While the field of Project Management (PM) has historically relied on data for measurement and reporting, as well as predictive models and process controls/automation, we are currently at an inflection point due to recent advances in generative artificial intelligence (Gen AI), relying on computationally complex large language models (LLMs). The initial generative AI proofs-of-concept (Google’s Bard, OpenAI’s Chat-GPT) were built to showcase the power of LLMs in enhancing productivity and distilling knowledge in large unstructured public databases; they have generated excitement but have thrown up attendant risks (privacy risks, edge learning LLM hallucinations, data management barriers, and computational complexity). In addition, in this nascent phase, it has been a struggle to find clear, consistent, scalable use cases.
According to Gartner research cited in a recent Harvard Business Review article, by 2030 some “80% of project management tasks will be run by AI, powered by big data, machine learning (ML), and natural language processing.” That same article outlined 6 aspects of the PM function that are most likely to be adjusted as these tools develop:
Current Adoption
Given the still-developing nature of the tools, one of the major findings of the research is perhaps unsurprising at this point: AI has yet to be widely integrated into PM workflows. However, several patterns are emerging:
- PMs are cautiously exploring AI engines without full deployment. As one PM commented, “I haven’t seen AI in a practical setting within the PM function. I have used robotic process automation in the Power BI area. [Another] use is for AI to search in the content of what people are putting into the code to identify errors and problems systematically rather than using a PM flagging these through more of a manual process. Data ingestion models will evolve, and this will empower AI adoption. Extracting intelligence from chatbots as well may be a further element for ML utilization.”
- Many PMs see the potential to adopt AI as a productivity enhancer, rather than as a threat to their roles. Indeed, one PM referred to it as a “shadow employee.”
- Scheduling and critical paths are seen as high-potential areas for AI-led automation over the next 3-5 years. “AI could use timeline complexity and template data for this,” noted one PM.
- An additional use case–and one more likely to come onstream in the near future–is to leverage AI’s data analysis capabilities for risk management. As one PM noted, “AI can analyze historical project data and external factors to identify potential risks early on, enabling proactive risk management strategies and reducing the likelihood of project failures.”
The Future Role of PMs
Perhaps the most significant finding related to the future state of AI across the PM function is the extent to which PMs are not just open to adoption, but excited about the potential productivity gains on offer.
In a world in which routine scheduling and reporting will be largely automated, PMs will have more time for high-judgment responsibilities, coupled with the ability to learn from a much wider range of prior outcomes.
- The use of standardized predictive/forecasting algorithms can enhance “proactive” PM; AI augmented control, governance, and encryption tools are emerging additions to the available predictive modeling platforms.
- From the point of view of risk controls and compliance in operational settings or at the edge of the enterprise (operational losses, technology deficits, cybersecurity, data integrity, privacy), a focus on modifications to existing data infrastructure to accommodate unstructured data, event, and digital data, as well as relational data formats that can be useful for LLMs, can yield early use cases.
- The initial generative AI proofs-of-concept (Bard, OpenAI) were built to showcase the power of LLMs in enhancing productivity and distilling knowledge in large unstructured public databases. The emerging use cases are in training LLMs on proprietary databases that solve for many of these risks and are already being deployed at places like McKinsey to collate knowledge bases and enhance productivity.
Conclusion: PMs are Leaning Into a Brave New World
Across Carpe Diem’s research, the consensus among PMs is that AI will radically reshape their field over the next few years–even if its adoption has been somewhat slow and cautious to date. The expectation of most is that, as these tools improve, they will offer productivity gains, help with data-driven decision making, and automate many repetitive, time-consuming tasks. As such, PMs who upskill, invest in emerging data infrastructure/management, and prepare for this AI-powered future will be best positioned to succeed.
These market insights from Carpe Diem Global Partners are gathered from the firm’s extensive client work leading Board, CEO, CXO, and CHRO executive search engagements for public and private multinational companies. For deeper, custom insights, contact Michael Whitehead at mwhitehead@carpediempartners.com.