Artificial intelligence has become one of the most influential technologies shaping modern organisations, public institutions and professional practice. From predictive analytics and automation to generative models capable of producing text, images and software code, AI is often described as a technology advancing at unprecedented speed. Yet this perception of relentless progress obscures a deeper historical reality. AI has not evolved in a smooth or linear fashion but through recurring cycles of enthusiasm, disappointment and renewal.
This insight forms the foundation of a recent scholarly article by Matteo Cristofaro, published in the Journal of Management History, entitled “Surfing the AI waves: the historical evolution of artificial intelligence in management and organizational studies and practices”. The research was conducted at the University of Rome Tor Vergata and offers a historically grounded framework for understanding how artificial intelligence has shaped, and been shaped by, management theory and organisational practice.
Rather than focusing solely on algorithms or computing power, the article examines AI as a socio technical phenomenon. It highlights how repeated misalignments between technological innovation, academic theory and organisational adoption have driven both progress and failure. For researchers, clinicians, policymakers and organisational leaders, this historical perspective provides essential context for making sense of the current generative AI boom.
Why artificial intelligence advances in cycles
A central argument of the article is that AI evolves through waves rather than steady accumulation. Drawing on long wave theories of technological change, the research shows that periods of rapid innovation are frequently followed by phases of stagnation or reassessment. These cycles are not accidents but structural features of how complex technologies interact with organisations and societies.
In management and organisational studies, this cyclical pattern becomes particularly visible. When technological capabilities advance faster than theoretical understanding or organisational readiness, AI systems tend to disappoint. Conversely, when academic models become highly sophisticated without sufficient technological support, their practical relevance declines. AI’s history is shaped by this persistent tension between promise and practice.
Recognising these cycles is critical for contemporary debates about artificial intelligence strategy, ethics and governance. It suggests that current concerns around trust, transparency and human oversight are not novel challenges but recurring themes that have surfaced in every major phase of AI development.
The first wave and the promise of symbolic intelligence
The first wave of artificial intelligence emerged during the 1950s and 1960s, a period marked by post war optimism and faith in rational planning. Early AI research focused on symbolic systems that represented human reasoning through formal logic and explicit rules. Intelligence was understood as a process of manipulating symbols according to predefined instructions.
This approach aligned closely with early management theories, particularly the concept of bounded rationality. Organisations were viewed as information processing systems, and symbolic AI appeared capable of supporting managerial decision making by formalising reasoning under constraints. Early expert systems and decision support tools promised to enhance planning, forecasting and control.
Despite their theoretical elegance, symbolic AI systems struggled to cope with the ambiguity and complexity of real organisational environments. They required stable conditions and extensive manual updating, limiting their scalability. As expectations outpaced practical results, confidence in symbolic AI began to erode.
The AI Winter and the limits of overconfidence
By the 1970s and 1980s, disappointment with AI’s capabilities led to a period known as the AI Winter. Funding declined, public enthusiasm faded and many ambitious projects were abandoned. Governments and corporations questioned whether artificial intelligence could ever fulfil its early promises.
Within management and organisational studies, this period prompted critical reflection rather than rejection. Scholars explored alternative approaches that emphasised uncertainty, probability and learning rather than rigid rule based reasoning. Early neural networks, Bayesian models and fuzzy logic gained attention as ways to handle complexity more effectively.
Organisational applications of AI became increasingly narrow and specialised. Systems such as medical diagnostic tools demonstrated impressive accuracy in controlled settings but faced resistance due to lack of transparency and difficulty integrating into everyday practice. The AI Winter highlighted a recurring lesson that technical sophistication alone does not guarantee organisational value.
Learning from data in the machine learning renaissance
The third wave of AI emerged during the 1990s and early 2000s with the rise of machine learning. Instead of relying on predefined rules, these systems learned patterns directly from data using statistical and probabilistic methods. This shift transformed AI’s role within organisations.
Machine learning aligned with the growing emphasis on data driven decision making and analytics based strategy. Organisations began using AI to forecast demand, detect fraud, optimise supply chains and personalise customer experiences. The availability of large datasets and improved computing power fuelled this renaissance.
However, this wave also introduced new challenges. Many machine learning models operated as opaque systems whose internal logic was difficult to interpret. For managers and professionals responsible for high stakes decisions, this lack of explainability raised concerns about accountability, bias and trust. Once again, AI’s technical success outpaced its organisational acceptance.
Big data, deep learning and generative systems
The fourth wave of artificial intelligence began in the 2010s with the convergence of big data, deep learning and cloud computing. Advances in neural networks enabled AI systems to excel in areas such as image recognition, speech processing and natural language understanding. Artificial intelligence became embedded in everyday digital platforms and organisational processes.
A defining feature of this wave is the rise of generative AI. These systems are capable of producing original content, from written text and software code to images and molecular designs. In organisational contexts, generative AI promises productivity gains, creative support and accelerated innovation across sectors such as healthcare, finance and research.
Yet the challenges associated with earlier waves have intensified rather than disappeared. Deep learning models are even less transparent than traditional machine learning systems. Ethical concerns related to data privacy, environmental impact and algorithmic bias have become central to public debate. The gap between rapid technological advancement and slower organisational and regulatory adaptation remains a defining issue.
Towards a fifth wave of human AI collaboration
The article argues that artificial intelligence is now entering an emerging fifth wave focused on collaboration rather than substitution. In this phase, AI is increasingly framed as an augmentative partner that supports human expertise rather than replacing it.
This shift reflects lessons learned from decades of AI experimentation. Fully autonomous systems often struggle in environments where contextual judgment, ethical reasoning and emotional intelligence are essential. Hybrid models that combine human insight with machine intelligence are better suited to organisational decision making.
Explainable AI plays a critical role in this transition. By making algorithmic outputs more interpretable, organisations can build trust and accountability into AI systems. This is particularly important in fields such as medicine, finance and public administration, where transparency and responsibility are essential.
Reference
Cristofaro, M., & Giardino, P. L. (2025). Surfing the AI waves: The historical evolution of artificial intelligence in management and organizational studies and practices. Journal of Management History. https://doi.org/10.1108/JMH-01-2025-0002
