Say Yes To The Mess Part 2: Frank Barrett On Improvisation, AI, and Leading Without A Script

Frank Barrett, guest · [02:31]so you know where this is going. Yeah, when he, it was in the 1960s, he was already famous. he'd been playing with Miles Davis and he played with Charlie Parker and John Coltrane and everybody knew who he was, but he quit playing because he was tired of hearing himself play. So it's sort of that same thing of habits, routines, take on a life of their own such that You're no longer playing them, they're playing you. And so he, he quit playing for a while, and, no one knew whether he got sick or whether he had a drug problem, they didn't know, but he went underneath the Brooklyn Bridge in New York every day and started to play his horn, and every time he played something that sounded familiar, he'd stop, and then start again, and just to play something fresh, and if it sounded familiar, he'd stop. So he practiced every moment at trying to play in a way that was fresh. And he came out with a new album a couple of years later. At the time when it came out, it was not understood by critics or fans. They didn't, it sort of got panned at first. And now it's on just about everyone's list of the, one of the best jazz albums ever recorded. And the name of it is The Bridge. It's named after his experience there at The Bridge. And there's a movement right now to rename The Brooklyn Bridge, The Sonny Rollins Bridge. But that's an example of, in a disciplined way, becoming suspicious of what becomes automatic.

Frank Barrett, guest · [04:35]Well, I've talked to, I've mentioned two of them already. The first one, the Mastering the Art of Unlearning is, you know, I've kind of said that. It's, ways to trick yourself into having to pay attention in fresh ways, following Sonny Rollins' model. And the second point, as I mentioned, is, this notion of affirmative competence. Say yes to the mess. That's where the title of the book comes from. That say yes, it's a very hopeful enterprise that you say yes to it's emerging with no guarantee that it's going to be the right thing. And it really is very, it's, it's sort of similar to the Challenge that executives face. I mean, you're in the, you're in public, there's lots of, information flowing by. It's highly equivocal information. Equivocal is different from ambiguous. Ambiguous means it's not clear. Equivocal means there's many possible interpretations with no guarantee that what you choose is going to be the right thing, which is kind of what executives are facing now, that you commit to a way of going forward with no guarantee it's the right thing. So there's a leap, and in jazz we do that too, and it's in public, so you can't take it back. And, and notion of a, the third point is performing and experimenting simultaneously, which means whenever there's going to be experimentation, there's clearly going to be unintended things happening, meaning there's going to be mistakes. So you have to embrace errors as a source of learning. Some of the greatest inventions in the world came about accidentally through, and we all know stories, the invention of penicillin. People know that story, the invention of the post-it notes from 3M. There's lots of famous stories of that. And it's true in jazz too. Jazz musicians make mistakes all the time, but we just treat it like another note that perhaps suggests another possibility, another possible way of moving forward, and it opens up new worlds. So It's there that, you want to make sure that the voice of shame isn't winning when you make a mistake, because if shame takes over, it drives out improvisation. You can't feel shame and improvise. The shame voice is interesting because it's the research on shame is having a sort of renaissance. I used to think that shame is not an emotion because it's not tied to activation, because emotions are supposed to be tied to activation. You know, you feel love, you want to be drawn towards someone, you feel fear, you want to away. That's the role of emotion is to activate. But shame was not tied to activation, and they finally realized it's the only emotion that's not tied to activity because you freeze. You're frozen when you feel shame. They call it the shame freeze, and that's why it drives out improvisation. So when you make mistakes, you got to have some voice. That is very forgiving, forgiving of yourself, forgiving of others around you. That's what makes it work. Also in comedy, improvisation troops, they're very devoted to that notion is you, you don't shame someone for making a mistake. In fact, you take it as possible pathway forward.

Frank Barrett, guest · [07:38]The fourth point is a minimal structure, maximal autonomy. So there is structure in jazz, but it's not, you know, you don't just pull out notes out of thin air, but it's just the right amount. It's also minimal consensus and maximal autonomy. We talk about the importance of consensus all the time, and there is agreement, but it's too much agreement in jazz is just as dangerous as no agreement. So you just have enough agreement about just the right things and nothing more. And basically it's the chord changes of a song through time that allows people to coordinate and update each other as to where you are in the flow of the song, and that's what makes it work. The other basic agreement that's non-negotiable is the rhythm. The rhythm is non-negotiable. The beat is non-negotiable, but you can do lots of things around it. You know, you can play polyrhythms, you could accent different parts, but when you have the non-negotiable minimal structure, it creates freedom. It sounds ironic, but it's, that's what it is. So there's some constraint, but a lot of freedom.

Daniel Atlin, host · [08:36]And that's similar to management as well, I guess, right? You empower people, but you give them a, at least the beat, whatever, you know, the high level thing that you're supposed to be focused on, but give them the, the ability to actually do it in their own way as well.

Frank Barrett, guest · [08:47]That's right. It's sort of a kind of guided autonomy in organizations. In the military, they had this phrase that's somewhat similar called commander's intent. So once you know the intent of the commander, you should act on it, and he's probably, he or she is probably not going to tell you how to act exactly in this situation. But if you embody their intent, and you know their intent well enough, you can make guesses as to what's appropriate in that. That's a kind of a minimal structure.

Frank Barrett, guest · [11:42]Because all there's so much passion around leadership, all these books about leadership. Great leaders, and for a while it was Jack Welch was the ideal leader, and then it was Steve Jobs was the ideal leader, and for a while it was Elon Musk. And it's, the downside is when we valorize leaders, we forget about this other thing. Who talks about followership? It never gets talked about because you think it's shameful. What, you want me to raise my son to be a good follower? And it's because we have a dearth of vocabulary. We don't have a vocabulary to talk about followership as a noble calling. And in jazz, we do. There's a way to talk about followership as noble, and we call it comping, which is short for a company. So if you're playing piano and the saxophone is soloist, your whole mission in life is to make him brilliant, make him or her brilliant. So you listen well, you feed ideas, you get out of the way, you support their thinking out loud. We could learn that lesson in, in, organizations as well. You know, imagine doing evaluations and evaluating someone based on when were you a catalyst for others to come up with good ideas? When did you support someone on a project even though you didn't get a credit? When did you ask good questions that led to new insight just by your attention? Those are all what we might call followership behavior. They're supportive, they're sort of catalytic behaviors. Those are the kind of things jazz musicians do for each other. You got a solo, I mean, you take your turn soloing, but if you're just a good soloist and you're not good at comping, No one wants to play with you.

Frank Barrett, guest · [17:17]So what I meant by double vision is you see what a person's capable of, even though they're not acting that way now. That's what I meant by double vision. It's almost a mystical quality. I've gotten interested in reading mystics lately. If you have a hammer, every place you look is a nail. I start to see this theme that you can actually witness emergence in real time. And one way to think about it, I use the phrase double vision, is when Miles was disrupting, creating provocative competence, if you had asked musicians in advance, what do you think about playing in these strange forms? They would have said, no, I'm not going to show up that day. But Miles knew they were capable of it better than they knew it. It's interesting to keep a vision of what someone's capable of. Even when they're not living up to it. And here's one model. I think of mothers of adolescent sons, my mother in particular, my parent, my father too, but mothers for some reason are more figural for me around this. Teenagers, adolescents are self-absorbed, think only about, you know, what's in front of their nose, in front of their face. And a good parent figure, a good mentor will see what that person's capable of way before the person shows any sign of it. That's double vision. That's kind of what I have in mind. You have to have vision, you also have to have sight. So you see what's actually happening. You can give feedback to the real event, but to keep both in mind at the same time.

Frank Barrett, guest · [22:27]Interesting how it's, what's going to happen and what it might mean. It could be a gift, can offer patterns and insights. I think we're talking about universities here. It's awful for an undergraduate to use AI. It's awful for a high school student to use AI. You can't offload the foundational work that's needed to understand, develop concepts, to wrestle with ideas, to read, the linear reading in a page, across a line. It's not appropriate. Go to it to do your work for you. But there is ways it's magical. Like I, I had several papers I've written for different calls and different actions on different topics. I went, I went back to get a master's in divinity at Harvard. I took time out because I was so tired of what I was teaching. I was so stale. I provided my own provocative competence as a full professor doing the same thing over and over again. So I took a leave of absence for three years and went and got a master's in divinity. Did it increase my salary? No, but it gave me a new richness. I read all kinds of things I'd never read before. And I wrote all these papers on different topics in different areas. And, and some of them had nothing to do with each other. And just for fun, I put five of the papers into Claude and said, what, what do you see? What themes do you see? And it gave, it saw patterns I didn't see. It saw themes of a possible, it could be a possible book. Connections I hadn't noticed. Now, the fact that I noticed I was able to appreciate those connections is an example of compressed expertise, because I had been so deeply into each one of those papers, each of those knowledge areas, that I noticed a breakthrough. I could see where they saw a breakthrough that I didn't notice. Seeing context in that way, recontextualizing, AI could be a gift. But if I used it to write those papers, I would have cheated myself. There's no way to learn without putting yourself at risk. Deep learning involves risk. There's no other way around it. And leaping out and making guesses about what things mean, that work, we can't offload that to AI. That's what I'll say about it.