246. Shared Wisdom: How Communication Defines Culture and Builds Community

[00:00:00] Matt Abrahams: Community and communication not only have the same root word, but they work together to make us more effective. My name is Matt Abrahams and I teach strategic communication at Stanford Graduate School of Business. Welcome to Think Fast Talk Smart, the podcast. Today I'm excited to have a conversation with Sandy Pentland. Sandy is a professor at the Massachusetts Institute of Technology, where he helped create and direct the MIT Media lab. He's also a Stanford Institute of Human-Centered AI Fellow. Sandy is recognized as a pioneer in computational social science and wearable computing. He has just released his newest book, Shared Wisdom: Cultural Evolution in the Age of AI. Well, welcome Sandy. I have long admired your work and I'm very excited for our conversation.

[00:00:49] Sandy Pentland: Well, I'm really happy to be here.

[00:00:50] Matt Abrahams: Shall we get started?

[00:00:51] Sandy Pentland: Yes, please.

[00:00:52] Matt Abrahams: Well, let's start with Honest Signals. That's where I first got to know you and your work. Can you define what honest signals are and how you studied them? And then can you share how mastery of these signals can actually be effective tools to help us build trust and foster relationships, more than words might even be able to? 

[00:01:09] Sandy Pentland: A way to think about this is that if you look at our closest ancestors, big, great apes, they actually communicate with hoots and pants and other sorts of sounds. Language seems to be built on top of that. But we didn't get rid of it. We kept it at its contrast to the words. So you can tell when someone's excited, of course, because they get all, oh they, that's a signal, right? And you can tell when they're not paying attention because there's these sort of awkward pauses sometimes. And those are the things that are the signals about what's going on in the person's head and about the relationship. And the ones that are most interesting, I think to me, and perhaps to your question, are the ones about the patterns of interaction. So it's not just about the person, it's about how they interact with other people.

[00:01:51] And we find that the pattern of people using each other, communicating with each other for help, both ways, is almost a perfect predictor of trust. Which is really interesting. So it's a relationship where you give me something and I give you something back, and we do that for a while. It also is something that's very closely related to friendship. People who do that, say about out work things, tend to become friends. And not just occasionally, very regularly. And if you look at groups of people, you see these same sorts of patterns define who's interested in a topic and who's not. In other words, what's the community? And so this is not something we're generally aware of, but you can look at these patterns and once you begin thinking about them, you can say, yeah, that makes sense. This guy must be interested in it and he gets something valuable out of it. So yeah, there's a certain level of trust there. 

[00:02:45] Matt Abrahams: What I found really interesting is how you actually looked at these patterns. How did you study and look at the reciprocity that you found? 

[00:02:51] Sandy Pentland: Oh, we've done it a number of ways. The one that people always remember is we built these little badges that would wear around your neck, and they had a microphone that didn't record the words, but did record how you moved and where you were and your tone of voice. And that turned out to be very informative about what was going on. For instance, who was leading a group, who was dissatisfied with it? And that's what the Honest Signals book is all about. 

[00:03:18] Matt Abrahams: Did you find that there are certain signals that are more likely or more effective in determining power in relationships than others? So is it vocal intonation versus gestures versus physical orientation? 

[00:03:31] Sandy Pentland: I think that's not the right way to think about it. What it is, is certain patterns of interaction, and those are accented by excitement or delay. In fact, we built a little, uh, sort of AI tool, which is actually in most call centers, and what it does is it tells the person when to listen, when to shut up, when to be excited, when to be a little cautious about it. What it does is it gets rid of the call center people fighting with the customers, so everybody has a better time. 

[00:03:59] Matt Abrahams: What I've taken away from the Honest Signals work that you've done is that observing the patterns can be very insightful to what the results will be, and many of us don't pay attention to those patterns.

[00:04:11] Sandy Pentland: No. 'Cause we're focused on the words or we're thinking about something else. Yeah.

[00:04:15] Matt Abrahams: I'm curious, now that we've shifted more to remote and hybrid work, what are your thoughts about that impact on Honest Signals and are there new things that we should develop and hone to help maintain stronger ties now that we're doing a lot of this remotely.

[00:04:29] Sandy Pentland: The obvious thing about say remote things and Zoom and all that is that it gets rid of a lot of the body language. The tone of voice is diminished. You can't tell about a lot of things. What I try to do with these things is have casual interactions, and in fact, we've built a little tool to help you do that. It's all free, all that. It's called Deliberation.io. It lets people talk about things in a sort of less structured way. And prevents overly large voices and gives people sort of summaries of what's going on. But the real thing is you have to have a personal connection that goes with these patterns.

[00:05:04] So when I do Zoom calls, I almost always start with five minutes of how are the kids doing and, you know, what's the weather, so that there's a sense that people actually care about the human and not just about the work. That's not the best thing because it's so much better to have that water cooler conversation or the little conversation after the meeting or before the meeting. Those are when things really get established in terms of trust and alignment and things like that. 

[00:05:32] Matt Abrahams: It's really interesting 'cause when you and I were walking over to the studio we're in today, you did just that. We had this small talk to sort of get connected and it's very useful. You've pioneered the use of data to map social networks. How can this data be used to identify communication bottlenecks in a company or help create more effective and efficient teams once you understand the network? 

[00:05:53] Sandy Pentland: We started with these little things that recorded sort of where you were, the tone of voice, whether you were talking to people or stuff, and we began to see that there were these bottlenecks. You say, these people don't talk to those people, they're gonna have a hard time coordinating, right? But what you can do with these is you can look at patterns of Slack messages, look at email, look at things like that, and you can do a really good job of saying, well, this group talks to this group, but not to that group. And maybe that's a problem and you want to think about that. We are building things here, we call AI buddies. It's basically a replacement for the manual that you never read, or all the newsletters that you never read. But what it does is it informs you about what other people are doing that are relevant to you and what's going on in the company or the organization, just to give you more context.

[00:06:42] And more sort of social awareness, and that's a huge amount of it, particularly with remote work, with international organizations, you lose context, you're not really in the loop anymore, and that's something that AI's can do really easily and fairly safely. Because, you know, they're not likely to hallucinate things like that very much. It's sort of like an automated newsletter that's meant to the particular thing you're doing right now, and people love it. It's good. It helps people connect. It helps business leaders actually change the organization. I was talking to the chairman of a large consulting firm, and he was going to reorganize how he ran his three hundred and fifty thousand employees, so he accepted the idea that there'd be fewer people in office, but use these sort of AI buddy techniques to have more remote people, more in the loop, and also to bring in people that are sort of on particular gigs or projects and have the confidence that they'll know what to do because they have this customized news and reminder service.

[00:07:42] Matt Abrahams: I really like that idea of giving people the information they need when they need it, to fill in those gaps and give the context, and AI allows you to make it very specific. 

[00:07:50] Sandy Pentland: In many ways, not a new idea at all, but the new AI's are actually pretty good at this without all the concerns about hallucinations or whatever. And you don't need the frontier model to do this at all. You can do this with any of the open source ones. 

[00:08:04] Matt Abrahams: So something that's germane and available to everybody. In your book, Social Physics, you discuss how ideas flow through a group. What's the best way for a leader to seed a new idea in a team, and what communication patterns will help that idea gain momentum over time?

[00:08:20] Sandy Pentland: You told me about this question and I sort of don't like it, and the reason is, is this idea of the leader in control and the leader defining the language about it, and those are both mistakes. Because in a group, what you need to have is you need to have people understand why something happens and what are the consequences and what do other people think. The reason we have so much polarization in this country, turns out that people just don't know what the other people think. If you set up a situation that let them know that, then they become rather dramatically more aligned with each other. It's really rather shocking. So the same thing is true in companies and so forth.

[00:08:58] So what a leader should do is say, here's something I'd like to discuss. Here's this idea. Have people talk about it, comment about it. And that's why we built this thing, Deliberation.io. You can just use it, it's free, but what it does is it doesn't allow the loud voices to dominate. And what you wanna do with this is you want to look at what is the language people use. I did some stuff with some political survey people, and they were talking about inflation and stuff, but humans, most people don't talk about inflation. They talk about cost of living. Oh yeah, sure, they're related, but one is far more salient than the other. Much more connected. And so, you need to have your conversation in the terms that the people understand.

[00:09:41] You need to listen to them because then they feel like they've been heard and they actually have been heard because they understand more about it. They've influenced a little bit. And all of that is this notion of shared wisdom, which enables action by the community, right? So collective action. 'Cause that's what you really want. You don't really care about the ideas and the conversation. You want to actually do stuff. But that depends on having shared understanding, shared wisdom about what's the right thing to do. And that depends on people understanding each other. 

[00:10:12] Matt Abrahams: So it's not the leader coming in and saying, thou shalt work on this. It's really posing questions that bring forth what's salient and even linguistically, what those words are that people use to describe what's salient. 

[00:10:24] Sandy Pentland: Another sort of aspect of this is getting the right people in the room, the people that actually have skin in the game. So org charts, I have this joke is if you have an org chart, you have a map of how to have a stupid organization, because org charts don't reflect the sort of piece by piece, task by task type of thing that needs to happen. There's things that cut across all these different things. You need to have a way of getting those people discuss things in the same room and feel like they can really do it. It's intended to deal with some of that, not that it's perfect, but, but it gives you an idea of how you can actually build digital media that work for these sorts of things.

[00:11:00] Matt Abrahams: So using the tools to help. So you have to have the right people in the room. You have to use the right language, and you have to give people an opportunity to understand how others see it. And that's how you propagate ideas.

[00:11:11] Sandy Pentland: That's right. They have to sort of in their own heads say, oh yeah, that's why we're doing this.

[00:11:15] Matt Abrahams: You mentioned earlier that you study communities. I'm curious, what is the most surprising or counterintuitive finding your research has shown on how people interact and form communities? 

[00:11:25] Sandy Pentland: I think the main thing is how smart communities can be. There's this sort of general sort of jokes, you know, stupid communities, madness, you know, et cetera. But actually when we look at things like people making financial decisions, people making other sort of decisions, if you have good patterns of communication between people and they have skin in the game and they're actually focused on it, they're usually better than the math. Take something like economics or trading where you know you can do all this quantitative stuff. The community version of that, which you know, has people who know the math, but now they're also talking to each other, works better over the long term than the people who just use the math.

[00:12:05] And the reason is, is that the community version of it has a broader view of what are the risks and opportunities than the math does. So it's a sort of myopic view when you're just trying to solve the equations or engineer it. You need to ask, how does this fit with everything? Have things changed? What are the things that are coming up? And we see that again and again. So it's possible to have wisdom, which is a capacity to make good decisions as a group, that outshines the scientific things. Not ignoring them, but incorporating them in a broader context. And I think that's the thing that people don't know and don't respect. 

[00:12:43] Matt Abrahams: I wanna turn our attention to your new book. In your new book, Shared Wisdom, you state, and I'm quoting, it is important to remember that story sharing is as least as much for communities as it is for individuals. Can you help unpack this quote and discuss what story sharing does for us? 

[00:12:59] Sandy Pentland: So the way to think about human society, and we'll imagine that it's five hundred years ago or ten thousand years ago, is that there's a constant conversation between people about, oh, this thing's interesting. Oh, that's terrible. Why did you think of that? And that's a type of deliberation to establish community norms. We like this. This is a good thing. We don't like that. That's a bad thing, right? And it's not a formal thing like where we sit down, we're gonna have the rules. It's this community sense or wisdom about how things should operate. And so that's really critical that you have people understand that why things are this way and not that way.

[00:13:39] And had the stories of, oh yeah, so and so ate those berries and got deadly sick. Don't do it, right? The sharing of stories educates the community 'cause it's passed on. It helps them define their culture. And they're incredibly sticky, also. If you look at Australian aborigines, they have stories, these, uh, songs about where to go and what to do that are authenticated to be more than seven thousand years old. It's just incredible. And what that's doing is defining the worldview and culture of that group. And different groups have different cultures, different songs that they sing.

[00:14:17] Matt Abrahams: Storytelling is a great vehicle to get your information across, but you're taking it to a new level. That stories help a society, a culture, a community to get their ideas across. So it's not just about the individuals. 

[00:14:29] Sandy Pentland: Stories are the stuff of culture. You know, people talk about, oh, we need a good culture in this company, or whatever, right? That has to do with the stories you're telling each other. And it's not the official like newsletter or the CEO, it's the things that people tell each other. Usually before or after the meeting or, and the lunch room, something like that. And those stories permeate in various ways to be able to define the attitudes of people, which is the culture. And what we know also is that culture determines a great deal about the decisions we make and about the outcomes we get. So you really want it to get it to be right. And there's two ways. One is to make sure that good stories spread, but you also need to be able to have the stuff of stories to be able to spread. So you need to do things that are actually will help shape the community by saying, this is what we do. When somebody has a problem, we do this, and the conversation should end up being, yeah, we see why that's true. 

[00:15:26] Matt Abrahams: Lots of rich ideas there. One thing I'm taking away from what you just shared is, if I am in a leadership position, in an organization, in a group, one of the ways to really understand the culture is to listen to the stories that are being told. Versus taking a survey or saying the culture is this, because I have said it is that. So this notion of listening and observing can be really helpful. Before we end, I like to ask three questions of all my guests. One I make up just for you and two are similar across everybody. Are you up for that?

[00:15:54] Sandy Pentland: Yeah, sure.

[00:15:55] Matt Abrahams: You are knee deep in AI. I'm curious what, what most excites you about AI and where it's headed? 

[00:16:03] Sandy Pentland: The part that excites me most about it, first of all, I'm completely irritated by all the frontier model type of stuff, because that's not what people are gonna use. People are gonna use this much lighter weight specific type of stuff. I also don't like the sort of model where there's this big central company that runs everything and owns all your data. So I'm a real advocate of sort of personal AI tools to help people get along. Also for small businesses and so forth. I think that what we can get is a lot more of the things that we've been talking about. You can get things that are much more tuned to the preferences and needs of particular people in particular communities. And I think that will do an awful lot for making this more agile, for being able to address problems and feel a lot more like a member of the community that is doing something that we care about.

[00:16:52] Matt Abrahams: I really like that idea of using AI tools to help us even feel more tightly connected as part of a community. 

[00:16:57] Sandy Pentland: Yeah, so AI for conversations, AI for community, not AI for big brain super intelligences that order you around, right? 

[00:17:07] Matt Abrahams: Yeah, absolutely. Question number two, who is a communicator that you admire and why?

[00:17:12] Sandy Pentland: I like Steven Pinker. He has the courage to take things that everybody believes and aren't really necessarily exactly right and attack it sort of head on and with a lot of evidence and good argument. And at least what he does is even if you don't buy what his thesis is, he changes the conversation substantially and he picks things that are really important.

[00:17:35] Matt Abrahams: I love the fact that you and he are friends because you both do such interesting work. He was a guest on the show and really enjoyed talking to him about indirect communication. Question three, our final question. What are the first three ingredients that go into a successful communication recipe? 

[00:17:51] Sandy Pentland: I think that the things you need to do is first of all, know what people are talking about, right? So you need some way to be in the loop about what everybody's doing, and that requires avoiding dominance and status things, because I'm not gonna tell my boss or my boss's boss what I really think. So it has to be something where there's a certain level of anonymity, but also social consequence about it. And then you need to have a conversation about what should we do so that people understand what everybody's views are and why we're making certain choices. And then you need to also ask people, how are you going to help with doing this? What's your role? And I think that people are willing to do things.

[00:18:33] Matt Abrahams: Absolutely. So I heard awareness of what the issues are and trying to minimize the loud voices in the room, synthesizing that with the people, and then inquire into the action that people are willing to take. And I think those are really important ingredients if you wanna affect change, and your research certainly is affecting change and how people interact verbally and non-verbally. Thank you for sharing your wisdom with us today, and best of luck on the book Shared Wisdom. 

[00:19:02] Sandy Pentland: I just hope it has some effect on our institutions and the way we do things. So thank you very much for having me. 

[00:19:11] Matt Abrahams: Thank you for joining us for another episode of Think Fast Talk Smart, the podcast. To learn more about nonverbal communication, listen to episode 137 with Dana Carney. To learn more about communication networks, check out episode 65 with Michael Arena and Glenn Carroll. This episode was produced by Katherine Reed, Ryan Campos, and me, Matt Abrahams. Our music is from Floyd Wonder. With thanks to Podium Podcast Company. Please find us on YouTube and wherever you get your podcasts. Be sure to subscribe and rate us. Also follow us on LinkedIn, TikTok, and Instagram, and check out fastersmarter.io for deep dive videos, English language learning content, and our newsletter. Please consider our premium offering for extended Deep Thinks episodes, and much more at fastersmarter.io/premium.