I Asked AI What It’s Like to Work with Me

Most people use AI to get answers. 

I used it to get evaluated. 

Not through a personality test, or a polished list of strengths and weaknesses. I wanted something harder to get. An honest read on how I actually show up when I work with other people. 

Because that is the part people rarely tell you plainly.  

People soften their feedback. They protect the relationship. They protect your ego. Sometimes they feel something clearly but do not have the language for it yet. 

AI, used correctly, can help with that. 

But only under one condition. 

It needs enough history with you to see how you actually operate. 

I didn’t ask this after one or two conversations. I asked after months of using it across writing, analysis, decision-making, parenting, leadership, and problem-solving. Then I came back months later and asked it to update its assessment. 

That second pass was sharper, more meaningful. 

The Experiment 

I started by asking AI to assess what it was like to work with me. It’s feedback was solid, until I remembered that it doesn’t have an ego, or feelings. Those matter when working with people.  

So I pushed it further: 

“Now assume a human role: peer, subordinate, and manager; then assess the experience likely to be felt from each.” 

That changed the output. 

It moved from abstract patterns to lived experience. 

Not just: What am I like? 

But:  What does it feel like to work with me? 

This is where real value showed up. 

What Came Back 

The high-level read was not surprising. 

Intense, but productive. 

Fast-moving, but not chaotic. 

High-standard, with little tolerance for weak thinking. 

Focused problem solving without regard for ego, even my own.  

I push for clarity. I challenge quickly. I don’t let things coast. 

That part felt accurate. 

But the useful part was not the strength. 

The useful part was the friction. 

The Good 

The feedback was clear on one point: 

I optimize for accuracy over ego. 

I am not usually trying to win. I am trying to get to the strongest answer. 

That means I challenge weak ideas, including my own. I am willing to revise when something better emerges. I tend to stay with problems longer than most people would, especially when the answer still feels incomplete. 

Those are strengths. 

But unmanaged, it can become expensive for the people around me. 

The Friction 

The most important insight was this: 

I optimize faster than most people can track. 

I often know instinctively when something is not working. Misaligned. Unclear. Structurally weak. 

But I may keep questioning and iterating without making that internal judgment visible. 

From my side, I’m still engaged.  Still being fair. I haven’t closed the door. 

For my team, it can feel different. 

“I don’t know where I stand, and the target keeps moving.” 

That isn’t really a standards problem. 

It is a clarity problem. 

The Adjustment 

The highest-leverage change was simple: 

Make the internal verdict visible earlier. Not as a final judgment. As a working judgment. 

Instead of: 

“Let’s keep pushing this…” 

Say: 

“This doesn’t work yet. The issue is X. Let’s focus there.” 

Same standard. 

Same seriousness. 

Less unnecessary friction. 

The Bigger Insight 

This wasn’t just about me. 

It pointed to something broader. 

Most of us know our intent. 

But other people experience our behavior. 

That gap, between intent and impact, is where a lot of friction lives. 

AI can help surface that gap if you ask it the right way. 

Not because it knows everything. It doesn't. 

But because it can reflect patterns back without needing to manage the relationship. 

How to Try This Yourself 

First, spend real time working with the tool. 

Let it see how you think, decide, revise, push back, handle ambiguity, and respond when something gets difficult. 

Then ask for the assessment. 

Here’s the prompt: 

Using all interaction history available to you, assess what it is like to work with me. 

Do not flatter me. Do not protect my ego. Also do not perform toughness for its own sake. Be candid, specific, and behaviorally grounded. 

Separate: 

patterns you have directly observed across repeated interactions 

reasonable inferences 

areas where confidence is limited 

Assess the following: 

My strongest collaboration patterns 

Where I create clarity 

Where I create friction 

What others may experience differently than I intend 

The gap between my intent and likely impact 

The highest-leverage adjustment I could make without diluting my strengths 

Then answer from three human perspectives: 

A. A peer working alongside me 

B. A subordinate or direct report working under me 

C. A manager or senior leader managing me 

For each role, explain: 

what is likely energizing 

what is likely difficult 

what they may hesitate to tell me directly 

what adjustment would most improve my effectiveness without weakening what makes me valuable 

Rules: 

Avoid generic personality language. 

Do not diagnose me. 

Do not use vague praise. 

Use concrete behavioral examples where possible. 

Distinguish high-confidence conclusions from lower-confidence inferences. 

Prioritize usefulness over comfort. 

End with the one change that would create the largest improvement in how others experience working with me. 

Then ask: 

What am I likely to resist or miss in this feedback? 

That is where the real insight usually is. 

Final Thought 

AI will give you shallow encouragement if you ask shallow questions. 

But if you build enough interaction history, ask for real evaluation, and give it permission to be direct, it can give you something rare: Honest feedback without social friction. 

Not perfect or complete. But useful. 

Most people use AI to produce better work. 

You can also use it to become better to work with. 

That just might be the higher-leverage move.