Project Management

Please login or join to subscribe to this thread

When someone says, “we should use AI,” how do you unpack what’s really being asked?

linkedin twitter facebook   Artificial Intelligence  
avatar
Michael Brinn
PMI Team Member
Product Manager, Learning| PMI Denver, Colorado, United States

What signals help you tell different kinds of AI work apart—and what tends to go wrong when everything gets lumped together?

Have you ever been in a conversation where “AI” meant different things to different people? What tipped you off?

Share your experiences navigating what’s really being asked when someone says “we should use AI” in the comments below.

Sort By:
< 1 ... 76 77 78 79 80 81 82 83 84 85 86 ... 94 >
avatar
BATHMANABAN S Trichy, Tamil Nadu, India

avatar
BATHMANABAN S Trichy, Tamil Nadu, India
I don't have personal experiences or real-world conversations, so I can't say I've been in one myself. But in many discussions, I've analyzed, "AI" often means very different things to different people.
avatar
Arthur Bialowas Tallahassee, Fl, United States
When someone says, "we should use AI," I think it's important to first understand the actual problem they're trying to solve. I would ask what task they want to improve, what outcome they're hoping for, and whether AI is the best solution. By asking these questions first, it's easier to choose the right AI tool and make sure it adds value instead of creating unnecessary complexity.
avatar
Inemesit Adeniyi Portage Regional Economic Development Portage la Prairie, MANITOBA, Canada
When I hear 'Lets use AI' the first thing that comes to me is identifying the business problem this usually determines the use case. When the use case is identified, the type of AI to employ usually stands out, I consider anonymizing data and activating the human loop in every stage of my work very important I have come to realize that so many people see using AI as an escape route not to put in the works. but we should remember that, especially, with social phenomena things can be very dynamic.
Feb 25, 2026 5:29 AM
Replying to Eduard Hernandez
...

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
That's an interesting point. As AI becomes more autonomous, I think trust and accountability will become even more important. If an AI agent makes a decision, who should be responsible for the outcome?
avatar
Anonymous
Since there was push to AI, I have started using it for summarising my weekly report
avatar
Mark Anthony Monje Fernandez Kowloon, Hong Kong
I’ve been in conversations where “AI” meant completely different things to different people. One clear signal is when expectations about the output don’t match. For example, a client may treat an AI‑generated house image as something close to a final design, while an architect or engineer sees it only as a sketch that still needs to be checked for structural safety, regulatory compliance, constructability, and cost. The gap becomes obvious once we talk about what the deliverable is supposed to be.
Trouble usually starts when all AI work is treated as the same. People may choose a tool that doesn’t fit the task, assume the data they have is enough, or rely too much on results that only look complete. In construction, this can lead to unrealistic timelines, incorrect cost assumptions, design mismatches, compliance issues, and confusion about who is responsible for verifying the output.
So when someone says, “We should use AI,” I begin by clarifying the basics: What problem are we trying to solve? What information do we actually have? What result do we expect? Who will check the output? How will accuracy and privacy be handled?
From experience, AI tools can help with analysis and planning, but they don’t replace professional judgment. Before using them, the team needs a shared understanding of the problem, the task, and who is accountable for reviewing the results. Once that’s clear, it becomes easier to define scope, risks, resources, and the value the tool can realistically provide.
avatar
AFOLABI KAMORUDEEN AJIBOLA Lagos, LA, Nigeria
Which decision or process do you want AI to improve? Estimating, planning, procurement, construction monitoring, quality control, HSE, document management, or project reporting? Once we identify the problem, we can determine whether AI is the appropriate solution
avatar
Jeison Cambronero Zuniga Santa Ana, Piedades, SJ, Costa Rica
I think currently the capacity of AI is being developed by the entries that users have and as we use it more for simple tasks or information the development of AI is limited and most likely to fail at some point cause we are not teaching it mostly to advance or have higher capability, so people needs to be careful from now on, on discerning if the answers of AI are going to be good or not, very easily misinformation can happen,
avatar
Abdelhamid Hammam Naghammadi, KN, Egypt
While it is recognized that AI can provide valuable support, clarification is required regarding the specific AI system to be used, as numerous AI platforms and tools are currently available.
< 1 ... 76 77 78 79 80 81 82 83 84 85 86 ... 94 >

Please login or join to reply

Content ID:
ADVERTISEMENTS

"If we knew what it was we were doing, it would not be called research, would it?"

- Albert Einstein

ADVERTISEMENT

Sponsors