner prompts an LLM with something like, "Write me a book on quantum physics," determining whether the teacher takes precedence, or if the AI should be credited as an author or even as a tool, or if learners should be recognized as contributors becomes complex. The dynamic book is actually an ever-changing co-construction resulting from a partnership between learners, teachers, authors, developers, and perhaps others, highlighting a problem not limited to book authorship but extending to myriad other creations.

A related issue is academic recognition. Currently, promotion and tenure in academia often hinge on contributions like book authorship. In a world of collaboratively created, constantly evolving dynamic resources, how will academic structures recognize and reward such contributions? This issue could dramatically impact on how research is conducted, as academia typically encourages collaboration but then, paradoxically, fights over first authorship, with universities often using authorship order to gauge the value or proportion of contributions to a publication. While the example provided here is embedded in the academic world, similar difficulties appear in countless other contexts.

Finally, there is resistance to change at a very practical implementation level. Teachers and institutions, steeped in traditional models of teaching, might resist adopting such disruptive technologies, despite their benefits. Although the adoption of such systems is clearly inevitable in the long run, the resistance experienced today poses a significant hurdle to the integration of dynamic books and similar resources into educational frameworks.

III TOWARD A LEARNER-CENTERED FUTURE

The dynamic book exemplifies the shift toward learner-centric education, enabling personalized, multimodal, and immersive learning experiences and mirrors societal trends toward personalization brought about by the availability of search tools such as Google, ChatGPT and resources such as YouTube. These tools can and do provide personal solutions when interrogated. The appearance of LLMs magnifies that potential and creates a generation of do-it-yourself, independent researchers [2], [3]. In turn, this has given rise to concerns that people will no longer remember things but, instead, will just know where to find the answers that they are looking for. This has become known as the Google Effect [4].

These developments not only enhance learning outcomes and human intellectual growth but also force a rethink of educational structures and accompanying social structures to accommodate collaborative and adaptive technologies. The dynamic book used here as an example is not just a tool for delivering content. It symbolizes a redefinition of the learning process itself, tailored to the unique needs and preferences of individual learners. This sets the stage for systemic transformation, where education evolves to meet the demands of a rapidly changing world and society. It also demands a realignment of thought that may be difficult to implement given that, intellectually we appear to be lagging behind the technology.

IV THE LIMITS OF CURRENT EDUCATIONAL MODELS

Most current educational systems are rooted in the assembly-line logic of the Industrial Revolution, which standardizes learning processes and outcomes. These models assume the production of graduates with similar knowledge and skills, designed to fit predefined roles. This approach has persisted even as technology advances, with AI tools often used to support existing paradigms rather than to challenge or transform them especially as the job market is also based on assembly line protocols.

For example, the application of AI in education often remains limited to static curriculum design. AI tools like ChatGPT have been increasingly employed to enhance traditional teaching, such as generating exercises or quizzes that fit within the kinds of syllabi that are currently at the heart of most educational systems. However, these applications fail to leverage the full transformative potential of AI, such as dynamically adapting the curriculum to individual learners' evolving interests, skill levels, or moment-to-moment needs [3]. Additionally, assessment conformity remains a significant issue, where education still predominantly evaluates students on uniform metrics like standardized tests, which prioritize rote memorization over critical thinking or creativity. Even when AI is able to generate personalized assessments, its potential is largely untapped due to institutional inertia, lack of time to develop new systems, and reluctance to engage in large-scale changes.

As we engage in this discussion, a significant theme emerging is the difference between renovation and innovation. Renovation involves using technology to improve existing systems without questioning their foundational assumptions. For instance, AI might streamline administrative tasks or enhance lesson delivery though people and organizations still operate within traditional structures and frameworks. Innovation, on the other hand, demands a fundamental rethinking of educational goals and structures. It involves asking questions such as: Should education, inter alia, continue to prioritize standardization, or should it focus on cultivating diverse intellectual ecologies tailored to individual learners?

Incorporating dynamic books and AI tools into education calls for innovation, not renovation, and that could have far-reaching effects. For example, a language course could abandon standardized tests in favor of AI-driven, real-world simulations that assess learners' conversational proficiency in various cultural contexts.

V AN INTELLECTUAL PARADIGM SHIFT

Like it or not, the integration of AI into education will inevitably, but perhaps not immediately, drive us toward a change in intellectual paradigms, most likely emphasizing diversity, flexibility, creativity, and continuous lifelong learning within both formal and informal settings. Surprisingly, these changes may not be led by education professionals but by the general public who as we have seen, have become accustomed to “individualized solutions” for “individualized” problems.

Two key principles illustrate this shift:

The shift from uniformity to individuality. AI enables the creation of personalized learning pathways. A simple example might be an English-speaking learner looking for listening practice to understand French better while others may be looking at pronunciation issues to enable them to complete a communicative task. Such flexibility requires educators to embrace variability in both student outcomes and student learning patterns, possibly recognizing diverse forms of achievement rather than conforming to standardized benchmarks such as the PISA test [5].
The strengthening of intellectual development. To keep pace with AI-driven advancements, education must emphasize skills that machines cannot replicate such as creativity, critical thinking and ethical reasoning. For instance, instead of teaching students to memorize sterile vocabulary lists, educators could use AI to illustrate and reinforce vocabulary usage in context, fostering deeper intellectual engagement even using Virtual Reality (VR) to enable students to participate in communicative events through immersion, e.g. Kaplan-Rakowski [6]. These changes will lead inevitably encourage to learner autonomy. People will be doing it for themselves. The shift from teacher-led instruction to learner agency will undoubtedly challenge traditional power dynamics in education, disrupting personal and institutional comfort zones, rhythms and principles. This effect is all-pervasive and not limited to the educational field.

VI RESISTANCE TO CHANGE

Despite clear benefits, adopting AI-driven innovations often meets resistance. Educators, policymakers, and institutions may hesitate due to fear of the unknown, where teachers accustomed to traditional methods might feel unprepared to navigate adaptive, AI-powered tools. Intellectual and institutional rigidity as well as suspicion also play a role; schools and universities, bound by traditional accreditation standards and administrative protocols, may struggle to accommodate flexible, technology-driven approaches, though with time, these standards and protocols will certainly change. More importantly. the shift from teacher-led instruction to learner-driven self-direction challenges traditional power dynamics in education, disrupting personal and institutional comfort zones, rhythms, and principles. The same is true of other areas in life.

VII CONNECTING THE THREADS

These challenges underscore the need for systemic change, linking back to the dynamic book’s potential for personalization and adaptability. The same flexibility that enables learners to regenerate books or request VR experiences must inform the broader educational ecosystem. Under these circumstances, education should evolve into a dynamic, fluid system where AI serves as a catalyst for intellectual growth, not just a tool for convenience.

By transitioning from renovation to innovation, educators and institutions can harness AI to create more inclusive, learner-centered environments. The dynamic book is only one example of the transformative potential of such technologies; its success depends on a willingness to rethink entrenched systems and embrace the possibilities of a new intellectual paradigm.

As we study the transformative potential of AI, it becomes evident that embracing its capabilities requires addressing not just systemic inertia but also foundational intellectual challenges. AI is influencing the interplay between technological progress and human intellectual development. In so doing, it highlights the urgency of rethinking our practices in our increasingly AI-centric world.

VIII THE EXPONENTIAL GROWTH OF TECHNOLOGICAL POWER

Two pivotal laws—Moore's Law and Huang's Law—illustrate the rapid acceleration of technological capabilities.

Moore's Law (Gordon Moore, co-founder of Intel Corporation) predicts that the number of transistors on a microchip doubles approximately every two years, leading to exponential increases in computing power.
Huang's Law (Jensen Huang, co-founder of Nvidia Corporation) highlights similar growth in GPU (Graphics Processing Unit) performance that powers AI at the moment, driven by advancements in specialized hardware and software optimization.

While Moore's Law offers predictable growth, Huang's Law introduces an unpredictable, step-like pattern characterized by breakthroughs and sudden accelerations or slowdowns.

These principles have serious implications, including a technological lag in education. While technology evolves exponentially, educational and other systems often lag behind: they adapt but only at a linear pace. Just as an example, while AI tools like ChatGPT can process complex datasets in seconds, many classrooms still rely on static textbooks and traditional teaching methods.

IX THE NECESSITY FOR INTELLECTUAL GROWTH

To bridge this gap, educators and learners must cultivate intellectual agility and enhance their cognitive capabilities in an attempt to keep pace with the rapidly evolving technology. Choosing not to engage in this effort is not an option. It will inevitably lead to a widening gap, leaving human beings increasingly behind in our technology-driven world.

The following will illustrate. The curves represent Moore’s Law, combined Moore’s Law and Huang’s Law together, with Dynamic General Intellectual Growth (normal intellectual growth) and Dynamic Accelerated Intellectual GrowthThe above curves clearly demonstrate the speed of the technology-based developments, especially when combined together with the relative sluggishness of “natural” intellectual growth. The discrepancy is alarming. However, there is significant projected improvement in the development of human intellect once action is taken to blend it with technology in new and unpredictable ways.

Continuing our exemplification of the issues through language education, it seems that our only hope of narrowing the gap significantly is to work in a new intellectual ecology that combines AI and other novel intellectual and scientific developments. Again, that will undoubtedly mean changing our paradigms. As we grapple with that concept, we will be obliged to abandon old habits and practices and replace them with new, currently unpredictable, ways of thinking and doing things at an ever-increasing rate. And new ways of thinking are a quasi-inevitable outcome of working with new technology as we suddenly realize, with justifiable excitement, that we can now do what had previously been undoable or even unthinkable even though some of the ideas may have previously been generated several decades in the past. For instance, Lian’s dialog generators [7] and other generative systems. This new freedom applies to both teachers and learners (whether taught or not) and, necessarily, covers assessment and evaluation too. It also means that the days of complacency, of “resting on our laurels” are gone forever. The days of stopping to catch our breath will no longer exist. Yet, in order to give us a chance to catch up we may need to impose such moments perhaps even by temporarily “pausing progress”. But even if we have that intention, will it in fact be possible to do so given the energy and excitement provided by AI developments in addition to the rate of technological change?

So, the only logical and realistic way forward for all of us is actually to embrace the technology. Yet, quite often, instead of encouraging technology use to assist with our (learning) problems we often punish it. In the context of English as a Foreign/Second language, for instance, both students and researchers are penalised for relying on LLMs to help them write papers. We claim that it is unethical, we say it is cheating. We expect all knowledge and all text to emerge entirely from within the minds of individuals. But the truth is that academic text has never been produced entirely or exclusively by individuals. Education systems have a tradition of providing significant assistance to both native and non-native speakers of a language through writing centres and academic support centres where well-meaning staff seek to train students to conform with academic standards and write the way that academics are supposed to write. Alternatively, students are sent to writing “workshops”. Or they are obliged to attend semester-long courses where writing scholars initiate them to the mysteries of moves and turns. If that does not work, they may be sent to various online webinars to listen to eminent writing experts explain how good academic text is written.

Unfortunately, most students and researchers do not become truly independent, expert writers. Some do, but the majority still experience serious problems and those who wish to publish their texts, even significant experts, still need to rely on external support such as native speaker expert writers especially if they are obliged to write in a language which is not theirs (e.g. English). Most are simply not self-sufficient.

So, essentially and to put it bluntly, we expect people to copy or otherwise reproduce the texts of others through the use of formulaic expressions and standard discourse structures. In the end, no academic writer is particularly original discursively. And we send students to such places as writing centres precisely to be able to reproduce these formulaic and other structures. We want them to comply with our notion of “good writing”, to reproduce/copy it and we reward those who appear to do it well.

The arrival of ChatGPT has meant that everybody is now able to produce “good writing” and write correct English. This means that readers of essays, assessors of dissertations and reviewers of academic articles need no longer strain to understand the previously defective texts of researchers who were not native speakers of the language they were writing in. At last, the issue of correct language and, perhaps, discourse structure has been potentially resolved and readers or assessors can now focus peacefully on the content of the text. Unfortunately, though, using LLMs is often construed as cheating – perhaps because they can be asked for more than just language and discourse advice. Strangely, though, asking for help from Grammarly is acceptable, as is the use of native speakers to check language – and it is also acceptable, and always has been acceptable, even encouraged, to consult subject matter experts (e.g. teachers, colleagues or team members) for advice.

So, the problem is not that people are getting help: they could always get help from a variety of sources, the problem is not that people have not become expert writers as most would never become expert writers anyway (especially in a foreign language), the problem is not that they could now produce perfect language, but the fact that help is so readily (and cheaply) available for BOTH writing AND content (especially) and that the effort to arrive at good writing is greatly reduced making life much easier. Somehow, the lack of effort is frowned upon: there does not seem to be enough “suffering” to merit the sought-after reward such as publication of an article or success in a course. Curiously, also, this problem seems to arise for a reason that has nothing to do with the creation of knowledge. It arises because in the context of assessment of texts or exams, or academic reviews, we are programmed to accept essentially the same response from all those who are writing, i.e. repetition of the dogma of the field. This, of course, is exactly what ChatGPT is programmed to provide and does provide to everybody on demand. ChatGPT will reproduce that dogma perfectly well. This is why ChatGPT can pass examinations set by reputable universities [8]. The university expects, on the other hand, the dogma but would call it something like “basic” or “essential” knowledge. In other words, students are actively expected to repeat the common knowledge of the field. If we change our form of assessment to require evidence of intelligent intervention rather than mere repetition of ‘facts’, then we will be happy with the intellectual content of the students’ response and, of course, experience a sense of relief at the quality of the language used. The responsibility for this change lies entirely with us.

While changing the form of testing/assessment is certain to be a major source of new work for the education world, it will also likely result in significant improvement in educational outcomes if implemented.

So, whereas we are currently going through a punishment phase for using technology, the likelihood is that in 10 years’ time, probably much sooner, technology use for both text and intellectual production will become normalized as we adjust to both features and the punishment will disappear (which makes one feel sorry for those being punished today). A similar scenario was true of calculator use before that was normalized in mathematics education. In contrast to calculators, however, the level of change resulting from current developments will be extraordinary as we will be forced to rethink our entire educational universe – especially conformity-based education.

In other words, we cannot keep thinking and doing the same things that we are doing now on the basis of today’s principles in the hope that all will be well. And we need to make that change as rapidly as possible.

X THE RISK OF FALLING BEHIND

Continuing the focus on education as a reflection of society in general, if we fail to adapt quickly, the gap between human intellectual capacity and technological power will continue to widen. Stopping this (or delaying it) will require decisive action as, while technological development is inevitable, developing constructive synergy between mind and machine is not. And traditional education will risk becoming irrelevant if it cannot keep pace with AI-driven advancements. For instance, preparing students for jobs that may no longer exist in a decade could certainly undermine the value of their education. It will also result in missed opportunities as the failure to leverage AI’s potential for personalization and innovation may result in failed opportunities to empower learners with the critically important intellectual skills required for the future.

To avoid these serious problems, institutions will need to embrace AI not as a threat to traditional education but as a catalyst for transformation, evolution and progress. By fostering intellectual agility and collaboration between humans and machines, education can evolve into a system that prepares learners for the challenges and opportunities of an AI-driven world. This shift is not merely desirable—it is imperative. The alternative is to fall further and further behind in a world that is accelerating beyond the capacities of traditional approaches.

The above changes are highly significant. But the extent of the change implied is much broader than we have seen so far, with fundamental intellectual changes being made in all aspects of life.

While, so far, we have illustrated with examples primarily drawn from the language education area, the impact of these changes will be felt everywhere. Research is one such area.

Historically, research has tended to be divided into two paradigms (some would argue inaccurately):

Quantitative Research that (apparently) prioritizes objective, group-focused, and statistically rigorous methods, favoring numerical analysis for broad applicability, and
Qualitative Research that focuses on subjective, individual-centered, and narrative-driven approaches, emphasizing depth and context over generalizability.

This is a much-cherished distinction with proponents of each field carefully guarding their territories even though the emergence of so-called mixed-methods approaches appears to combine both.

Nowadays, tools like ChatGPT can analyse large qualitative datasets (e.g., interviews, open-ended responses) and apply quantitative methodology to them such as quantifying themes, frequencies, and patterns. This new ability will bridge the divide between quantitative and qualitative, enabling researchers to apply statistical rigor to qualitative insights, while quantitative methods gain depth from qualitative contexts and a better focus on research questions.

This convergence undermines the historical tension between the two paradigms, encouraging collaboration, innovation and clearer, more valuable, research outcomes. At the same time, this new paradigm also democratizes methodologies across disciplines. And it is clear that, with time, the GPTs will get better at it.

This development also has transformative applications in all fields. As an example, again taken from language education, personalized Intervention through subgroup identification may become more readily available. Traditional group-level statistics often obscure individual variability and subgroup patterns. AI tools based on a combined quantitative and qualitative analysis will enable the identification of subtly different learner profiles. For example, a language class of 30 students might include subgroups with distinct needs – some struggling with anxiety, others needing a more collaborative environment, while still others may prefer solitude. Characteristics of this kind are often difficult to identify let alone address, especially on a large scale. The combined qualitative-quantitative AI-driven model of analysis would be able to deal with all issues possibly by recommending tailored interventions for each subgroup.

These targeted strategies, in turn lead to the adoption of such interventions as Precision Language Education [3], [9] or rhizomatic systems [8], [9], [10] where learning paths are optimized not only according to specific group and individual learning profiles but also according to learners’ perceptual and biological characteristics, addressing much more subtle features of their learner makeups than is currently available.

Although we have been talking about this new mixed model in the context of education, the model is applicable to all research contexts and all areas of endeavour. As a result, the entire world will change because the entire world depends on research. Despite possible protests, the currently enshrined qualitative and quantitative models of mixed research will inevitably fail over time and give way to a new hybrid model. And, in due course, this sequence of events will be repeated countless times in countless areas of life as we move on to the next phase of human intellectual development.

The ability of large language models (LLMs) to conduct convergent qualitative-quantitative research is an emergent property rather than a planned feature of their design. This capability arises from the inherent structure of the model and the statistical patterns it derives from its training data. The model itself may be new and exciting, but its components already exist and are embedded in the world as we know it. The AI suddenly makes the convergence possible and that is because the novel paradigm and enhanced capabilities observed above are due to human ingenuity in developing, interpreting, applying or further training the LLMs. This reflects the human ability to think beyond the limitations of current AI systems. These systems, despite extraordinary powers, cannot synthesize in original and optimal ways without external direction.

This is because, at its core, AI, especially LLMs, is designed to simulate known aspects of human intelligence rather than change it in truly novel ways. The current augmentation of human capabilities through AI is primarily based on improving speed, scope and accuracy rather than by creating fundamentally new intellectual capabilities. In other words, it mimics rather than create.

Current AI, with its quasi-omniscient data processing and rapid operational speed, significantly amplifies human ability to perform tasks, thereby saving time and cognitive effort. However, even seemingly original outputs, like novel strategies in games, are the result of rule-based activities programmed or learned from human sources.

This brings us to an interesting impasse. While we can now perform countless tasks better than before to the extent that we can make new discoveries, while we can use the synthetic and logical capabilities of LLMs to arrive at novel human-driven solutions (such as the dynamic book) or new paradigms (such as the convergent qualitative-quantitative research paradigm) we have not yet broken free from the limitations of our current intellectual systems and the limitations of AI itself (as a mirror of current human intelligence). This means that the disconnect between human intelligence and technological development will continue to grow.

Even if we accept the Dynamic Accelerated Intellectual Growth curve, which suggests that technological augmentation can boost intellectual development, this growth remains insufficient when compared with the rapid rate of technological progress.

In other words, the strategy of augmenting our intellectual capabilities with technology but following the traditional evolutionary path of intellectual development might not suffice. We need to go beyond that.

A simple example from the language world will illustrate. Below are two graphs. They show the results of national proficiency tests for English in ASEAN and other Asian countries for the last decade or a little more. One graph represents the results of the EF Proficiency Index test for English (proficiency band scores) whereas the other summarizes results for IELTS, TOEFL and TOEIC tests.

As can be seen, proficiency levels on a per country basis are flat-lining. This means that over the past decade or so, there has been no improvement in national proficiency scores despite significant efforts in terms of policy, expenditure and commitment to English language education during that time. A few countries like China and Viet Nam have even regressed from Moderate Proficiency to Low Proficiency (on EF EPI) while others such as Thailand were already at the lowest level but have not improved. This is highly disappointing for the countries concerned and makes one wonder as to why this is so.

The typical response to this issue from the academic community includes a contradiction between proficiency requirements and government testing requirements (training students for the test rather than proficiency), lack of teacher education including the ability to speak the language, lack of methodological training, oversize language classes and reluctance of the population to learn English. Somehow, it is largely the fault of governments, teachers, students and society in general. There is a sense in which the language teaching principles are fine to reviewers and observers, but badly applied e.g. the study by Phetsangkhad that barely mentions them while dealing at greater length with other issues[12]. In other words, the belief of the profession and the language education industry in the fundamental principles of its teaching and learning approaches, its own well-foundedness, is never questioned. As a general rule, it is not one of the variables examined when determining strengths or weaknesses. While it may be possible to improve current teaching and learning practices, the principles involved appear beyond reproach. We firmly believe that we know how to teach and failure is someone else’s fault. Taking a closer look at this issue for a moment in light of the proficiency tests applied above may be helpful in dealing with this issue.

The reality is that all those doing the tests listed above are volunteers. They do not have cultural or motivation problems. In fact, they are keen to improve their English and score well. They work hard and do not just follow government programs with their testing-focused assessment (if this actually applies in their country). They almost certainly supplement their formal classes with informal classes in private language centers or with private tutors and engage in other language-learning activities. In other words, they are hard-working, motivated, ambitious people who have experienced a wide range of teaching approaches and who are not subject to the claimed limitations of governments and other issues as outlined above. The only thing left to explain the failure of the programs is the one variable that is never questioned: how they learned or were taught English. Why is this not a concern? It is difficult to know the truth but there seems to be general satisfaction with teaching/learning approaches as there is no widespread move to replace them. This inevitably leads to a complacency mindset. Almost no one believes that we have reached a plateau in our understandings of how to teach languages and there is no widespread belief that a change is necessary. Yet the evidence speaks for itself. None of the failure factors mentioned above seem to apply to the above test-takers. Under those circumstances, methodology is the only remaining factor that can account for their lack of success.

So, what we need in order to move forward here and elsewhere in this new age is an alternative mindset. We need to break from the belief that all is well. This applies to the AI world too and translates into the need for a new form of artificial intelligence that diverges from traditional AI (Artificial Intelligence) and AI (Augmented Intelligence) by focusing on creating intellectual systems that do not merely replicate or enhance human cognitive processes but instead offer new, innovative ways of thinking and problem-solving. Let us call this new way Alternative Intelligence (also AI). Alternative Intelligence may one day be embedded in software but, for the moment at least, it is intrinsically embedded in human intellect. It is a way of questioning the intellectual status quo in order to move life forward. It is a way of breaking the vicious circle of endlessly repetitive backward-looking research, of substituting new ways of thinking for old and proposing (at times radically) different intellectual paradigms.

Such a new AI will have to be developed explicitly and deliberately as the old AI will necessarily carry with it all the limitations of real human intelligence and social practices, including its preferences and biases.

Alternative Intelligence (AI) will need to be fundamentally different, truly innovative, seeking to develop cognitive models that are not bound by the biases or limitations inherent in human cognition and the data used to train traditional AI models. In other words, it will be truly forward-looking. Unlike traditional AI, it would aim to predict, adapt, evolve and create the future rather than repeat past patterns. As much as possible, it would also be free from the biases and limitations inherent in human thought and society as reflected in the data of conventional AI systems. Finally, it would represent a paradigm shift in how we understand and implement intelligence in machines, potentially leading to solutions for problems that current AI methodologies are ill-equipped to handle. To achieve all of this, it would necessarily be iconoclastic in nature and map out changes in both artificial and human intelligence that are likely to irreversible.

To bring it to life, this new concept would necessarily bring to machine intelligence insights from philosophy, ethics, neuroscience, advanced computing and many other fields: true interdisciplinarity not limited by traditional academic hedges. Most of all though, the realization of Alternative Intelligence would also necessitate a synergistic partnership between AI and humans, blending cutting-edge technology with human perspectives, including those of outliers, who are typically overlooked. As a result, Alternative Intelligence would necessitate creating and preserving an enduring connection with the human world (at least until creativity can be fully synthesized). As for human intellect, its ongoing conversations with machines would enable it to grow much faster than previously possible, creating a true win-win scenario.

Not surprisingly, the actual management of this kind of AI would be fraught with unpredictable difficulties. We would need to ask questions such as: “Should Alternative Intelligence be permanently active or should it be switched on and off?”. That question alone has the potential to give rise to all kinds of ethical and other considerations. And, of course, there would be serious challenges in actually creating it.

Finally, the writing of articles such as this inevitably raises doubts in the mind of the author as to the value of its main conceptual feature: Alternative Intelligence. Does the concept have value for the field or is it a false problem? Is it an issue worth raising and writing about that may have an impact on future versions of Artificial Intelligence or is it something that can already be accomplished today by using current AI as augmented intelligence or with agents? So it was decided to test this by consulting an AI expert that is well-equipped to know the answer, ChatGPT itself. Here is a transcript of how it summarized the value of the ideas presented here after reading a draft of this paper.

The author’s prompt (after uploading the draft) was:

Do we need new technology to achieve Alternative Intelligence, or can we achieve it using ChatGPT or something like agents?

ChatGPT responded:

“Achieving Alternative Intelligence transcends the capabilities of current technologies like ChatGPT and agent-based systems. While these existing tools offer a solid foundation, realizing Alternative Intelligence requires the development of novel architectures and innovative paradigms that fundamentally differ from traditional AI. This new form of intelligence must integrate diverse data modalities, implement robust bias mitigation frameworks, and support dynamic, real-time learning and adaptation to anticipate future challenges effectively.

Moreover, leveraging existing technologies in creative ways — such as hybrid and modular systems, enhanced training techniques, and fostering human-AI collaboration — can significantly contribute to this evolution. To successfully develop Alternative Intelligence, a multi-faceted approach is essential, encompassing cutting-edge research, interdisciplinary collaboration, ethical considerations, and scalable infrastructure. By combining these efforts, we can pave the way for a transformative intelligence that not only overcomes the limitations of current AI but also propels us into a promising new era.”

It is both informative and reassuring to see that ChatGPT agrees as to the complexity of the task of creating Alternative Intelligence and acknowledges that it is both original and non-trivial. It seems aware of its own limitations and appreciates that its performance can be transcended in the manner suggested here. This judgment is also comforting in relation to the validity of the argument made in this paper while outlining a possible path forward for humanity to improve its intellectual position.

There are many challenges that the implementation of Alternative Intelligence raises, not only in relation to the growth of the human intellect but also to the general impact of advanced AI systems on society in general. The changes suggested here are not domain-specific but are embedded in an ecological intellectual structure with far-reaching impact.

XI CONCLUSION

In the spirit of this study, and given its exceptional abilities to synthesize and summarize, the last word is given to ChatGPT to bring this paper to a conclusion. When asked to produce that conclusion, it wrote:

“The integration of AI into human intellectual systems signals a profound transformation, reshaping not just education but the broader structures of knowledge creation and dissemination.

This shift transcends specific domains like education, representing a broader redefinition of intellectual processes across disciplines. Tools like AI-powered books highlight the potential for adaptive, personalized, and interactive environments that challenge traditional, static frameworks and the values underpinning them, raising questions about authorship, collaboration, and intellectual labor.

However, AI’s reliance on historical data and backward-looking models limits its capacity to address humanity’s future needs. To transcend these constraints, the concept of Alternative Intelligence offers a forward-looking paradigm that fosters novel thought processes and creativity. Unlike traditional AI, which enhances existing capabilities, Alternative Intelligence seeks to break free from cycles of past knowledge and explore innovative intellectual frameworks.

This transformation is not just an opportunity but a necessity. The rapid pace of technological advancement starkly contrasts with the slower evolution of human intellectual systems. Without decisive action, society risks falling behind, unable to adapt to or shape the emerging landscape. Alternative Intelligence provides a pathway to bridge this gap, enabling humans to redefine intellectual boundaries and align with the possibilities of a rapidly changing world.

The implications of this shift extend beyond specific applications, inviting a rethinking of how knowledge is created and shared across disciplines. By questioning entrenched paradigms and prioritizing adaptability, humans can leverage AI to craft entirely new intellectual landscapes. Ultimately, the convergence of human and machine capabilities offers the potential to unlock unprecedented growth and innovation, shaping a future that transcends the limitations of current systems and redefines intellectual progress for a new era.”

REFERENCES

[1] Council of Europe, Common European Framework of Reference for Languages: Learning, teaching, assessment. Council of Europe, 2011. Accessed: Dec. 30, 2024. [Online]. Available: https://rm.coe.int/168045b15e

[2] A.-P. Lian, “Reflections on Language Learning in the 21st Century: the Rhizome at Work,” Rangsit J. Arts Sci., vol. 1, no. 1, pp. 3–15, 2011, [Online]. Available: http://www.rsu.ac.th/rjas/article.php?id=10

[3] A.-P. Lian and P. Sangarun, “Rhizomatic learning systems and precision language education: A perfect match (Chapter 8),” in New Directions in Rhizomatic Learning From Poststructural Thinking to Nomadic Pedagogy, M. S. Khine, Ed., in Routledge Research in Education. , United Kingdom: Routledge - Taylor & Francis Group, 2023, pp. 119–141. [Online]. Available: https://www.routledge.com/New-Directions-in-Rhizomatic-Learning-From-Poststructural-Thinking-to-Nomadic/Khine/p/book/9781032453088

[4] B. Sparrow, J. Liu, and D. M. Wegner, “Google effects on memory: cognitive consequences of having information at our fingertips.,” Science, vol. 333, no. 6043, pp. 776–8, Aug. 2011, doi: 10.1126/science.1207745.

[5] OECD, Measuring Student Knowledge and Skills. Paris: OECD, 2000.

[6] R. Kaplan-Rakowski and A. Gruber, “The impact of high-immersion virtual reality on foreign language anxiety.,” Smart Learn. Environ., vol. 10, no. 46, 2023, [Online]. Available: https://doi.org/10.1186/s40561-023-00263-9

[7] A.-P. Lian and B. K. Joy, “The Butcher, The Baker, The Candlestick Maker: Some Uses of Dialogue Generators in Computer-Assisted Foreign Language Learning,” Aust. Rev. Appl. Linguist., vol. 6, no. 2, pp. 60–71, 1983.

[8] A. Basiouny and C. Terwiesch, “ChatGPT Passed an MBA Exam. What’s Next?,” Wharton Business Daily, Philadelphia, Jan. 23, 2023. [Online]. Available: https://knowledge.wharton.upenn.edu/podcast/wharton-business-daily-podcast/chatgpt-passed-an-mba-exam-whats-next/

[9] A.-P. Lian and P. Sangarun, “Precision Language Education: A Glimpse Into a Possible Future (Feature article),” GEMA Online® J. Lang. Stud., vol. 17, no. 4, pp. 1-15 (SCOPUS), Nov. 2017, doi: 10.17576/gema-2017-1704-01.

[10] A.-P. Lian, “Technology-Enhanced Language-Learning Environments : a rhizomatic approach,” in Computer-Assisted Language Learning: Concepts, Contexts and Practices, J.-B. Son, Ed., New York, NY: iUniverse, 2004, pp. 1–20. Accessed: Apr. 24, 2017. [Online]. Available: http://www.andrewlian.com/andrewlian/prowww/apacall_2004/apacall_lian_ap_tell_rhizomatic.pdf

[11] A.-P. Lian and M. V. Pineda, “Rhizomatic Learning: ‘As… When… and If…’ A Strategy for the ASEAN Community in the 21st Century,” Words, vol. 2, no. 1, pp. 1–28, 2014, [Online]. Available: http://journal.wima.ac.id/index.php/BW/article/view/508/487

[12]P. K. Phetsangkhad, “Exploring the Thai EFL Landscape: Implications for English Language Teaching,” Law Humanity. Q. Rev., vol. 2, no. 1, pp. 44–50, 2023, doi: 10.31014/aior.1996.02.01.50.

SYNAPSES, CIRCUITS AND SYNERGIES: TOWARD A TRANSFORMATIVE ECOLOGY OF HUMAN INTELLECT

Bài viết khác

Bài viết khác