Agentic AI or Automating Clicks?

Agentic AI can automate workflows, analyze data, and fire off emails, but it still can’t pour concrete or dig a tunnel. Agentic AI lives in the digital world and operates entirely through software, APIs, data, and cloud tools. Until project robots arrive, AI agents are only automating the actions we take in front of our computer. Of course, creating a status report, managing emails, and examining resource issues are all important and become more efficient using AI. Projects exist in the physical world, changing reality through infrastructure projects or triggering consequences in the real world such as improved cost, quality, or work efficiency. That’s the main disconnect: projects happen in the physical world, while agentic AI operates entirely in the digital one. Project managers are the ones who bridge that gap.

How much can agentic AI really help with construction projects where physical materials and manual work are essential to achieving milestones? AI-based predictions of impending problems and clear recommendations for a resolution are valuable, especially in a dynamic and uncertain environment. This is dependent on ensuring the underlying data is structured, complete and timely. That’s a problem for many projects (Decker et al., 2022). If even one team member fails to update a critical task, the status report becomes inaccurate, leading to misinformed stakeholders. For agentic AI to be effective, it needs more than accurate data. In the dynamic environment of a project, agents need access to reliable data which is a critical responsibility for project managers who want to rely on AI agents.

The future belongs to PMs who know how to combine real-world judgment with digital assistants that can automate the busywork and amplify insight.

Reference
Decker, D., Edelman, D., & Sharma, A. (2022).
How data can help tech companies thrive amid economic uncertainty.
McKinsey & Company.