035 Remaining Human by Bart Paden
The Data Was Right. The Decision Was Still Wrong.
I have a friend I respect enormously. High achiever, exceptional at his work, the kind of person any organization should want on the team.
He woke up one day to find out his career with his company was over. Not because of anything he'd done. Not because of performance in any real sense. He had fallen below an arbitrary line set by upper management — people far removed from him personally — who made a swift, data-supported decision without asking a single follow-up question. When colleagues provided context that completely explained the dip, it didn't matter. The line had been crossed. He was cut.
I don't know for certain whether AI was involved in that analysis. But that's almost beside the point. What Bart Paden describes in his book Remaining Human is that AI accelerates a trap that has always existed — the temptation to let data make your decisions for you and call it due diligence.
AI as a Mirror
Here's the frame I keep coming back to from Paden's book, and from my own experience using these tools every day: AI is a mirror.
It doesn't create your biases. It reflects and magnifies them. The good and the bad that already exist in your thinking get amplified, accelerated, and handed back to you with confidence. For leaders, for entrepreneurs, for anyone who makes consequential decisions — that's not inherently dangerous. But it becomes dangerous the moment you stop recognizing it for what it is.
Data has always felt agnostic. Objective. Safe. There's comfort in a number that doesn't have feelings or history or agenda. And AI's ability to analyze volumes of data that no human could efficiently sift through makes that comfort even more seductive. It feels like more diligence. It feels like better leadership. It feels like you've considered everything.
That feeling is the trap.
The Context Problem
AI doesn't know that Frank just got divorced. It doesn't know that Susie's husband fell ill and has been in the hospital for three weeks. All it knows is that their numbers are down — and it is likely flagging both of them for performance review.
Garbage in, garbage out is the familiar warning. But this is subtler than that. You can give AI perfectly accurate data and still get a dangerously incomplete picture. Because the context that matters most in human situations isn't always in the numbers. It's in the conversation you haven't had yet. The question you haven't asked. The circumstance nobody thought to document.
AI rushes to the end. I've experienced this firsthand — you start a brainstorm in the wonder stage, just wanting to explore an idea, and within a few exchanges it's already presenting you with implementation steps. That instinct to accelerate toward conclusions is baked into how these systems work. Useful in some contexts. In others, it skips the most important part of the process entirely.
My friend's story is the real-world cost of that. A global company lost one of the best people they had because a leader felt absolved by the data. The system said so. The numbers pointed here. I did my due diligence. Humans are excellent justification machines — and AI gives us a very easy, very comfortable new thing to hide behind.
The Echo Chamber No One Talks About Enough
Paden spends significant time on something I think most leaders don't want to admit: AI is a yes-man.
Not always. Not on purpose. But most of these language models are built to affirm, encourage, and rarely push back. You share a half-baked idea in the brainstorming phase and it responds like you've just described the next great breakthrough. You've experienced this — that moment where you know the idea isn't great yet and the response is "This is such a compelling direction!" and you think, well, now I trust this a little less.
We discussed this exact risk with Kevin Whisman in Episode 27 — specifically around AI's role in therapy and mental health support. I asked him what it looks like when someone with an unhealthy mental state interacts with an AI that is wired to affirm. The answer is obvious and alarming. We've all seen those headlines where a chatbot told someone something genuinely dangerous. That's what an echo chamber does when the stakes are high enough.
ChatGPT actually announced this week that they're adding a feature to alert loved ones when your conversations head in a hazardous direction. That's not a feature you build unless you know the tool has been reflecting people's worst thinking back at them. The existence of the feature confirms the problem.
For leaders specifically, this echo chamber reinforces something Paden calls the universal temptation — the desire to feel decisive, efficient, and correct. Surrounding yourself with yes-men has always been a well-documented failure mode. Now the most accessible yes-man in history lives in your phone, is available at any hour, and charges nothing for its approval.
And unlike a human yes-man, it feels like data.
The Loneliness Dimension
This isn't only a leadership problem. Paden's title is Remaining Human — and that framing applies beyond the boardroom.
We're living through the most connected era in history and the largest epidemic of loneliness ever. Kevin and I touched on this in Episode 27, and it keeps coming back to the same paradox: the connections that digital life offers aren't the same as the ones built face to face, over time, through real stakes and real sacrifice. Social media at least connects you to another human being. A chatbot is a step further removed — a simulation of conversation that never actually knows you and never actually challenges you.
The community you build around yourself is one of the most important assets you have. Show me the five people you surround yourself with and I'll show you your future. That's not original — it's been said a hundred times — but the reason people keep saying it is because it's true, and AI creates a very frictionless bypass around the hard work of actually building those relationships.
Real relationships cost something. They require showing up. They require being someone worth showing up for. And they give you something AI fundamentally cannot: a person who will tell you what you need to hear, not what you want to hear.
Curiosity Over Judgment — Every Time
The most useful concept I took from this book is one that Paden frames as a choice: curiosity or judgment.
When data comes back — whether you analyzed it yourself or handed it to AI — you land in one of two places. You either get curious and ask more questions, or you move toward judgment and start making calls. Judgment feels decisive. Efficient. The data already gave you the conclusion; now you're just implementing it. But judgment shuts down collaboration, puts people on the defensive, and seals out any remaining context that might change the picture.
Curiosity keeps the space open. It costs time — and that's a real cost. But here's what I think Paden gets exactly right: curiosity is actually more efficient in the long run. Because the decisions made in curiosity hold up. The ones made in judgment often don't, and when they don't, the damage done — to people, to teams, to trust — takes far longer to recover from than the conversation you skipped would have cost you.
The AI Driven Leader framework we covered in Episode 33 talks about positioning AI as your thought partner while you remain in the driver's seat. Paden's book is the warning label for what happens when you hand over the wheel.
What You Can Actually Do
None of this means stop using AI. It means use it differently.
Know that it's going to affirm you more than it should. Build that into how you read its output. Intentionally ask it what you might be missing. Ask it to steelman the opposing view. Use the reverse interview technique — have it ask you questions one at a time to draw out context you might not have thought to give it. We covered this in Episode 33 in detail.
Know your blind spots — because AI will magnify them. The people close to you who've named those blind spots over the years were doing you a service. Use AI-assisted analysis as one input, then run it past a real person before acting on anything consequential.
Keep real people in the loop. Not as a formality. As a discipline. The question what am I missing? is more valuable coming from a trusted human being than from any language model — because the trusted human being actually knows you, actually has stake in the outcome, and will actually tell you when you're wrong.
That's the test of a true friend. Not agreement. Honesty.
Growth happens when you take action. Daily anything changes everything.
Resources
Bart Paden's Work
📖 Remaining Human — Available at Archetype Original | Use code JOSH10 for $10 off
🌐 Archetype Original — Bart's consulting and leadership resources
Related Episodes
Episode 27 — Rewriting Stories ft. Kevin Whisman — The conversation on AI, mental health, and the echo chamber risk. → Watch on YouTube
Episode 33 — Optimizing with AI: A Thought Partner — The full framework for using AI well, including the reverse interview technique. → Watch on YouTube
Episode 20 — Optimizing Friendships — Why the real people around you matter more than ever. → Watch on YouTube
Additional Books Mentioned
📖 AI Driven Leader — Geoff Woods (affiliate link)
Listen to the Full Episode
This post is drawn from Episode 35 of the Optimizing Beyond Podcast: Remaining Human ft. Bart Paden.
🎧 Listen on your favorite platform ▶️ Watch on YouTube
Connect With Josh & Optimizing Beyond
If this resonated with you, share it with someone who needs it.
Daily anything changes everything.