What You Can Learn From Asking AI for Brutal Honesty About Yourself - with Bestselling Author Daniel Pink

**Daniel Pink** (0:00)
And I had to do a SWOT analysis of me as a person. And this is the one that hit me in the gut. Your greatest risk isn't failure. It's irrelevance.
If you keep playing it safe, you'll continue to sell books and give keynotes, but the cultural edge will slip to voices that are rougher, riskier, and hungrier. The danger isn't a crash.
It's a slow fade. And so that was like, whoa, like that could be right. Hi, my name is Daniel Pink. I'm the author of a bunch of books on science and business and creativity. And I am excited to talk about AI, my own personal blind spots, and the future of humanity, all in less than an hour.

**Jeremy Utley** (0:47)
I thought it'd be kind of fun to start here, Dan.
I had you on my show at Stanford, Masters of Creativity, to talk about your book Regret several years ago. And I started following you, getting your newsletter, watching your videos, which I love, by the way. And then the other day, you posted this video about mastering yourself with AI. And I realized there's this enormous space of opportunity for AI around self-knowledge. And so to me, I loved the video, by the way. I loved watching it. I loved how visceral it was. Clearly, you're getting kind of punched in the gut occasionally. And so I just thought, that's why I reached out and said, hey, Dan, would you want to talk about some of the stuff you do with AI? Because I think it's an underutilized capacity of AI to help us understand ourselves better. And so I'd love to just start there.
How have you worked with AI to understand yourself better?

**Daniel Pink** (1:40)
Well, a few things. I did that explicit experiment, Jeremy, where I essentially enlisted AI as a kind of brutally honest, trusted advisor and wrote a set of specific prompts, putting ChatGPT in that mode. And the reason I used ChatGPT is that ChatGPT had much more of my history than the other large language models did.
And I wanted to know things like some of the things that my friends' advisors maybe wouldn't tell me. So what are my blind spots? And one of my favorites was what do people say about me behind my back?

**Jeremy Utley** (2:21)
And so why do your friends and advisors not tell you this stuff?

**Daniel Pink** (2:25)
I think sometimes these kinds of things can be uncomfortable for them. And I think other times they don't necessarily see it.
And it's both. And I think those things are hard to disentangle. If I really like somebody, I might be less likely to see their flaws. Whereas at some level, your enemies know you better than your friends.
And I'm not positioning these large-language models as an enemy. But again, these models, they'll do what you tell them to do. You know? So their default mode is very sycophantic. But if you say stop being sycophantic, be an asshole instead. Right.

**Jeremy Utley** (3:03)
I think at some point you said gloves off, right? Gloves off. No, no.

**Daniel Pink** (3:08)
The model said that. I didn't say that.

**Jeremy Utley** (3:10)
Gloves off.

**Henrik Werdelin** (3:11)
What was the thing that people said behind your back? And how true was it?

**Daniel Pink** (3:15)
Well, you know what, guys? Let me pop this up on my screen here for a second. So I just want to make sure that I have everything accurate about what I actually asked it. I'll do the PDF with these prompts that people seem to really dig.

**Jeremy Utley** (3:27)
And we could link that as well in the show notes.

**Daniel Pink** (3:30)
Yeah, if you want, you guys can have it. I mean, it's boom.
We're in business now. All right, so hit me. What do you guys want to know here now?

**Henrik Werdelin** (3:38)
We want to hear what do people say about when you asked a GPD about what people said behind your back, what did it answer and how true do you think it was?

**Jeremy Utley** (3:47)
I mean, folks should go watch the whole video. Just as an aside, you should watch the whole video.
But for you, Dan, we're getting the behind the scenes. What really stuck?

**Daniel Pink** (3:54)
Okay, here we go. So the first question that I asked ChatGPT, and I ended up seeing some themes here, and often some uncomfortable themes. There were things that ChatGPT said about me that I actually disagreed with, not because they were some kind of narcissistic injury, some kind of threat to my self-esteem, but because I thought they were actually analytically wrong. There were other things that were like, oh, yeah, okay, that's fair, even though it was uncomfortable. And so the first question that I asked is really at the heart of this, which is sort of self-discovery, self-awareness. I said, what do you know about me that I don't know about myself? In other words, what are my blind spots?