🚩 Tactical Memo 059: The Day Your Team Stops Asking

In partnership with

Want to get the most out of ChatGPT?

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide

Read time: 7 minutes

👋 Hey, it's Justin. Welcome to Tactical Memo, my weekly newsletter for leaders who plan to win the next decade of work.

For more, check out my courses, workshops, and webinars here.

👉 Why Read This Edition: You'll learn how to spot the moment your team has quietly stopped trusting you, and what to do before the silence costs you the project.

The thread that died

Leena Rinne, the VP of leadership at Skillsoft, told Fortune in March about an employee who came to her with a strange complaint.

The employee had received a Slack message from her boss, didn't understand it, and suspected it was written by AI. So she ran it through her own AI tool to interpret. The AI then offered to draft a reply.

She paused and told Rinne, "I literally think my boss's AI is talking to my AI. That is the actual conversation happening right now." Rinne calls it social offloading. When the human work of leading, the part that takes judgment, empathy, and a little courage, gets handed to a machine. I've watched this play out from a different angle. The boss isn't using AI to write the message. The boss is using AI to make the call. FortuneHR Magazine

And the team can tell.

Here's how it usually starts.

Six months in

A senior leader I'll call Mark runs a 40-person product org at a mid-sized SaaS company. Smart guy. PhD. Early adopter.

About six months ago he started running every meaningful decision through ChatGPT before bringing it to his team.

Pricing changes. Org redesigns. Hiring debates. Roadmap trade-offs.

The team noticed. Of course they noticed.

Someone would raise a problem in the team Slack channel. Mark would say, "Let me think on this." Twenty minutes later he'd come back with a fully formed answer. Bullet points. Pros and cons. A clean recommendation.

It read like a strategy memo. It didn't read like Mark.

The Tuesday afternoon

The moment I'm thinking of started in a Slack thread on a Tuesday.

A senior PM on the team flagged a real problem. A vendor was three weeks behind on a critical integration. The launch date was at risk. The PM laid it out cleanly, named two options, and asked Mark which way he wanted to go.

Mark read it. He copied the thread into ChatGPT. He came back forty minutes later with a five-paragraph response that neatly weighed both options, surfaced a third the PM hadn't considered, and told the PM to "align with the vendor and re-baseline the schedule."

The PM thanked him. Said he'd run with it. Closed the thread.

Then he went quiet.

Not loud quiet. Not visible quiet. The kind of quiet a senior leader misses if they're not paying attention.

The next time a problem came up, the PM didn't bring it to Mark. He brought it to the VP of engineering. The time after that, he handled it himself and told Mark about it after the fact.

Within a month, three other people on the team had done the same thing.

Mark's Slack DMs got quieter. His 1:1s got shorter. His team meetings turned into status updates. He thought he was running a smoother org. He was actually running a colder one.

What the team was telling him

When I talked to one of those PMs months later, here's what he said.

"I stopped bringing him problems because I knew what was going to happen. He was going to disappear for thirty minutes, come back with something polished, and it was going to feel like I'd just been processed."

That's the line that matters.

Processed.

People can feel the difference between a leader who is wrestling with their problem and a leader who is laundering it through a model.

When Mark answered fast and rough, with his actual voice, the team trusted him. When Mark answered slow and polished, they stopped.

Andy Grove had a line in High Output Management about how a manager's real output is the output of their team. If the team stops bringing you the hard stuff, your output drops to zero. You just don't see it for a while because everything looks calm on the surface.

Calm is what it looks like when people have given up on you.

Why this is the real AI leadership question

There's a version of "leading in the AI era" that gets thrown around in LinkedIn posts. Use more tools. Automate more tasks. Build the best stack.

That's the easy version. It's also the wrong one.

The harder version is this. AI now writes a better-sounding answer than you can in most situations. It will be more organized. It will sound smarter. It will cover more bases.

And the second your team realizes that's what they're getting from you, your authority erodes.

Not because the answer was bad. Because the answer wasn't yours.

Trust at work has never been about having the smartest answer in the room. It's about a person standing behind a call. Owning it. Taking the hit if it goes sideways. Saying "I decided this" instead of "the analysis suggests this."

A model can't do that part. A model can't be on the hook.

When a leader hides behind AI output, they're telling the team something whether they mean to or not. They're saying, "I don't want to be on the hook for this."

The team hears it. Every time.

What to do this week

I'm not telling you to stop using AI. I use it every day. So should you.

The line is about what part of the job you let it do.

Use AI to think. Use it to pressure-test. Use it to surface the option you missed. Use it to draft the boring parts of the memo so you have time for the parts that matter.

Don't use it to make the call. Don't use it to write the message that should sound like you. Don't use it to deliver the answer to the person who came to you scared.

Here's the diagnostic. Pull up the last five non-trivial decisions you communicated to your team. Read them out loud.

Ask yourself four questions.

1. Does this sound like me, or does it sound like a model?

2. Did I name the trade-off I was actually worried about, or did I list every trade-off cleanly?

3. Did I tell them what I decided, or did I tell them what the analysis suggested?

4. If this call goes wrong, is it clear I'll own it?

If three or more of those answers are bad, your team already knows. They've been adjusting around you for weeks.

Fix it the same way you broke it. Quietly and consistently.

This week, take one decision you'd normally polish through ChatGPT and don't. Write the response yourself. Keep it short. Keep your voice in it. Tell them what you decided and why. If you're uncertain, say you're uncertain. If it's a hard call, say it's a hard call.

The people on your team don't need a better answer. They need to know there's a person on the other end.

One question to sit with

When was the last time someone on your team brought you a real problem before they had a clean answer for it?

If you can't remember, that's the data.

If this helped you, send it to someone who could use it this week.

Until next time,

Justin

✍️ From the Desk of Justin Bateh, PhD
Become an indispensable AI-era leader