Abstract
General-purpose technologies (GPTs), also known as platform technologies, are catalysts for major changes in economic activity. They are general purpose in the sense that they have multiple applications across a wide range of industries and support waves of innovative products and services. The discovery of different ways to use the new form of GPT can leave an industry profoundly different from how it was previously. For example, the printing press, the steam engine, the telephone, and the computer have all launched waves of creative destruction that have made lasting changes in organizations and society (Schumpeter, 1942). The theory of multi-decade phases of economic growth, known as Kondatriev cycles, is underpinned by the explanation of far-reaching economic effects caused by GPTs (Rosenberg & Frischtak, 1984).
Predicting how and when a GPT will transform an industry is difficult because, like all other innovations, uptake of the technology takes the form of an S-curve (Rogers, 1962). In the early stages of development, the functionality of the technology may be limited and adoption slow. Many GPTs, such as those based in IT and communications, exhibit network effects, which means that the first movers may not see much initial reward for using the technology. In the rear vision mirror of history, it seems obvious that the steam engine was going to launch the industrial revolution, and the computer would change most facets of modern life, but this is due to the benefit of hindsight.
In reality, the transformative effects of GPTs are very hard to foresee. Agarwal et al. (2018) succinctly capture this problem with the observation that when the first iPhone came out, nobody foresaw the rise of ride-sharing apps and said: “Well, it’s curtains for the taxi industry” (p. 155).
We now see a new generation of GPTs that will change 21st-century organizations and find uses in project management. Predicting how they will be deployed in project management is a primary motivation for this special issue. Artificial intelligence (AI) and machine learning (ML) have already found a wide range of business applications, but their impact is only just starting to be seen in project management. Similarly, blockchain (distributed ledgers), originally used for bitcoin, is now used across a wide range of transactions—from finance and insurance to food supply chains and diamond mining. The possibilities for coordinating interorganizational projects with blockchain exist as well (Lu et al., 2022). On the horizon is a new form of computing called quantum computing, which has the potential to handle data and complex computations in a way well beyond the digital technologies we know today (Humble et al., 2021). In the same way computers have changed project management over the last few decades of the 20th century, quantum computers may support complex project management capabilities that are unimaginable for today’s project managers.
The declining cost of information technology is one of the major factors in the development of digital technology for project management. Sensors that turn information from the operating environment into digital information have existed for years but they have become cheaper and more powerful and are also complementary technologies for project management. In economic terms, these have the effect of lowering information costs in projects to almost zero, resulting in important project management technologies such as building information modeling (BIM) and augmented reality (Bryde et al., 2013). These technologies are supporting dramatic changes in project delivery models (Whyte, 2019). These declining information costs will also enable a new wave of GPTs that will further accelerate the transformation of project management
Practitioners and consultants are starting to think about how AI might be used in project management. Recent reports and blog articles suggest that repetitive tasks, such as cost estimation, will benefit from AI trained on previous, similar projects (Ludden, 2019). Going further, it is possible that AI could synthesize risk and issue logs from previous projects to predict the future success of a similar project, although this technology is not yet available (Ludden, 2019). Projecting into the future, PwC has speculated on the emergence of fully autonomous project management systems within 20 years that could replace the project management office (PMO) as we know it (Lahmann, 2018).
This wide range of AI futures is confusing for practitioners, and a more thorough investigation of current practices and possible impacts on project management practices is urgently needed. AI has already had a transformative impact on many industries. In the retail sector, AI is used to predict shopping behavior and make purchase recommendations. In manufacturing, AI has improved productivity through predictive maintenance of production equipment. AI can be conceptualized as a prediction engine that builds algorithms based on previous observations to determine what will, or should, happen next (Agrawal et al., 2018). Prior data observations can be given to the algorithm for a specific repetitive task (narrow AI) or the algorithm can learn by gathering data and updating the model to a wider range of situations (general AI) (Wang & Goertzel, 2012).
“The Expectations of Project Managers from Artificial Intelligence: A Delphi Study,” by Vered Holzmann, Daniel Zitter, and Sahar Peshkess (2022), deploys a Delphi study to examine possible futures for AI in project management. Using the expert opinions of 52 practitioners, the study shows that the predominant opinion is that AI will be deployed in schedule, cost, and risk management. The least relevant functions for AI are likely to be resource management functions, such as team management, recruitment, and communications, probably because these involve using data that are less able to be analyzed numerically.
In a demonstration of how ML and AI can be used to manage project schedules and analyze the causes of megaproject overruns “Reference Class Forecasting and Machine Learning for Improved Offshore Oil and Gas Megaproject Planning: Methods and Application,” by Ananth Natajaran (2022), examines reference class forecasting in energy megaprojects. While there have been many studies of factors contributing to poor time and budget performance in megaprojects, this article is novel in that it uses the data set to train the ML algorithm for predicting project-specific factors that create time and budget overruns. The combination of ML and reference class forecasting to manage project delivery risk is a powerful concept worthy of further attention from researchers and practitioners.
The fields of information economics and the theory of the firm can be a basis for theorizing how the cost and availability of information can change the organization of business processes (Williamson, 1991; Stiglitz, 2002; Birchler & Butler, 2007). In economic terms, AI lowers the cost of prediction but raises the value of information and decision-making as complementary assets. Until powerful, general AI technology is developed, human judgment will be needed to turn insights from narrow AI into solutions to complex problems. As AI produces more predictions, the value of the capacity for good judgment based on these predictions increases as well (Agrawal et al., 2018). This would place a premium on the skills of experienced project managers.
Megaprojects are information-rich and complex decision environments, where the consequences of an event are hard to foresee due to the temporal and spatial interdependencies among subprojects (Steen et al., 2017). Being able to predict events that contribute to cost overruns and delays in megaprojects is the domain of general AI, and the technologies that would facilitate this are still in development. However, megaprojects do have many repetitive functions, such as inventory and contract management, that are manageable with narrow AI and this is probably being done now in many megaprojects.
The extant literature on data analytics and AI that may impact megaproject management is reviewed by Sachindra Chamode Wijayasekera, Syed Asad Hussain, Amrit Paudel, Bhuwan Paudel, John Steen, Rehan Sadiq, and Kasun Hewage (2022) who follow these ideas in their article, “Data Analytics and Artificial Intelligence in the Complex Environment of Megaprojects: Implications for Practitioners and Project Organizing Theory.” The authors show that while there is a range of challenges in the adoption of digital technology, the organizational outcomes are potentially transformational in the design and delivery of megaprojects. These benefits include managing projects through the complete life cycle, enhancement of data accessibility, and more effective integrated project delivery. As a further contribution, the authors suggest that information economics can be used to theorize the impact of digital technology and AI, which leads to predictive propositions regarding the future direction of megaproject management.
The other review article in the special issue, “Digital Technologies in Built Environment Projects: Review and Future Directions,” by Eleni Papadonikolaki, Ilias Krystallis, and Bethan Morgan (2022), focuses on projects in the built environment and adds another dimension to the state of digital technology in projects by considering different levels of analysis of the digital technology and the level of technological complexity. One table in this article shows that the literature is largely dominated by studies of project-level technologies that can be characterized as system innovations. BIM can be included as an example of this category. The authors conclude that more studies are needed on individual experiences of specific technologies in the built environment. One observation from their matrix of analysis level and technological complexity is that there are fewer studies of radical digital technologies and their effect on projects and project-based organizing. Given the developmental acceleration of GPTs, we suspect there will be more opportunities for researchers to study radical technologies in project management.
An overlooked GPT is social media. It is hard to understate the way that social media has changed many aspects of modern life and economic activity. Social media also affects project management, especially in communication and stakeholder engagement. It allows the mobilization of people for or against a project with a force and scale that were unimaginable 10 years ago. This means that the news media is no longer a one-way flow of information from the official news outlets to the public. News is now co-created in an online environment as citizen journalists create their own media and respond to posted media articles.
This public process of news co-creation and opinion shaping is a rich data source for investigating the public legitimacy of projects and how stakeholders change narratives about projects. “Mobilizing Megaproject Narratives for External Stakeholders: A Study of Narrative Instruments and Processes,” by Johan Ninan and Natalya Sergeeva (2022) is an example of how these data can be used to investigate complex stakeholder relationships surrounding megaprojects. Using a sample of news media and public responses regarding the High Speed Two (HST) rail megaproject in the United Kingdom, the article categorizes the online material into stories, labels, and comparisons to theorize the mobilization of narratives to manage external stakeholders.
This collection of articles shows the breadth of opportunities that 21st-century GPTs offer to project management researchers. We encourage the research community to be proactive in thinking about how these technologies will change the organization of projects. In reviewing papers for this special issue, we received many conceptual papers and systematic reviews. These are useful and some have been included in this special issue but, ultimately, there is a need for empirical investigation of these phenomena if our theories of project-based organizing are going to stay abreast of the changing world of project management practice.