Jennifer CalimagStatewide Compliance Manager| State of ArizonaPhoenix, Az, United States
The usage of AI is a controversial topic in my industry because of the many threats to data and security. My signal is always to make sure we are adhering to data and security protocols and not go rogue in utilizing AI, for example, taking meeting notes (because of public record requests) and unsanctioned software. Saving Changes...
Good question, AI means different contexts and domains depending on who you are talking to; usually, they think of a robot ready to use their knowledge and give you all the assertive answers. But the AI is a tool who simulate the human intelligence and needs context and domain knowledge, which turns into data to train some model, and other components such as RAG, MCPs Servers, guardrails, and agents. etc, so it's important to understand how the AI works to get all the potential in our programs and projects. Saving Changes...
After 13+ years in AI across retail pricing, sensor fusion, computer vision, robotics, and now LLMs, my first move is always the same: ask what flavor of AI we're actually talking about. To me "AI" is an umbrella, not a synonym for ChatGPT et al.
The signal I listen for is what the person pictures themselves doing. Are they describing a tool they'd open (ChatGPT, Claude, Gemini), a capability they want embedded in a product (recommendation engines, computer vision, forecasting), or an agent they want operating without supervision? Same word, three very different scopes, costs, and risk profiles.
What goes wrong when we lump them together: the requirements conversation never closes. The PM scopes one thing, engineering builds another, the buyer expected a third. By the time anyone notices, the budget's spent.
The move I make before scoping anything: ask the requester to walk me through what the AI would actually do, end to end. If they can't, we're not ready to scope. We're ready to discover. Saving Changes...
Anonymous
We can't live and work without AI Saving Changes...
We should use AI in Project management Saving Changes...
H Murlidhar RaoIT Infrastructure Project Manager| NoneDubai, United Arab Emirates
The biggest signal is whether people mean strategy, automation, analytics, generative AI, or ML operations — because each needs a different problem statement, data shape, and success metric. What goes wrong is treating ‘use AI’ like a single solution, which leads to mismatched expectations, bad vendors, wasted budget, and disappointment when the tool solves the wrong problem. I usually ask: what decision or workflow are we trying to improve, what data do we have, and what does ‘success’ look like?
...
1 reply by Sonya Calef
May 19, 2026 4:49 PM
Sonya Calef
...
And what will the organization do to offset the environmental harm and quality of life impact to the people living by a huge polluting data center? What budgetary constraints are there on future litigations because of using GenAI and the definitely harmful infrastructure it requires? What will the organization's insurance company think about entering such high-risk solution space? IT has got to step up and do better vs. being enamored with a new and very harmful animal in the zoo. The security considerations are not trivial.
Often the response was that AI is a hyped fad and won't be used, but today there's a feeling that we're behind the times. That's why AI is used very superficially to write a text, an email, or take meeting minutes. They can't delve deeper and use data from several years of operation from over many unidades to support decision-making, estimates, case studies, etc.
To echo what others have said, when my teammates have talked about using AI, it really only seems to mean using ChatGPT to look something up. That’s all well and good, but making full use of AI is a bit more comprehensive than that. It likely also doesn’t help that only a handful of my team members have received training on putting together prompts (and let’s just say they’re the folks who are least likely to use AI in the first place).
...
1 reply by Sonya Calef
May 19, 2026 4:45 PM
Sonya Calef
...
ChatGPT is the world's most expensive search engine that has a steep environmental and quality of life cost, and it was never meant to be a little lookup tool. Frankly, I can get the work done in less time than engineering a prompt because I have put in the effort to become a professional the hard way - reading, writing, being mentored, and doing for 30 years. I do not need to hand my thinking and cognitive processes over to an overgrown Google. GenAI's value is in large dataset analysis. If you don't have very large data to analyze for specific outputs, it's a very, very expensive solution looking for a problem.
Saving Changes...
Carol WalshSenior Financial Analyst| DXC TechnologyMadison Heights, Mi, United States
I agree with Douglas Boyd, how do we know which AI system to use?
...
1 reply by Sonya Calef
May 19, 2026 4:41 PM
Sonya Calef
...
The same way you determine which new application to build or buy: requirements, requirements, requirements. Business process analysis and optimization. Fit-gap analysis. The core processes and solid business analysis will never let you down regardless of what technology happens to be doing at the moment. AI is not the answer to many problems.
When someone says “we should use AI,” the conversation is rarely about technology itself. It is usually about pressure for speed, efficiency, innovation, or competitive leverage. The first step is to clarify intent.
Three signals help distinguish what is really being asked.
First, decision proximity. Is AI automating a task, augmenting human judgment, or moving toward managing objectives autonomously? These are fundamentally different categories of work. The closer AI gets to consequential decisions, the stronger the need for governance, traceability, and explicit oversight.
Second, problem clarity. Is there a clearly defined business problem with measurable impact, or is AI being treated as the starting point? When the solution precedes the problem, misalignment and inflated expectations follow.
Third, accountability design. Who owns the outcome if an AI-driven recommendation fails? When responsibility becomes diffuse, risk scales faster than performance.
In many organizations, “AI” simultaneously means efficiency, experimentation, and cost reduction to different stakeholders. Misalignment becomes visible when decision flows and ownership are unclear. A common tipping point is when stakeholders use the same word “AI” but describe different success metrics.
The real shift is not from manual to automated. It is from “man in the loop” to “man in control.” Without deliberate design of responsibility, capability increases while accountability erodes.
Clarity of purpose, category of AI work, and ownership separates disciplined transformation from technological noise.
"In the real world, the right thing never happens in the right place and the right time. It is the job of journalists and historians to make it appear that it has."
PMI Team Member
