Jean-Marie PijuanIT Manager| Federal GovernmentOgden, Ut, United States
Hello,
I am a data scientist and a Project Manager. I develop AI tools\apps and use AI "a lot". In your opinion, how will\does AI assist in developing and managing project(s) and what do you think are the pros and cons of using LLMs, GenAI, etc...? Should PMs or even the PMI develop its own AI infrastructure (I have my own ideas on this) . I may be wrong but I really think that we are left behnd using old oncepts.
Hi Jean-Marie, AI is transforming project management from reactive tracking to proactive intelligence. It helps by automating documentation, summarizing meetings, detecting risks early, forecasting delays, and supporting better decision-making. LLMs and GenAI reduce cognitive load, accelerate planning, and improve knowledge reuse.
Pros: speed, scalability, improved pattern recognition, enhanced decision support. Cons: overreliance on automation, data quality issues, bias risks, false confidence in predictions. In summary , AI should augment strategic judgment, not replace it and project management must evolve beyond traditional static models to stay relevant. Cdlt AS
Saving Changes...
Jean-Marie PijuanIT Manager| Federal GovernmentOgden, Ut, United States
Good evening Sir, I agree completely and Ihope that PMs use AI as a tools and not a solution. One of the keys to incorporattin AI in project management is its ability to predict issues or outcomes for example, provide recmmendations to cmplete taks timely but it can go even further. ps. Aerospace is alse one of my fields of study. Thank you for your comment. v/r, Jean-Marie Pijuan
Jean-Marie — I appreciate the way you framed this, especially the infrastructure question. That’s where the real shift is.
I don’t think AI’s primary impact on project management is automation. It’s reframing where judgment lives in the delivery system.
Traditional PM models assume:
Data is scarce
Analysis is periodic
Reporting cycles are fixed
Decisions happen in structured forums
AI changes those assumptions. Information becomes abundant, analysis near real-time, and weak signals surface earlier. That shifts the PM role from “information organizer” to decision architect and boundary designer.
If AI is meaningfully embedded, the conversation becomes architectural:
Where does the model sit in the delivery stack?
Is it ingesting structured portfolio data, unstructured artifacts, or both?
Is it advisory, or does it trigger workflow actions?
How is data lineage preserved so outputs are auditable?
What feedback loop exists to recalibrate based on outcome variance?
At that point, we’re not adding a tool — we’re introducing a probabilistic actor into governance.
That raises deeper design questions:
Who owns model outputs?
What confidence threshold warrants escalation?
What decisions remain explicitly human?
How do we prevent over-optimizing measurable variables while underweighting political, cultural, or regulatory nuance?
On pros and cons:
Pros:
Faster synthesis across fragmented data sources
Earlier detection of risk patterns and dependency conflicts
Real-time scenario modeling at portfolio scale
Reduced latency between signal and action
Risks:
Illusion of precision in probabilistic forecasts
Amplification of data quality issues
Governance drift if accountability isn’t redesigned
Shadow usage if guardrails aren’t explicit
I don’t think PMI needs to build technical infrastructure. But the discipline absolutely needs to define operating principles for AI-augmented decision systems — especially around traceability, override authority, and learning loops.
We’re not being left behind because the concepts are old. We’re at risk if we embed intelligence into delivery without redesigning the accountability model that surrounds it.
I’d be curious whether, from your perspective as someone building AI systems, you see the bigger transformation happening in execution — or in governance design. Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
Unfortunately in the last time some people and organizations are using generative AI as a synonym of AI with is the first step to fail. Generative AI is just a tool like the ohters type of AI that we are using from more than 40 years ago. Something obvious for you when I read your expertise on the matter. So the first cons is this: do not understand what generative AI realy is. Saving Changes...
AI is reshaping delivery by turning data into foresight. PMs gain faster risk detection, clearer prioritization, and sharper decision windows. LLMs accelerate analysis and communication, but bias, security, and over-reliance remain real risks. A PMI-led AI ecosystem could standardize ethics, data quality, and PM-specific models so the field evolves beyond old concepts Saving Changes...
[what do you think are the pros and cons of using LLMs, GenAI, etc...?] It's hard to do anything online and not hear about the pros of using LLMs/AI tools. I think the biggest con, or risk, is overreliance on AI tools - turning to AI first. AI makes it easy to outsource critical thinking, which can lead to shallow understanding, reduced ability to define problems, false confidence, and can become a crutch. Used well, it can help sharpen thinking - like a tutor. Overreliance can have the opposite effect.
[Should ... PMI develop its own AI infrastructure?] What use case(s) are you thinking of, and what old concepts do you think are causing us to be left behind? Saving Changes...
Community Champion
