Latest / Machine Dreams / FBI’s Former Lead Kidnapping Negotiator Chris Voss on AI, Negotiation, and the Human Advantage
Transcript
- speaker-1: Uh-huh. speaker-0: Mm-hmm. Negotiation has always been deeply human. We listen, we read emotion, we build trust, and we try to understand what the other person really, really wants. But we now seem to be moving into a world where AI does some part of that for us. The big question, however, is what happens in that world. This is Machine Dreams, where we speak with the people shaping AI about what actually works, what doesn't, and what the future of AI could look like. My name is Kalawale Samuel Adebayo, a technology analyst working at the intersection of AI, cybersecurity, business, and society. And of course, I'm not here alone. I'm joined, as always, by my co-host. speaker-2: Hi everyone, I'm Leia Stern, global communications strategist and former venture capitalist. Sam, how you doing? I hope you're not alone. speaker-0: I mean I I have you so so so I'm not alone. it's always good to be on the show with you to talk about some of these really important things in AI today. So speaker-2: Yeah. Excited. Me too. It's great to see you. Today we're joined by Chris Voss, who spent twenty-four years at the FBI and became its lead international kidnapping negotiator. His entire method rests on reading a human being in real time. The pause before an answer, the word someone chooses instead of the obvious one, what they won't say. Currently he runs the Black Swan Group, wrote the book Never Split the Difference, and there's also a newer companion book out by members of his team. called Fight Less, Win More. Chris is now teaching companies how to hand their work over to AI. So I wanna know what happens to a skill built on reading people when the person across the table isn't one. Chris, welcome to Machine Dreams. It's great to have you on. speaker-1: it's a pleasure. Yeah. I'm looking forward to the conversation and and I I an obscure I could an obscure reference to do androids dream of electric sheep. Do you know what I'm talking about? Yeah. speaker-0: Chris, speaker-1: Alright. speaker-0: great. Thank you for being on the show. We appreciate it. So let's just get right into it. You're known for your wealth of experience in tactical empathy. As a matter of fact, a couple years ago, I must have told you the first time we had a chat, I took a course on tactical empathy and tactical negotiation. So that you know, you're known for that as well as many other things. but I find it really interesting the idea that AI which is what everybody's talking about these days, could now start to do some parts of negotiation increasingly on our behalf. what do you make of that idea and what does what does that really, really look like? speaker-1: I I think it's a great advantage. It's a it's a a tool that helps. You know, in point of fact, AI's been doing this for quite a while. When people started f putting up bots to respond. you know, th there's just more sophisticated versions of bots. And way back when somebody asked me, you know, could you have a bot for negotiation? I thought, well, first of all, if it never gets tired and it's never in a bad mood. You've eliminated a couple of problems in negotiations because people make mistakes when they're tired or they're in a bad mood. So if it can help maintain your mood, if you're talking with somebody on the other side, if you're talking to a bot who's not in a bad mood, it's more likely to avoid doing the stupid stuff that people do when they're in a bad mood. So for starters it's it's not a bad idea as a supplement, as a tool that helps. speaker-2: That's interesting. And you know, as a former war correspondent, obviously I've been looking at some of the recent hostage cases over the last few years, and nothing looks the same. Hamas held people as strategic assets with no price attached, negotiated through Qatarian Egyptian mediators over months. Whoever took Nancy Guthrie demanded Bitcoin, sent the notes to TMZ, may never have had her at all. So Are those the two ends of what a kidnapper is now? Does does anything in your old playbook still cover both? speaker-1: Yeah, it does. You know, first of all, Hamas didn't invent that game. and it was originally that game per se was invented kind of by our first time we saw it was Al Qaeda two thousand four time frame, ISIS twenty twelve, twenty fourteen time frame. So they wanted to make it look like they were negotiating, but they weren't. It was publicity, it was PR, it was manipulation. But it wasn't negotiation, because negotiation in point of fact is trying to reach a mutual outcome. It's a By definition it's collaboration. And Hamas, Al Qaeda two thousand four, ISIS twenty twelve ish. They weren't looking to collaborate to reach an outcome. They were looking to manipulate. They were looking to exploit. It was in in both the previous examples and all three of those examples, it was intended murder, it was intended warfare. They tried to disguise it. You know, some some people's, you know, tyranny described as manners, if you will. is another paraphrase. So they were they were never negotiating in any of those instances. Now the Nancy Guthrie thing, it falls into a category. Domestic US kidnappings are a thousand times more dangerous than international kidnappings as a general rule. Because the bad guys ex expect to get caught. And so they put themselves in a position where Really, the victim is never found again. And if you don't smell it from the very beginning, what's the other side's intention? Are they trying to make a deal or are they trying to attack you? Are they predators? And y all those instances that you just mentioned, it's predatory behavior, which you deal with it accordingly. You gotta understand if the other side's not trying to make a deal and then you have to react. speaker-0: Great great Chris. let's come back to what you've said about AI in negotiations and you know, things like procurement, for instance. you've talked about the limitations that humans have. They get tired, they get nervous, you know, they can make mistakes and AI can help to supplement in that regard. But I I'm curious, Cita, thinking about what then is the advantage that a human has over these machines. speaker-1: Yeah, well it's kind of a the philosophical question. I really believe that humans will always have the market cornered on true insight, on you know, true i I don't know if it's insight, I don't know if it's creativity. You know, I'm one of those believers, first of all, the a ridiculous capacity of the human brain to start with, which is far more than we really realize or are willing to recognize, which is why instinct is such a power, gut instinct is such a powerful thing. And there's very loose science unproven these days that the brain may actually have quantum capabilities or the ability to sense energy that we haven't really been able to measure yet. Like when I was w working surveillance in New York, my instincts from following people, like you couldn't follow me. Like I was impossible to follow because I followed people all the time. And I I'd be on a crowded street in Manhattan and suddenly I'd know that somebody was staring at me from a block away who's behind me. Now I believe the human being humans have ability to pick up energy. And then if you start talking about quantum capabilities, you understand the entanglement theory, which is two objects several light years apart, one moves the other one moves simultaneously. There's a lot of science that we can't explain, which I think humans have the capability of. What am I rambling about? I don't think computers have those capabilities. And I think computers would be a great boost. And I know that people are talking about quantum computing is the next level computing. But I think that the the true insight, the instinct, the one percent, the two percent, the five percent that's required to make a great negotiation will always lie in the hands of a human being. speaker-0: Just off the back of that and as a follow up, Chris, are you saying that humans will always be better than AI at negotiation? speaker-1: Yeah, human being at at their top form. I don't think I think is unbeatable. You know, the the insight that comes from a human being whose energy level, you know, roughly ten o'clock in the morning, everybody's at their top form. You're out of bed, you've gotten up, your your nutrition is kicked into gear, your biorhythms, your mitochondria are at their peak, about two hours after you get out of bed. Human being at their top form, I think will always be unbeatable. There's a lot of humans that are not negotiating at that top four. I try to avoid making having important negotiations at four o'clock in the afternoon. I'm getting tired. I got decision fatigue. Machines are never going to have decision fatigue. So there's going to be certain instances where human beings are not at their their max capacity, where a machine computer, an AI, is going to be a great advantage, which will give me an insight in a moment based on my interaction with it. So you know, the issue the qualified answer is a human being at the top form I think will always be unbeatable and we're not always at our top form. speaker-2: Chris, you might be biased on this question, but who's still the best in the world at hostage negotiations, surveillance, reconnaissance? Is it the Americans? Is it the Israelis? Is it the Brits? who would you say is top of the game? speaker-1: Yeah, you know, I've kind of been out of that game for a while, and I'm relying more and more upon, you know, the great business negotiators that are now wading into these complex international negotiations. And on all sides, you know, the most visible of Steve Whitkoff or Wykoff, forgive me if I've said his last name improperly, Jared Kushner. There there are quiet people behind the scenes in the United Arab Emirates, in Saudi, even in Ukraine, and probably even in Russia. Quiet people that have just gotten really good at international business negotiations where you're looking for a collaborative outcome. You want to build a hotel someplace. Look, you gotta get along with the local zoning board to pull that off. You're gonna, you know, y you wanna put up a hotel someplace. They're gonna be people in the neighborhood that either gonna be on your side or against you, wherever it is. And that's just an example. So I think the best negotiators are the ones you're hearing very little about, and I think that's intentional. speaker-2: When you see the crossover, especially now that you're doing more in business, what are the skills that it takes to be good as a CEO, as a negotiator, with your money, with your company, skills that maybe you've transferred over. speaker-1: an an instinct that collaboration makes you more money than being cutthroat. And that that that takes a while to get there. and I'll continue to I get a mental block. I'm I'm good at concepts, I'm poor on names. The CEO of Nvidia, Huang, Jensen Huang. Like, this is a collaborative dude. Like, you know, he i even with the people that are his competitors, I mean, he is overwhelmingly collaborative. I've listened to him on Lex Friedman's podcast not that long ago. His people know what he's thinking about as a leader. You know, he says when he finally makes his decision, his people finally say, Well, what took you so long? Well, that means he's been communicating with him about his thought patterns for quite a while. He's determined to be collaborative with people that are openly competitive with him because he knows that in the bigger picture they're all going to be better off by being ridiculously collaborative. And and I think they're currently the most valuable company on earth. And they're you know, he's operating in Taiwan, he's operating in China, he's talking to everybody. That's the mark of a great leader. Now, a lot there are a lot of very good people that are just competitive. And you can and it's hard to know that you're not great. You could be very good. You could you could be a B plus player. You might even be an A player. The A plus players, they're collaborative, and that's the ear arcs of a guy's collaborative. speaker-2: They keep their friends close and their enemies and their competitors closer. speaker-1: Yeah and after a while the competitors and their enemies say like you know, this guy is he really my enemy? Like he just keeps trying to help me get better. It's it's kind of irresistible over time. speaker-0: I think that's that's some words on on Mabu right there about how collaboration makes you more money than being cutthroat. So that's very useful anecdote for people watching or listening. Chris, I wanna ask you this. obviously you've done a lot of work with negotiations and you've said a lot of really great stuff since we began the conversation, but I wanted to get a sense of what you will do. If you were training an AI to become a world class negotiator, what's the first thing you would teach it? speaker-1: Well, in general terms, you gotta listen between the lines. And as a human being, you're pulling data. What did they say? How did they say it? What do they look like when they say it? Now, a long time ago, there's been this widely disputed ratio of 738-55. 7% of the meaning comes from the words, 38% from the delivery, 55% from the body language. And what in point of fact Is the the numbers don't matter. What matters is when they're out of line. And then when they're out of line, if somebody's tone of voice is different than what they've said, then how do you explore that? How do you listen between the lines? A lot of that we can predict. it's going to be mostly negative, it's going to be mostly fear-driven. It's going to have more to do with their identity issues, their autonomy issues, and their agency issues. Who am I? Am I making this decision myself? Do I have influence on my future? That's identity, autonomy, agency. The fascinating thing is none of those are survival issues. And so realizing that those three are more important than survival, people will die over their autonomy. Business people will blow up a deal that'd make them better off if they feel forced into it. That's where you start to look between the lines and make educated guesses on where the person has come from. If you can teach AI that all in one, it could probably help you. I'm not sure an AI could nail all of it. speaker-0: What what I was really trying to get at with that question was I I just imagine a world where an AI versus an AI, you know, two different AI systems are negotiating. And I don't know if that's happening right now, but I I just wonder how much of the beauty of the human side of things that could take out of conversations and negotiation processes. And I was just curious to know what that might look like, you know, when you have machine versus machine doing high stakes negotiations or or is that something already happening now? speaker-1: Well, i so i it would depend upon how the machines were trained to begin with. You know, garbage in, garbage out. How paranoid were the people that were training them? You know, what were the fears that they embedded in into the into the code that was written in? and I and we're seeing those shortcomings a lot. Like certain I'm just discussing today with my team, we use multiple AI systems. And we're learning about the shortcomings of each one. For example, if the person that programmed the AI thinks that great negotiation is doing a great job of explaining, being eloquent, that's not going to be any good. If you're explaining you're losing. It's not a it's not a matter great negotiators aren't eloquent, they're great at comprehending what the other side said. Now, with all those qualifications A couple of good AIs are not going to give you bad outcomes. You're going start avoiding bad outcomes. You're not going to get great ones. But there's going to be there's a definite advantage to getting rid of the horrible ones. You know, a steady B player beats the erratic player who occasionally delivers a deal that just destroys your company. So you're going to avoid the really bad outcomes, and there's something to be said for that. speaker-2: Chris, you are the expert negotiator, but you seem to also be the expert guy on reinventing yourself. It seems like you've had a metamorphosis, you've had many lives, you've had many roles, and you've reinvented yourself again and again. Tell me the process of who you are now versus who you used to be, advice for people on how to come out of one role and take on a new one with using the skills you learned before and putting them to something new. speaker-1: Well, all you gotta do is love learning, and realize that there's endless, delightful, beautiful novelty in to continue to learn. there's a gentleman I have tremendous respect for, Michael Mulill founded this marketing company, Crisp Video, which was originally for marketing for lawyers, and now it's expanded into so many training things to help law firms make a lot of money. Anyway, one of Michael's favorite sayings is if you're not embarrassed of what you thought a year ago, you haven't learned enough in the last year. So th there's just I just keep learning. I just I'm having the best time, discovering new ideas. my former GM who moved on to a better opportunity, which was more challenging and satisfying for him, once told me, he said, you know, you learn more than anybody I ever know, ever met. You you listen to podcasts, you're reading books, like you're absorbing stuff all over the place. speaker-2: Which was my next question. What what are you doing to stay so learned and informed? What do you have a routine or do you read twelve papers in the morning? You mentioned podcasts. How do you stay on top of it all? speaker-1: Yeah, first of all, d don't expect yourself to stay on top of it all. But then find the people that you find the most interesting to listen to. and s they may be go to sources for different reasons. Like Andrew Huberman is my go to source on neuroscience. If I wanna start talking about neurochemicals and neuroscience, Andrew's Andrew's solid. I mean, he's as rock solid as anybody is out there. So When I need a dose of that, I'll you know, I'll scroll through his different episodes and I'll find one that I want to listen to and I'll listen to it while I'm I'm hitting my song or something like that. Finding those moments during the day to feed your brain. Then just becomes this I get dopamine hits from learning new stuff. speaker-0: Great stuff, Chris. Last question. And this is very personal. obviously there's a lot of challenges with AI systems, such as hallucinations and bias with, you know, AI systems producing incorrect or sometimes fabricated output. I'm thinking to myself right here, and I'm I'm sure a lot of our viewers and listeners probably are also thinking about this question. There is probably a conversation I'm having with a boss at work. And I'm thinking about how do I negotiate with my boss? How do I get the fair price? I'm trying to get sign on for a deal and I'm trying to negotiate the price for that deal. Or, you know, we kn negotiate every day across the activities that we perform. The big question I want to ask is how do I use AI in my personal life for negotiation or persuasion purposes? and how do I do that? in a way that I'm able to avoid things like hallucinations, bias in the output. speaker-1: Well, whatever AI you're using, continue to educate it. And so you're getting ready to get into a negotiation with somebody, anybody. Ask yourself, what are they probably thinking? H here's who they are w in me, here's how long I've known here's the nature of the relationship. What are they probably thinking? What they're probably thinking is basically what your empathy statement should be. Empathy is how the other side sees things. And so then Engage in the data. Give your AI a shot at guessing, making an educated guess of what the other side's probably thinking. When you talk to them, articulate it out loud, they're going to tell you if you're wrong or right. Now take that reaction, feed it back into your AI, and continue to train your AI. And eventually your AI is going to get really good at it. But you have to look at it as a little bit of a game, have fun with it, and enjoy the process. That way when you make mistakes, which you will. That'll actually make you smart. speaker-0: That's really, really insightful, really great. And I'm gonna take some some of that for my personal life 'cause I've got a lot of negotiation to do and so we'll see how that pans out. I probably will let you know if I get any results. speaker-1: And and if you want to get really good, then come to my company and let us coach you. speaker-0: Chris, thank you so much for making it to the show today. This was a really, really exciting, great conversation. We appreciate it. speaker-1: Fun talking to you guys. Thank you for having me on, it's a privilege. speaker-2: Thanks, Chris, and thank you to everyone listening. This is Machine Dreams. More honest, unfiltered conversations with the people shaping the future of AI are on the way. See you next time.