AI Driven PM: S2E2 - What Do Project Managers Actually Do?

 So AI can write your status reports now.

It can take your meeting minutes. Update your project schedule. Do resource forecasting. Some tools even claim they can manage your risks and predict your delays.

Cool.

So what's left for us?

I'll tell you what's left: Everything that actually matters.

Here's the uncomfortable truth most PMs don't want to hear: Right now, you're probably spending about 60% of your time on coordination and busy work. That means only 40% of your job is focused on creating real value.

You can't avert a risk you didn't see coming if you're busy documenting what happened last week.

You can't clear roadblocks if you're stuck in your third status meeting of the day.

You can't anticipate needs or generate new ideas if you're staring at a screen updating who said what in yesterday's standup.

So if AI can handle status updates, meeting notes, risk logs, timeline adjustments, and all that other stuff—what's our actual job?

Let's talk about it.

We Make Dreams Come True (But What Does That Really Mean?)

I've said it before: Project managers are dream translators. But what does that actually look like in practice?

Here are the four things PMs do that AI will never replace:

1. Sense Making

We translate chaos into clarity for stakeholders.

Think about it: You've got twenty moving pieces, five competing priorities, three sponsors who all want different things, and a team trying to build something nobody's fully defined yet.

Your job is to be the centralized storyteller.

You take an abstract idea and turn it into concrete requirements. You hear a technical problem and frame it as a business value problem so a sponsor can actually make a decision.

AI can summarize. But it can't make sense of the why behind the chaos.

2. Decision Architecture

We design how decisions get made and by whom.

I don't know if anyone's ever said it that way before, but think about how often people try to force project managers to make decisions. That's not what we do.

Nowhere in the PMBOK does it say the project manager makes the decision.

What we do is:

• Analyze options
• Present trade-offs
• Go for decision
• Enforce the decision that was made

We don't own the decision. We own the process that gets to the right decision at the right time with the right people in the room.

AI can suggest options. But it can't read the room when a sponsor is about to overrule their entire leadership team because of a personal bias.

3. Energy Management

We keep teams motivated and aligned on what matters.

This isn't scope control. It's energy regulation.

Have you ever walked into a room where people are proactive, enjoying each other, having fun, being creative? You can feel that momentum. That's positive energy driving innovation.

I've also walked into rooms where you literally can't walk in because it's so tense. The air is frigid. People are angry. You can feel the toxicity.

We regulate that energy. We create the conditions where innovation can thrive.

AI can tell you morale is low. But it can't facilitate the conversation that gets the team back on track.

4. Conflict Resolution

We navigate competing priorities and personalities.

I once worked on a project where the CIO, CTO, CFO, and CEO all came to me separately and told me not to listen to the other three—just them.

You can't automate your way out of that.

AI can highlight the conflict. But it can't build the trust or influence the relationships that resolve it.

---

So here's the mantra for this episode:

AI Handles the Transactional. PMs Handle the Transformational.

AI does the reports, tracking, transcripts, and data analysis.

PMs do the vision, relationships, judgment calls, and influence.

That's the augmentation model. And if you get that balance right, you're not just surviving the AI era—you're thriving in it.

Today's Focus: Use AI to Surface What Actually Needs Your Attention

If AI is going to free up 60% of your time, you need to spend that time on high-value work.

So today, I'm giving you three prompts designed to help you think like a strategist, not a secretary:

1. Project Health Diagnostic (your non-negotiable experiment)
2. Decision Architecture Mapper
3. Energy Audit

I ran all three live in ChatGPT and Claude. And honestly? Claude surprised me with how good its coaching was in this round.

Let me show you.

---

Prompt 1: Project Health Diagnostic (The Non-Negotiable)

This is your weekly experiment. I want you to run this on a real project.

Here's the prompt:

---

You are an experienced PMO director reviewing project health.

First, ask me 3–5 clarifying questions about the project's current state, team dynamics, and stakeholder concerns.

Then, using my answers, provide a diagnostic report that answers:

1. What are the top 3 risks that need PM attention right now?
2. Which stakeholder relationships need strengthening and why?
3. What decision is being avoided or delayed that's creating drag?
4. Where is the team's energy or morale most vulnerable?
5. What one action would have the highest impact on project momentum this week?

Current project context: [Enter Context Here]

---

What I Tested It On

I used our fictional Social Wishing app project (the bucket-list social network from Episode 1).

Here's the context I gave it:

• Month 2 of development
• 5 engineers, 2 designers, 1 QA
• Sprint velocity dropped 20% in the last two sprints
• Daily standups feel perfunctory
• Sponsor keeps asking about launch date, but MVP scope isn't finalized
• One senior engineer just gave notice
• 4 months and $200K left in budget

Both ChatGPT and Claude asked clarifying questions:

ChatGPT asked:

• Who owns MVP scope decisions?
• Are stories entering sprint fully defined or being clarified mid-sprint?
• Has the sponsor agreed on success criteria for launch?

Claude asked:

• Is the MVP not finalized because of genuine disagreement or because no one's forcing the decision?
• Do you know why the senior engineer is leaving?
• When stand-ups feel perfunctory, what's actually happening?

Here's Where Claude Crushed It

ChatGPT gave me a solid diagnostic. Top risks: uncontrolled MVP expansion, architectural fragility, team disengagement. Good stuff.

But Claude gave me coaching.

Check out this section on stakeholder relationships:

"The sponsor—urgently, but carefully.

Core problem here is a shared reality gap. They're operating in a world where their feature ideas are refinements. You're operating in a world where those are scope additions with real cost.

This isn't malicious. Sponsors often genuinely don't see the downstream impact of small additions.

What you need is a direct, structured conversation that makes trade-offs visible: Here's what's in MVP. Here's what each addition costs in time and money. Here's what we must cut to accommodate it.

The goal isn't to win an argument. It's to get them into the same decision-making frame as the rest of the team."

That's the kind of coaching I'd give a junior PM.

It's not just identifying the problem. It's framing how to talk about it in a way that doesn't create defensiveness.

And then it ended with this:

"The good news is none of this is fatal yet. Month 2, budget intact, and a team that's disengaged but not yet departed is recoverable. The window is short, but it's open."

That's empathy. That's perspective. That's what a senior PMO director would say.

Both tools recommended the same highest-impact action: Run a 90-minute MVP reset session with the sponsor and core team.

And both gave me an agenda for it.

---

Prompt 2: Decision Architecture Mapper

This one's for when decisions aren't getting made—or they're getting made by the wrong people.

Here's the prompt:

---

You are a project governance consultant.

First, ask me 2–3 questions about how decisions are currently being made on this project.

Then help me design better decision architecture by answering:

1. What categories of decisions exist on this project? (strategic, tactical, technical, resource, scope)
2. For each category, who should make the decision, who should be consulted, and who just needs to be informed?
3. What decision rights are currently unclear or creating bottlenecks?
4. What decision-making process should I establish for the most critical category?

Project context: [Enter Context Here]

---

What Happened

I told it:

• Sponsor requests features mid-sprint and goes directly to developers
• Senior engineer (who just left) had final say on architecture
• Sponsor is also acting as product owner

ChatGPT told me I had a "governance vacuum."

Claude built me a full RACI chart and told me:

"The critical shift here is that scope moves to you. With a fixed budget and no senior engineer buffer, ungoverned scope is an existential threat."

Then it gave me the script for the hardest conversation:

"This isn't designing the architecture. It's the single conversation where you tell the sponsor that scope authority is shifting to you."

Again—coaching, not just output.

---

Prompt 3: Energy Audit

This one's for when something feels off with the team, but you can't quite articulate it.

Here's the prompt:

---

You are a team dynamics coach working with project managers.

Ask me 3–4 questions about team morale, motivation, and alignment with project goals.

Then provide an analysis answering:

1. Where is team energy highest right now? (What's motivating them?)
2. Where is energy lowest? (What's draining them?)
3. What misalignment exists between team priorities and leadership priorities?
4. What story or narrative could I reinforce to re-energize and re-align the team?
5. What one conversation should I have with the team this week?

Team context: [Enter Context Here]

---

What I Learned

Claude asked me:

• "When the team talks about Social Wishing, what specifically lights them up?"
• "How did the team react to the senior engineer's resignation?"
• "Outside of stand-ups, where's the real conversation actually happening?"

That last question is phenomenal. Is it Slack? Hallway conversations? Lunch groups?

Because if the real conversation isn't happening in your ceremonies, you've got a trust problem.

Claude's narrative recommendation:

"We're two months in, and we just learned something expensive but valuable. The project needed better architecture before it could move. We're not behind. We're resetting with clear rules. People still here chose to stay. Now we're going to build the thing we actually believe in—with scope we can defend—and ship something we're proud of."

I would use that exact language in a team meeting.

---

Your Non-Negotiable Experiment This Week

Run the Project Health Diagnostic on one current project.

Use transcripts from your last few team meetings. Use emails. Use status reports. Give it real context.

Then act on at least one insight it surfaces.

Here's what I want you to notice:

1. Did AI surface something you were avoiding or hadn't articulated yet?
2. How much time did you save by having AI organize your thinking?
3. How much faster could you move if you had this clarity every week?

None of this is about status reports or meeting minutes.

This is about surfacing what actually needs your attention so you can spend your time on the transformational work that only you can do.

---

The Takeaway

Project managers aren't project secretaries.

We're sense makers. Decision architects. Energy managers. Conflict navigators.

AI should handle the busy work so we can focus on the stuff that actually moves the needle.

So stop spending 60% of your week looking backwards.

Start spending 80% of your week looking forwards.

Because that's where dreams come true.

---

If you would like to see the podcast live, check out this link: https://youtu.be/3sE6ynNyjIg

Next time: We're talking about the Halo Effect—why we keep promoting the wrong people into project management—and how AI can help you become the transformational leader your team actually needs.

Now go run that diagnostic. Your project is waiting.