Artificial Intelligence Before and Beyond ChatGPT: A Sociological Reconstruction of Its Historical Development

Author: Youssef Serkal, Independent Researcher

Received 10 June 2025; Revised 28 July 2025; Accepted 5 August 2025; Available online 22 August 2025; Version of Record 22 August 2025.

Abstract

Artificial intelligence (AI) is frequently presented in public discourse as a recent breakthrough associated with systems such as ChatGPT. Such a view is historically incomplete. AI has a much longer intellectual genealogy that extends from ancient cultural imaginaries and philosophical reflections on artificial beings to twentieth-century computer science and contemporary generative models. This article reconstructs that trajectory while arguing that AI should not be interpreted only as a sequence of technical innovations. It should also be understood as a social field shaped by struggles over legitimacy, capital, institutional prestige, and global inequality. To develop this argument, the article integrates three sociological perspectives: Bourdieu’s theory of economic, cultural, and symbolic capital; world-systems theory; and institutional isomorphism. Through this interdisciplinary lens, AI appears not merely as a scientific achievement but as a historically embedded social phenomenon influenced by myth, power, competition, and global asymmetry. The article concludes that the rise of generative AI is best understood as the latest phase in a long historical process in which technical development and social structure have remained deeply interconnected.

Keywords: artificial intelligence, sociology of technology, Bourdieu, world-systems theory, institutional isomorphism, generative AI, symbolic capital, global inequality

1. Introduction

Artificial intelligence is often described as though it emerged suddenly in the early 2020s, especially with the popularization of ChatGPT and related systems. This impression has been reinforced by media attention, political debate, commercial marketing, and public fascination with conversational models. Yet the idea that AI began with recent applications reflects a narrow historical memory. It ignores the long intellectual, technical, and cultural development that made contemporary AI possible.

A broader historical perspective shows that AI has evolved through repeated cycles of imagination, experimentation, institutional support, disappointment, and renewal. Its development has never been purely technical. Scientific progress has always been influenced by social conditions such as funding priorities, geopolitical competition, professional authority, and institutional legitimacy. For this reason, AI should be examined not only through computer science but also through sociology.

This article critically reconstructs the long history of AI and situates it within three major sociological frameworks. First, Bourdieu’s theory of capital helps explain how AI has functioned as a site of struggle over prestige, authority, and knowledge. Second, world-systems theory clarifies how AI has developed unevenly across the global order, with innovation concentrated in core regions and support labor often located elsewhere. Third, institutional isomorphism explains why organizations repeatedly adopt AI in response to pressures of competition, legitimacy, and imitation. Together, these perspectives reveal that AI is not simply a neutral technology. It is a historically produced and socially organized field of power.

2. Ancient Imaginaries and the Prehistory of AI

The roots of AI extend far beyond modern computing. Long before algorithms and data infrastructures, human societies imagined artificial beings endowed with movement, intelligence, or agency. Ancient myths and legends from different civilizations included stories of mechanical servants, animated statues, and human-made beings that blurred the boundary between nature and artifice. These narratives reveal a long-standing human fascination with the possibility of creating non-human intelligence.

Such imaginaries were not mere entertainment. They reflected deeper social and symbolic structures. Myths of artificial life often appeared in contexts where religious, political, or philosophical authority was being asserted. In this sense, they can be interpreted as early symbolic resources that expressed hopes for mastery over matter, knowledge, and order. From a Bourdieusian perspective, these stories may be understood as part of symbolic capital: cultural forms that reinforced the authority of those who claimed proximity to special knowledge, whether priests, rulers, or intellectual elites.

The significance of these proto-AI narratives is not that they anticipated contemporary machine learning in a technical sense. Rather, they demonstrate that the desire to create intelligence outside the human body has deep historical roots. AI, therefore, did not emerge only from laboratories; it also emerged from long traditions of cultural imagination. This prehistory matters because modern technological projects often draw legitimacy from older symbolic narratives about invention, control, and transcendence.

3. The Scientific Formation of AI in the Mid-Twentieth Century

AI became a formal scientific field during the mid-twentieth century, especially with the Dartmouth Conference of 1956, which is widely recognized as a foundational moment. Early pioneers such as John McCarthy, Marvin Minsky, Allen Newell, and Herbert Simon proposed that aspects of human reasoning could be simulated computationally. This early phase emphasized symbolic AI, in which intelligence was understood largely in terms of logical representation, formal rules, and problem-solving procedures.

The scientific ambition of this period was substantial. Researchers believed that machines might soon perform tasks associated with human cognition, including reasoning, language use, and learning. However, this optimism cannot be explained only by technical confidence. It also reflected the institutional and geopolitical context of the Cold War. Universities, defense agencies, and research centers supported AI not only because of its intellectual promise but also because it represented scientific modernity, national prestige, and strategic advantage.

World-systems theory provides an important perspective here. The early institutionalization of AI took place mainly in powerful core countries with access to research capital, military funding, advanced universities, and computational infrastructure. In this sense, AI emerged not simply as a universal scientific endeavor but also as part of a broader struggle for dominance within the global knowledge economy. The field’s early authority was inseparable from the position of its leading institutions in the international system.

4. AI Winters and the Problem of Legitimacy

The history of AI is not one of linear progress. It includes periods of intense optimism followed by disappointment and reduced support. The first major AI winter occurred in the 1970s, following criticism of the field’s limited practical achievements. A second period of decline emerged in the late 1980s, when enthusiasm around expert systems weakened and commercial expectations were not fully met.

These episodes are often described as technical failures, but such an explanation is incomplete. AI winters were also crises of legitimacy. Organizations that had invested in AI faced rising pressure from funders, policymakers, and professional communities to justify continued support. When promised outcomes were not delivered quickly enough, confidence weakened and institutional retreat followed.

Institutional isomorphism helps explain this pattern. Organizations do not adopt emerging fields only because they are effective; they also adopt them because they signal modernity and competence. When AI appeared innovative and prestigious, universities, laboratories, and firms aligned themselves with it. When doubts spread, those same organizations adapted by reducing visible commitment. The result was not only a scientific slowdown but a collective institutional adjustment shaped by external expectations. In this sense, AI winters were social as well as technical events.

5. Expert Systems and the Codification of Expertise

Despite these setbacks, the 1980s witnessed important advances through expert systems, especially in fields such as medicine, engineering, and business decision-making. These systems attempted to translate the specialized knowledge of experts into formal rules that machines could process. They represented a significant effort to operationalize human judgment and make professional knowledge reproducible through software.

This moment is particularly important from a sociological perspective. Bourdieu’s concept of cultural capital is useful for understanding expert systems as projects aimed at converting embodied expertise into codified and institutionalized form. Professionals possess forms of knowledge that are acquired through training, practice, and social recognition. Expert systems sought to detach that knowledge from the person and relocate it into technical systems.

However, this translation was only partially successful. Much expert judgment depends on tacit understanding, context, intuition, and situational interpretation. These features are difficult to formalize completely. The limitations of expert systems therefore revealed an important sociological truth: human knowledge is not always reducible to explicit rules. Even so, the ambition behind expert systems reflected a broader modern desire to standardize, transfer, and control knowledge. In that respect, they marked a critical stage in the social history of AI.

6. Machine Learning and the Reorganization of the Global Knowledge Economy

From the 1990s onward, AI increasingly shifted away from symbolic reasoning toward statistical approaches and machine learning. Rather than relying mainly on hand-crafted rules, these systems learned patterns from data. This methodological transition corresponded with wider transformations in digital infrastructure, the growth of computational power, and the increasing availability of large datasets.

This phase was deeply connected to changes in the global political economy. The expansion of digital capitalism, platform economies, internet-based commerce, and data-intensive business models created a new environment in which machine learning became highly valuable. AI was no longer just an academic field. It became an infrastructural tool for markets, administration, logistics, communication, and surveillance.

World-systems theory is especially relevant here. The resources necessary for machine learning, including large-scale infrastructure, research funding, and high-value computation, remained concentrated in core economies. At the same time, other regions were often incorporated into AI development in subordinate ways, such as through data generation, annotation labor, or market consumption. This uneven structure reinforced existing global hierarchies. AI did not erase inequality; in many cases, it reorganized and intensified it within new digital forms.

7. Deep Learning and the Cultural Logic of the 2010s

The rise of deep learning, especially after 2012, marked another major transformation in AI. Advances in neural networks, graphical processing units, and large-scale training methods allowed systems to perform impressively in image recognition, speech processing, and other tasks once considered difficult for machines. Deep learning rapidly became the dominant paradigm within AI research and industry.

Its success was partly technical, but it was also cultural and institutional. Deep learning was embraced not only because it worked well in many domains but also because it became the symbol of cutting-edge intelligence. Once major research centers and firms demonstrated successful applications, other organizations followed. Governments invested in national AI strategies, universities redesigned curricula, and companies rebranded themselves around AI capacity.

This pattern is well explained by institutional isomorphism. Coercive pressures emerged through funding frameworks and public policy. Mimetic pressures pushed organizations to imitate perceived leaders. Normative pressures came from professional communities that increasingly treated deep learning as the accepted standard. As a result, deep learning became more than a method; it became a cultural and institutional model of what legitimate AI should look like.

8. Generative AI and the Reconfiguration of Symbolic Production

The 2020s introduced a new stage through generative AI, including large language models and image-generation systems. Unlike earlier AI applications focused mainly on classification or prediction, generative systems produce novel text, images, code, music, and other forms of content. Their public visibility has reshaped global discussions about creativity, authorship, education, labor, and communication.

Generative AI can be interpreted through Bourdieu’s concept of symbolic capital. Organizations associated with these systems often gain prestige, visibility, and legitimacy. Possessing advanced generative AI has become a marker of technological authority. At the same time, such tools appear to democratize access to forms of cultural production that once required specialized skills. Writing, design, and ideation are now more widely accessible through interactive systems.

Yet this apparent democratization should be approached carefully. The ability to use AI tools is expanding, but control over the most powerful models remains concentrated in a relatively small number of corporations and research centers. Economic capital, computational infrastructure, and data access remain decisive. Therefore, generative AI simultaneously broadens participation and reproduces hierarchy. It opens creative possibilities while strengthening the structural advantage of actors already positioned near the center of the digital economy.

9. The Social Dynamics of AI Hype

AI has repeatedly moved through cycles of excitement and disillusionment. These cycles are not accidental. They are produced by the interaction of scientific ambition, institutional competition, media amplification, and financial investment. Hype helps mobilize resources, attract talent, and generate legitimacy. At the same time, inflated expectations create vulnerability when systems fail to meet public or commercial promises.

From a sociological perspective, hype should not be dismissed as simple exaggeration. It is structurally embedded in fields where visibility and symbolic value matter. Scientific and technological domains often function as arenas in which actors compete not only over evidence but also over attention and prestige. AI is especially susceptible to this dynamic because it occupies a powerful place in the public imagination. It represents innovation, efficiency, and future transformation.

Institutional isomorphism intensifies this pattern. Once influential institutions invest heavily in AI, others often follow to avoid appearing outdated. This collective movement amplifies expectations beyond what the technology may immediately deliver. When those expectations weaken, disappointment spreads quickly across the same networks. In this sense, AI hype is best understood as a recurring social mechanism rather than as an isolated communication problem.

10. AI and Global Inequality

The global expansion of AI has not occurred on equal terms. Research leadership, advanced infrastructure, and strategic control remain concentrated in a relatively small number of countries and firms. Meanwhile, many regions participate under conditions of dependency, whether as consumers of imported tools, providers of labor for data preparation, or sites for policy experimentation without equivalent technical sovereignty.

World-systems theory highlights how this arrangement mirrors broader historical patterns. Core actors control high-value innovation and the symbolic authority attached to it. Peripheral and semi-peripheral actors often provide labor, resources, data, or markets without capturing the same level of value or influence. The result is a new form of asymmetry in which AI may reproduce patterns similar to older economic dependencies, although through digital rather than industrial mechanisms.

At the same time, this structure is not entirely closed. Some emerging economies are developing local AI ecosystems, regional collaborations, and sector-specific expertise. These efforts suggest that semi-peripheral actors may carve out strategic niches. However, such progress remains shaped by structural constraints, including reliance on external platforms, hardware, and standards. AI therefore reflects both the persistence of global inequality and the possibility of limited reconfiguration within it.

11. AI as Cultural Capital in Education and Professional Life

AI literacy is increasingly becoming a valued resource within education and professional life. Students, researchers, managers, and workers who can effectively use AI tools often gain practical advantages in productivity, communication, and decision-making. In this sense, AI competence is becoming a contemporary form of cultural capital.

Educational institutions have responded quickly. Universities and training centers are incorporating AI into curricula, assessment strategies, and skills development frameworks. This is not only a matter of pedagogical necessity. It is also a matter of legitimacy. Institutions that fail to engage with AI risk appearing disconnected from current economic and technological realities. Here again, institutional isomorphism is visible, as educational systems converge around similar patterns of AI integration.

However, access to high-quality AI education remains uneven. Well-funded institutions are better positioned to provide infrastructure, specialized faculty, and advanced training. Under-resourced institutions may struggle to keep pace. As a result, AI has the potential to widen educational and professional inequalities even while it promises broader access to knowledge. The social consequences of AI in education therefore depend not only on the tool itself but on the institutional conditions of access.

12. Ethical Discourses as Symbolic Struggle

Ethical debates about AI have become central to public and academic discussion. Questions of bias, transparency, accountability, labor displacement, privacy, and social harm are now part of mainstream policy and research agendas. These concerns are important and necessary. However, they are also part of a broader symbolic struggle over who has the authority to define legitimate AI.

Institutions that position themselves as leaders in responsible AI often gain symbolic capital. Ethical language can increase trust, strengthen public legitimacy, and enhance geopolitical influence. For this reason, AI ethics is not only a moral domain; it is also a field of strategic positioning. Actors compete to establish standards, shape regulation, and define acceptable practices.

This process can reproduce global asymmetries. Ethical frameworks developed in powerful contexts may be presented as universal while reflecting specific institutional interests and historical experiences. Less powerful actors may be expected to adopt norms they had limited role in shaping. Thus, ethical discourse can become a mechanism of soft power as well as a genuine effort at governance. A critical perspective does not reject ethics; rather, it asks who defines it, whose interests it serves, and how it circulates internationally.

13. AI, Capitalism, and the Logic of Accumulation

AI is closely linked to the logic of capitalist accumulation. Across sectors, it is used to optimize production, predict consumer behavior, automate management, personalize advertising, reduce labor costs, and intensify forms of monitoring. From this perspective, AI is not only an intellectual achievement. It is also an instrument for reorganizing value extraction.

Generative AI makes this especially visible. By producing text, images, and other outputs at scale, it changes the conditions of creative and knowledge work. Tasks once associated with specialized human labor can now be accelerated, fragmented, and redistributed. This does not mean that human creativity becomes irrelevant, but it does mean that its economic organization changes. Value increasingly shifts toward those who control platforms, models, and infrastructures rather than only those who produce content directly.

A critical but balanced interpretation is necessary here. AI can enhance productivity and expand access to useful tools. Yet it can also deepen surveillance, precarious labor, and concentration of power. These outcomes are not inevitable results of technology alone. They depend on how AI is governed, owned, and integrated into wider economic systems.

14. Theoretical Synthesis: AI as a Social Field

Taken together, the three sociological frameworks used in this article offer a comprehensive way to understand AI.

From Bourdieu’s perspective, AI is a field in which actors compete over economic capital, cultural capital, and symbolic capital. Technical innovation is inseparable from prestige, authority, and social recognition.

From world-systems theory, AI is embedded in an uneven global order. Innovation, infrastructure, and strategic control remain concentrated, while dependency and unequal exchange persist in new digital forms.

From institutional isomorphism, AI spreads not only because of technical merit but also because organizations imitate one another, respond to funding pressures, and seek legitimacy in rapidly changing environments.

These perspectives do not deny the scientific importance of AI. Rather, they enrich it by showing that technological fields are always socially situated. AI is produced through institutions, narratives, hierarchies, and global structures. It is therefore both a technical system and a social field.

15. Conclusion

ChatGPT did not create artificial intelligence. It marked a highly visible moment in a much longer and more complex history. The roots of AI extend from ancient cultural imaginaries to formal symbolic reasoning, from expert systems to machine learning, from deep learning to generative models. Across all these phases, AI has been shaped not only by technical invention but also by struggles over legitimacy, authority, labor, and global power.

A sociological approach makes this history clearer. It shows that AI is not simply a neutral tool moving forward on its own internal logic. It is built within institutions, financed through strategic interests, interpreted through cultural narratives, and distributed through unequal global systems. Its development reflects broader patterns of social organization, including hierarchy, competition, and symbolic struggle.

For this reason, understanding AI requires more than engineering knowledge. It requires historical depth and sociological imagination. Only by combining technical and social analysis can scholars fully explain why AI has developed as it has, why it generates recurring waves of hope and anxiety, and why its future consequences will depend as much on institutions and power as on algorithms themselves.

#ArtificialIntelligence #SociologyOfAI #HistoryOfTechnology #GenerativeAI #DigitalInequality #InstitutionalTheory #CulturalCapital #WorldSystemsTheory

References / Sources

• Bourdieu, Pierre. Distinction: A Social Critique of the Judgment of Taste.

• Bourdieu, Pierre. Forms of Capital.

• Wallerstein, Immanuel. The Modern World-System.

• DiMaggio, Paul & Powell, Walter. The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality.

• Kaplan, Andreas & Haenlein, Michael. A Brief History of Artificial Intelligence.

• Zhang, Lin. Artificial Intelligence: 70 Years Down the Road.

• Hajkowicz, Stefan et al. Artificial Intelligence Adoption in the Physical Sciences, Natural Sciences, Life Sciences, Social Sciences and the Arts and Humanities.

• Floridi, Luciano. The Fourth Revolution: How the Infosphere is Reshaping Human Reality.

• Russell, Stuart & Norvig, Peter. Artificial Intelligence: A Modern Approach.

• Pickering, Andrew. The Mangle of Practice: Time, Agency, and Science.