10. Artificial Intelligence and Project Management
Although artificial intelligence (AI) has been around for over 20 years, hardly a day goes by without it appearing on our favourite news and social media sites. The work carried out by project management professionals is affected by advances in AI technology and many project management professionals are assisting organizations implement creative uses of AI technology.
Learning Outcomes:
After reading this Chapter, students will be able to:
- Define what artificial intelligence is.
- Discuss its impact on the project management profession.
- Discuss what organizations need to do to fully leverage AI’s capabilities.
Artificial intelligence defined.
Artificial intelligence is defined as “the capability of computer systems or algorithms to imitate intelligent human behavior”.1 Generative Artificial Intelligence (GenAI) is defined as “artificial intelligence that is capable of generating new content (such as images or text) in response to a submitted prompt (such as a query) by learning from a large reference database of examples.”2
GenAI models like ChatGPT (from OpenAI), Microsoft’s CoPilot and Google’s Gemini are chatbots that rely on large language models (LLMs). A chatbot is a computer program that is designed to emulate human conversation, and it’s often based on natural language processing (NLP). According to Tim Keary, Techopedia’s Technology Specialist, ChatGPT is the most well-known AI chatbot on the market and is extremely easy to use.3
For organizations, the appeal of this form of intelligence is it can simplify complex decision-making and automate routine processes, all in a quest to improve performance. AI is widely used in many diverse industries and sectors (profit and non-profit). It is so commonly used in our daily lives that we often forget it is there. Examples include web search engines (e.g. Google), purchase recommendation systems (e.g. Amazon, Crave, Netflix and YouTube), generative tools (e.g. ChatGPT, CoPilot) and generative creative tools (e.g. AI art).
Large language models (LLMs) are quickly improving their ability to perform a wide range of natural language processing (NLP) tasks. It’s interesting to note how well OpenAI’s various chatbot models perform on common exams. For instance, according to OpenAI, ChatGPT and ChatGPT4 would pass the Uniform Bar Exam with scores of 213/400 (10th percentile) and 298/400 (90th percentile) respectively.4 However, the results are strikingly different from the Chartered Financial Analyst (CFA) exams. A recent study by a team of AI researchers at Queens University, Virginia Tech, and J.P. Morgan concluded that ChatGPT would not pass the level I and level II CFA exams while ChatGPT4 has a decent chance of passing these exams.5 This is because financial reasoning is a unique challenge for LLMs.
There are 2 significant conclusions from this research. Firstly, AI’s impact varies considerably by profession. Second, free versions of OpenAI’s leading chatbots are not as effective as the paid version. Organizations that plan to invest in AI technologies will have to evaluate the pros/cons of market ready free versions, subscription-based alternatives and internally developed models. Cost/benefit, reliability and the protection of internal/customer data privacy are a few of the key considerations associated with this investment decision.
Although AI has the potential to deliver significant benefits, it does carry a lot of controversary. Since it is dependent on large databases of information, AI systems replicate the biases that are inherent in the data. Furthermore, many argue that the data used by GenAI applications do not respect intellectual property and copyright laws. Lastly, many AI systems do not explain how personal and confidential information is protected.
Impact of AI on the project management profession.
One of the first questions many organizations ask is “can AI improve organizational performance”? In the context of project management, there is lots of room for performance improvement. According to the Standish Group, only 35% of projects are considered successful.6
Successful projects deliver on their intended outcomes. In other words, they generate value for the organization. Ultimately, this incremental value assists organizations in achieving their vision/mission and business objectives. Common examples include improving customer satisfaction, increasing profitability and entering new markets with new products. It’s also important that this value is delivered in accordance with the expected timeline, cost and scope of the project. Given the high project failure rates, many organizations view artificial intelligence as an opportunity to improve overall organizational performance.
AI can assist project management teams in many ways. Examples include:
- Improve estimating accuracy (effort and cost).
- Draft project plans.
- Generate an initial work breakdown structure.
- Analyze resource allocations.
- Assess project performance.
- Compile status reports.
AI is advancing quickly and as such, this list is not meant to be exhaustive. It is important to note that widely available AI models are useful in generating initial drafts of plans, but refinement is necessary to ensure the information is relevant to the specific organization and the unique project.
Many of the uses for AI identified above will assist teams improve their productivity. This is an exciting opportunity because some of the time that is spent on creating plans can be spent communicating with project stakeholders. This alone can have a significant impact on project success.
Many project management professionals view AI as a companion in their decision-making processes. Common practices, typical risks and important assumptions can be easily identified. Without the proper context, they are theoretical at best. Successfully project management professionals can learn how to craft effective prompts, start with the theoretical responses provided by their AI system of choice and then apply their expert judgement acquired from working in a unique organizational time and place. To take theoretical information to practical and implementable solutions, an organization’s unique culture, structure, processes, resources and vision/mission must be considered. Generic information will never be as effective as tailored solutions.
Perhaps most importantly, what AI can’t do is connect with people. AI systems are not capable of identifying the individual people within an organization that are affected by a change. They are also incapable of comprehending the unique needs of these stakeholders. Similarly, they are incapable of empathizing with stakeholders who may fear change, lack the incentive to change and/or simply not see the rationale to change. Human beings inspire each other to move beyond their comfort zone and achieve something that was once unimaginable. The best project management professionals will always be effective change makers.
Project Management Offices (PMOs) have also found important uses of AI that include ways to improve the project selection, resource prioritization and project management training processes. This is leading to the creation of organization specific GenAI models. The Project Management Institute is also investing in a GenAI model for its members – PMI Infinity. The advantage of these GenAI systems is they offer content that has been generated and reviewed by members of the organization.
Organizational requirements for fully leveraging AI’s capabilities.
Organizations that are successfully using AI to improve their performance have invested in data and effective people change management practices.
Effective AI systems are built on huge volumes of information. If an organization wants to build their own AI system, it can take a lot of time to identify the information needed and to create a repository of reliable information. For instance, imagine an organization that wants to improve project estimates (effort and cost). This organization must identify the factors that influence estimate accuracy (use cases), collect all the relevant data points and build algorithms that allow the system to be used in a predictive fashion. In addition, the organization has to train users on how to write effective prompts so the model can be easily used. And as its name implies, Generative AI models evolve and this can lead to additional uses. Organizations need to invest the time in helping the AI system learn and the employees who use the system learn along with it. Presently, this is often out of reach for many organizations because they still using spreadsheets, slides, and other applications that haven’t evolved much over the past few decades.7
The other requirement is a culture willing to accept that AI systems can be reliable and worthy of investment. Organizations that are not willing to speak candidly about AI’s potential impact on employment levels, compensation, and roles and responsibilities will struggle to sustain their AI investment. Furthermore, learning how to craft effective responses must be a part of a well thought out AI training program for employees, regardless of the AI model used.
A discussion of AI’s impact would not be complete without a discussion about the potential for job loss. The project management profession includes many different roles – project coordinators, project managers, scrum masters and business analysts to name a few. Titles and responsibilities vary from one organization to another so it’s difficult to discuss the impact on specific roles. What we do know is that roles and/or people that are simply distributing publicly available information are more likely to be replaced by AI systems in the future. Professional success in today’s technology-infused environment is not about information recall. However, in the short to medium term, it’s important to note that the implementation of AI solutions is a driving force in the heightened demand for project management professionals. PMI estimates that 25 million new project management professionals will be required globally by 2030 as organizations embrace the use of new technologies8. It’s quite possible that some of the roles that may be candidates for elimination are the very roles that help build the AI systems that profoundly impact their jobs. Today’s workforce must be constantly retooling to maintain employability.
In conclusion, AI can be seen as a powerful tool for Project Managers to increase their productivity and improve the likelihood of project success, when the needed organizational investments have been made. AI is not a replacement for a Project Manager’s ability to connect with stakeholders, motivate project team members and creatively solve the many challenges often encountered in organizational transformation initiatives.
References
1 Artificial intelligence. (n.d.). Merriam-Webster dictionary. https://www.merriam-webster.com/dictionary/artificial%20intelligence
2 Generative Artificial Intelligence. (n.d.). Merriam-Webster dictionary. https://www.merriam-webster.com/dictionary/generative%20artificial%20intelligence.
3 Keary, Tim. (April 29, 2024). 10 Free Generative AI Tools That Are Great for Beginners. Technopedia. https://www.techopedia.com/6-free-generative-ai-tools-that-are-great-for-beginners
4 OpenAI. (March 14, 2023). GPT-4. https://openai.com/index/gpt-4-research/
5Callanan, E., Mbakwe, A., Papadimitriou, A., Yulong, P., Sibue, M., Xiaodan, Z., Ma, Z., Liu, X., and Shah, S. (2023). Can GPTmodels be Financial Analysts? An Evaluation of ChatGPT and GPT-4 on mock CFA Exams. Queens University, Virginia Tech and J.P. Morgan AI Research.
6 Nieto-Rodriquez, Antonio, and Vargas, Ricardo Viana. (February 2, 2023). How AI Will Transform Project Management. HBR.
7PMI. (n.d.). Narrowing the Talent Gap. https://www.pmi.org/learning/thought-leadership/narrowing-the-talent-gap#:~:text=The%20Talent%20Gap%3A%20Facts%20and%20Figures&text=This%20is%20expected%20to%20grow,professionals%20are%20needed%20by%202030.
8Nieto-Rodriquez, Antonio, and Vargas, Ricardo Viana. (February 2, 2023). How AI Will Transform Project Management. Harvard Business Review.
