Jamal SaidParnter and Senior Consultant| Meirc Training and Consulting
One of the biggest challenges I see when organizations talk about “AI” is that they treat it as a single capability when in reality, it spans very different types of work. The signals that help distinguish AI use cases are quite practical: • If the focus is on prediction then we are dealing with analytics or machine learning • If the focus is on content creation then we are talking about generative AI • If the focus is on automation then we are in the realm of process optimization • If the focus is on decision support I believe this is where AI becomes strategic What tends to go wrong is when all of these are lumped together under one label. This leads to:
Misaligned expectations
Poor tool selection
Confusion between efficiency gains and strategic impact
In my experience, clarity at the outset of what problem we are trying to solve, and what type of AI is required, is what separates meaningful adoption from wasted investment.
When it comes to conversations around AI, in one discussion, leadership was talking about AI as a strategic capability to enhance decision-making, while the operational team understood it as automation of reporting. At the same time, the IT team was thinking in terms of infrastructure and data pipelines. We were all aligned on the word “AI,” but not on what it actually meant. The tipping point is usually when:
• Different stakeholders describe different outcomes • The same term is used, but expectations diverge • Conversations move forward without defining the problem clearly
This is where a structured conversation becomes critical, aligning on purpose before jumping into tools.
When it comes to my experience wiht AI, I focus on about the underlying need. In practice, I’ve learned to pause and ask: • What decision are we trying to improve? • What inefficiency are we trying to eliminate? • What insight are we currently missing? Only then does the conversation become meaningful, AI is not a starting point, it is an enabler. Hence, organizations that start with tools often struggle. Those that start with problems tend to succeed. In leadership conversations, reframing the discussion from “AI adoption” to “value creation” makes, in my professional opinion, all the difference. Saving Changes...
Most individuals relate AI to LLM lihe ChatGPT. There are very few individuals who realize that AI is on an "agentization" process, evolving from the current assistant status.
Agentization refers to the process of turning an AI system (such as a LLM) into an autonomous agent that can:
Perceive its environment (through inputs, data, APIs, sensors, etc.)
Make decisions based on goals
Take actions using tools or external systems
Adapt based on feedback or changing conditions
Totalmente de acuerdo. La Inteligencia artificial debe ser vista como algo mas allá de los prompt y los grandes modelos de lenguaje (LLM). Excelente punto de vista Eduard. Saving Changes...
Vasti HerreraProject Manager| Manao.mxMexico City, Mexico
AI sounds catchy and trendy.ñ, but we must never forget it is another tool. The important thing is to define the problem first and then analyze if it is the right tool for this problem. Saving Changes...
Agree with the sentence that "we should use AI" as there are many time consuming things in daily life of PMs can be done quickly. for.e.g. MOMs from important meetings or Creating PPTs or having suggestions of complex problems by scanning organization database etc. Here is personal experience - Got chance to check on one AI experiment where tried to generate user stories based on requirement document scanning. It lead to stories creation but the accuracy was only 10-20%. 90-80% stories has to be rewritten by manual intervention to define actual requirements from user perspective. It got little time consuming further. so to identify the project and AI type is crucial. Also input data quality, security and privacy, ethical biases are some aspects to be considered. Saving Changes...
Oscar MiganiProject Management| Rural energy AgencyDodoma, 3, Tanzania, United Republic Of
I have tried to use PMI Infinity as one of the AI, it is good for project management inputs , and it simplify some assignment by making the work easy , i recommend PM to get used to this AI, for it is being one of the benefit of being PMI member Saving Changes...
I distinguish AI work by clarity of the problem, type of solution (automation vs. machine learning), data readiness, and measurable objectives. The issue with lumping everything under “AI”: scope creep, unrealistic expectations, and overcomplicating simple problems. Key signal of misalignment: different expectations between management, technical teams, and end users. Approach: turn the idea into a clear business case, define scope and measurable deliverables, and distinguish between a PoC and a full project. Saving Changes...
Anonymous
AI in project management is like having a smart assistant that not only organizes tasks but also predicts challenges before they happen, helping teams work smarter,not harder. Saving Changes...
When someone says, “we should use AI,” we should start with what is the core challenge that is needs to be solved, what should the future look like in an ideal scenario, and how will AI help us get there Saving Changes...
Emma WealthyProject Manager Executive| WPP MediaLondon, United Kingdom
Hi everyone,
From my own experience, it seems "AI" has become quite a broad term, meaning different things to different people. On one hand, it's undeniably a new and exciting frontier, and there's a natural eagerness among stakeholders (and ourselves!) to jump on the bandwagon. I've certainly encountered internal stakeholders who are incredibly enthusiastic about transforming or optimizing their day-to-day processes and workflows by "implementing AI."
I recently worked on a project precisely geared towards identifying and enhancing our internal processes, ultimately to better serve our clients. The initial framing, largely driven by this prevalent enthusiasm, was heavily centered on how AI could revolutionize these workflows.
However, as we moved into the discovery phase and engaged directly with various functional teams to understand their actual needs for process transformation, a consistent pattern emerged. A significant portion of the changes and optimizations they desired – things like saving time, improving data accuracy, or achieving 24/7 access to certain analyses – didn't actually require AI.
What we found was that many of these valuable improvements were already achievable using existing tools, systems, or even simple process adjustments within our current organizational capabilities.
The real challenge wasn't a lack of technological solutions, but often a lack of effective communication, awareness, or proper utilization of the resources already at hand. It became apparent that the initial push for "AI" was frequently born more out of the excitement surrounding the technology itself, rather than a deep analysis of specific functional gaps that only AI could address.
This experience highlights to me that whilst it's absolutely vital to explore AI's potential, we might be inadvertently overlooking immediate, non-AI opportunities for significant process improvement. Sometimes, the most impactful first step isn't about deploying a new AI solution, but about strategically reviewing our existing ecosystem and ensuring we're maximizing its efficiency.
The conversations around AI can become quite convoluted precisely because of this eagerness. There's often a push to implement AI "as soon as possible" without a sufficiently defined strategy. As project managers, I believe our role is crucial in guiding these discussions.
uWe need to be asking:/u
Why are we looking to implement AI in this specific process? What problem are we trying to solve?
How will AI specifically change or enhance the process?
What elements of the current workflow will be impacted?
And, most importantly, what specific AI tools or capabilities are we considering? "AI" is not a single solution; different applications will require different technologies. This demands a thorough discovery phase to move beyond the generic concept of "AI" to concrete, appropriate tools and benefits.
uMy key takeaway:/u AI should be approached not as magic and hastily applied, but as a powerful, strategic tool. A planned, controlled implementation focused on clear objectives, a comprehensive understanding of all available solutions (AI and non-AI), and robust monitoring of realized benefits will always yield better results than simply adopting the latest buzzword. Saving Changes...
PATRICIA ARCILA RAMIREZINGENIERA DE SISTEMAS| SECRETARIA DE EDUCACION DISTRITALBogota D.C, DC, Colombia
For me, the clearest sign is when someone says “let’s use AI” but doesn’t really know what problem they’re trying to solve. At that point, everyone starts imagining something different—one person thinks of a chatbot, another thinks about automating tasks, someone else is picturing something super advanced… and in the end, no one is actually talking about the same thing. What usually goes wrong is exactly that: expectations get really high without grounding the idea. So people expect AI to magically fix everything, but there’s no clear data or defined goal. Yeah, I’ve been in conversations like that, and I notice it pretty quickly when people start suggesting completely different approaches. That’s when I pause and ask simple questions like “what exactly are we trying to achieve?” or “what problem are we solving?” That helps a lot to get everyone aligned. Saving Changes...
PMI Team Member
