WEBVTT 00:00:03.780 --> 00:00:07.050 Matt Abrahams: You can catalyze community through compelling communication. 00:00:07.800 --> 00:00:11.370 My name is Matt Abrahams, and I teach strategic communication at 00:00:11.370 --> 00:00:13.260 Stanford Graduate School of Business. 00:00:13.590 --> 00:00:18.244 Welcome to this live episode of Think Fast Talk Smart, the podcast. 00:00:18.875 --> 00:00:22.174 The power of community to enhance learning, entrepreneurship, and 00:00:22.174 --> 00:00:23.945 connection is truly amazing. 00:00:24.544 --> 00:00:28.085 I witnessed this firsthand at the recent Stanford SEED summit 00:00:28.115 --> 00:00:29.705 in Cape Town, South Africa. 00:00:30.244 --> 00:00:34.415 I invite you to listen in as Stanford professors Jesper Sørensen, Christian 00:00:34.415 --> 00:00:39.510 Wheeler and incoming GSB Dean Sarah Soule share their thoughts with and 00:00:39.510 --> 00:00:44.644 answer questions from the close to five hundred SEED members in our audience. 00:00:45.255 --> 00:00:46.125 So let's get started. 00:00:46.125 --> 00:00:47.565 Jesper, thank you for joining us. 00:00:47.565 --> 00:00:48.825 Excited to have you. 00:00:48.825 --> 00:00:53.775 When you and I spoke earlier on the podcast, we talked about strategy and 00:00:53.775 --> 00:00:58.575 telling stories around strategy, and I know that you've done some recent work 00:00:58.575 --> 00:01:02.515 that continues that line where you've done research with your colleague, Glenn 00:01:02.535 --> 00:01:08.820 Carroll, on the use of analogies in being effective in the work that you do? 00:01:08.820 --> 00:01:10.260 Can you share a little bit more about that? 00:01:10.350 --> 00:01:10.710 Jesper Sørensen: Sure. 00:01:10.800 --> 00:01:14.520 As many of you know, Glenn and I published a book a few years ago, and the paper 00:01:14.520 --> 00:01:18.030 that we recently wrote on analogies, we think of this as the missing chapter. 00:01:18.030 --> 00:01:22.590 The chapter that didn't get done in the book, and we use analogies all the time. 00:01:22.650 --> 00:01:26.460 So we say this is kind of like that when we're trying to explain something, right? 00:01:26.460 --> 00:01:29.640 So we think sound is like a wave, right? 00:01:29.640 --> 00:01:32.310 There's a classic way of thinking about explaining principles 00:01:32.310 --> 00:01:33.480 in physics, for example. 00:01:33.480 --> 00:01:36.270 And we use this all the time in business as well. 00:01:36.420 --> 00:01:39.210 So I think what most familiar to people would be, we use 00:01:39.210 --> 00:01:40.740 it in startup pitches, right? 00:01:40.740 --> 00:01:41.520 All the time. 00:01:41.730 --> 00:01:45.750 So we'll say, my company is the Uber of, or something like that. 00:01:45.750 --> 00:01:51.030 And then what we're doing is we're trying to persuade people that our 00:01:51.524 --> 00:01:56.414 organization and that analogy has the same potential and maybe even the same 00:01:56.414 --> 00:01:59.865 business model as the source organization. 00:02:00.014 --> 00:02:03.945 So that's a much more common way of reasoning, I think that people do. 00:02:03.945 --> 00:02:08.115 It's a particularly powerful for idea generation and it's also 00:02:08.115 --> 00:02:10.035 particularly powerful for persuasion. 00:02:10.454 --> 00:02:13.635 As a form of communication it's really quite compelling. 00:02:13.875 --> 00:02:18.650 Uh, at the same time, it's super dangerous as a basis for strategic thinking. 00:02:18.650 --> 00:02:21.320 So what our article is about is basically about how can you think 00:02:21.320 --> 00:02:25.430 about what makes for a good analogy from a strategic perspective. 00:02:25.490 --> 00:02:29.840 And I think one element of this is we tend to think about analogies as being 00:02:29.840 --> 00:02:32.450 about what we call horizontal comparisons. 00:02:32.450 --> 00:02:36.859 So we're capturing features and comparing them to each other, but we actually 00:02:36.859 --> 00:02:39.470 think that the more important thing to do is what we think about as the 00:02:39.470 --> 00:02:44.665 vertical relationships, which is really about what's the logic underlying each. 00:02:44.665 --> 00:02:48.804 When you say that your business is the Uber of something else, what is it 00:02:48.804 --> 00:02:53.005 about Uber's business model that you are saying applies in your business? 00:02:53.005 --> 00:02:55.975 And so the article really tries to flesh those ideas out. 00:02:56.185 --> 00:02:59.125 Matt Abrahams: I like that idea of, it's not horizontal but the vertical 00:02:59.125 --> 00:03:00.625 logic that's really important. 00:03:00.864 --> 00:03:05.005 That answer was as good as your previous answers, so thank you. 00:03:05.524 --> 00:03:08.734 Sarah, I recently had the pleasure of hearing you give a lecture where 00:03:08.734 --> 00:03:12.125 you talked about some recent research you were doing on how corporations 00:03:12.125 --> 00:03:14.435 respond to societal issues. 00:03:14.554 --> 00:03:16.894 Would you share for all of us a little bit about that research 00:03:16.894 --> 00:03:18.125 and what we can take away from it? 00:03:18.454 --> 00:03:18.815 Sarah Soule: Sure. 00:03:18.815 --> 00:03:19.774 Absolutely, Matt. 00:03:20.075 --> 00:03:24.635 One of my PhD students and I got very interested in corporate political 00:03:24.635 --> 00:03:28.804 statements, and some of you maybe have made these statements before. 00:03:29.195 --> 00:03:32.375 Many of you have read these statements before, but we were 00:03:32.435 --> 00:03:36.575 interested in these as a growing phenomenon, at least in the US case. 00:03:36.905 --> 00:03:42.185 And we collected a number of these around a lot of different social and political 00:03:42.185 --> 00:03:47.405 issues, and we started to notice some very interesting patterns in some of these. 00:03:47.765 --> 00:03:52.030 And in particular, one of the things that we noticed is that usually 00:03:52.120 --> 00:03:56.350 organizations and companies, they usually put forward and put out very 00:03:56.350 --> 00:03:58.630 positive information about themselves. 00:03:58.990 --> 00:04:03.370 In extreme cases, we may call this greenwashing or some other kind of 00:04:03.370 --> 00:04:07.750 pinkwashing and so on, but typically we think about companies putting forth 00:04:07.750 --> 00:04:09.550 something positive about themselves. 00:04:10.090 --> 00:04:13.630 But in these political statements, we noticed that companies were 00:04:14.115 --> 00:04:17.654 doing something very different, and that we came to call confession. 00:04:18.255 --> 00:04:22.635 They were releasing negative information about themselves with respect to whatever 00:04:22.635 --> 00:04:24.825 the political or social issue is. 00:04:24.945 --> 00:04:29.534 So for example, in the environmental space, rather than putting out something 00:04:29.534 --> 00:04:33.745 that might resemble something on the spectrum of greenwashing, companies would 00:04:33.745 --> 00:04:38.545 put something out talking about how dirty their supply chains are, things like this. 00:04:38.695 --> 00:04:42.535 And so we came, as I said, to refer to this as confession, company 00:04:42.535 --> 00:04:46.480 confessions, but we were also interested in understanding how 00:04:46.480 --> 00:04:51.315 people, the general public, consumers, might react to those statements. 00:04:51.555 --> 00:04:55.275 And so we conducted a number of online experiments and asked 00:04:55.275 --> 00:05:00.195 people to rate these companies with respect to different dimensions of 00:05:00.195 --> 00:05:01.965 corporate social responsibility. 00:05:02.415 --> 00:05:05.925 And the way this worked is that we created some vignettes, some sort of 00:05:05.925 --> 00:05:10.185 fake statements that a company might put out based on some of the company 00:05:10.185 --> 00:05:11.805 statements that we had collected. 00:05:12.225 --> 00:05:14.985 And we asked people a series of questions. 00:05:15.375 --> 00:05:19.335 And what we were manipulating in the experiment was the extent to which 00:05:19.335 --> 00:05:24.405 they confess something negative about themselves versus neutral versus positive. 00:05:24.645 --> 00:05:29.805 And what we found was that companies that confess something were rated 00:05:29.805 --> 00:05:33.825 more highly with respect to corporate social responsibility than those 00:05:33.825 --> 00:05:37.385 that just put a neutral statement out or a positive statement out. 00:05:37.815 --> 00:05:41.534 But there's a little wrinkle in this finding, and that's that if the 00:05:41.534 --> 00:05:46.995 company was confessing something that was particularly surprising or might 00:05:47.085 --> 00:05:52.485 seem counter to what an individual's priors might be, they were discounted 00:05:52.485 --> 00:05:56.685 very heavily and believed to be not at all responsible with respect to 00:05:56.685 --> 00:05:58.395 corporate social responsibility. 00:05:58.485 --> 00:06:03.735 So for example, if we have a lumber company that confesses to something 00:06:03.794 --> 00:06:08.174 with respect to deforestation, they get a bump in how people perceive 00:06:08.174 --> 00:06:11.395 them in terms of their level of corporate social responsibility. 00:06:11.675 --> 00:06:16.140 But if the lumber company confesses to using sweatshop labor in 00:06:16.140 --> 00:06:19.380 an emerging market, that they are discounted very heavily. 00:06:19.739 --> 00:06:24.210 So in a sense, I think what we are picking up on is hypocrisy, and that's 00:06:24.210 --> 00:06:28.409 that consumers and general public is very astute at picking up something 00:06:28.409 --> 00:06:33.210 that feels like it could be hypocritical or feels like it could be an empty 00:06:33.210 --> 00:06:37.289 statement that doesn't connect to their general business and so on. 00:06:37.289 --> 00:06:38.864 So that's the gist of what we've been doing. 00:06:39.585 --> 00:06:41.295 Matt Abrahams: Which I find really interesting. 00:06:41.295 --> 00:06:45.015 I'm curious, do you think that this notion of corporate confession is 00:06:45.015 --> 00:06:49.335 something that companies should be thinking about in terms of how they 00:06:49.335 --> 00:06:50.715 present themselves in the world? 00:06:51.015 --> 00:06:54.105 Sarah Soule: I think I would say that, and this is maybe obvious, but it's 00:06:54.105 --> 00:06:57.435 astounding to me how many of these corporate statements we saw, which 00:06:57.465 --> 00:07:01.365 really did seem a bit performative and seem quite hypocritical, I think 00:07:01.365 --> 00:07:06.200 the takeaway for me is that we can tease out something that feels like 00:07:06.380 --> 00:07:10.760 individuals sensing that there's something awry in what the company is saying. 00:07:11.000 --> 00:07:14.270 And so I think what I would say is that when one, if a company is going to use 00:07:14.270 --> 00:07:18.140 this strategy of releasing corporate statements, they really need to be 00:07:18.140 --> 00:07:19.910 very careful about what they're doing. 00:07:20.270 --> 00:07:24.560 That whatever they're confessing to is directly related to their line of 00:07:24.560 --> 00:07:29.210 business, their values, their mission, their strategy, and not weighed into 00:07:29.240 --> 00:07:34.055 political issues that are far afield from where they play and what they do. 00:07:34.115 --> 00:07:36.125 Matt Abrahams: So stay close to what you know and what you do. 00:07:36.185 --> 00:07:36.485 Sarah Soule: Yes. 00:07:36.485 --> 00:07:36.995 Matt Abrahams: Thank you. 00:07:37.325 --> 00:07:41.255 So Christian, you co-teach a class on spontaneous management, I love 00:07:41.255 --> 00:07:45.785 that idea, where you blend ideas from your area of study and improvisation. 00:07:46.235 --> 00:07:50.045 What are one or two of the ideas you can share that you teach in your class that 00:07:50.045 --> 00:07:53.285 can help all of us in the ever-changing environment we live in today? 00:07:53.730 --> 00:07:54.030 Christian Wheeler: Sure. 00:07:54.030 --> 00:07:55.410 I'll say a couple of things. 00:07:55.410 --> 00:07:57.180 We get used to postponing judgment. 00:07:57.240 --> 00:08:00.930 We have a tendency to rush to categorization, to assume that we 00:08:00.930 --> 00:08:05.040 understand things before we really do, because it saves us cognitive energy. 00:08:05.040 --> 00:08:08.370 It makes us feel that we have an understanding and a confidence 00:08:08.370 --> 00:08:09.600 about the way things are. 00:08:09.600 --> 00:08:14.550 But what it takes away from us is a curiosity about things around us, 00:08:14.550 --> 00:08:17.910 the ability to notice things that might peak our interest, and it 00:08:17.910 --> 00:08:20.160 also leads to inaccurate judgments. 00:08:20.540 --> 00:08:23.960 We also have a tendency to judge ourselves often in a moment. 00:08:24.140 --> 00:08:27.650 Right now, I may be having some self-talk thinking about whether this is a good 00:08:27.650 --> 00:08:29.480 answer or people liking this answer. 00:08:29.960 --> 00:08:30.650 I don't know. 00:08:30.980 --> 00:08:34.610 This is not the time for me to be judging this answer because the more 00:08:34.610 --> 00:08:37.730 I'm in my head, judging this answer I'm giving, the less I'm thinking 00:08:37.730 --> 00:08:39.289 about what I'm gonna say next. 00:08:39.470 --> 00:08:42.260 So I can listen to the podcast when it's out, reflect on my 00:08:42.260 --> 00:08:45.770 performance, think about how I could do better, but now is not the time 00:08:45.770 --> 00:08:47.570 for me to be judging my answer. 00:08:47.660 --> 00:08:50.480 Then I would also say another thing is being present. 00:08:50.689 --> 00:08:54.079 A lot of the class talks about capitalizing on serendipity, 00:08:54.140 --> 00:08:57.620 about being attentive to nonverbal signals from other people. 00:08:57.800 --> 00:09:00.199 This requires that we are in the moment. 00:09:00.290 --> 00:09:04.640 If we're honest about how we spend much of our time, it's like this, right? 00:09:04.640 --> 00:09:06.020 Staring at our telephones. 00:09:06.020 --> 00:09:08.689 You can't be present if you're staring at your telephone. 00:09:08.900 --> 00:09:12.630 And what it does is it has all of these effects on being able to 00:09:12.630 --> 00:09:16.079 capitalize on chance circumstances, being able to notice things that 00:09:16.079 --> 00:09:19.079 might capture our interest, but it also takes us out of this interaction 00:09:19.079 --> 00:09:20.490 that you're having with other people. 00:09:20.579 --> 00:09:23.910 And research shows that not only does it diminish the quality of 00:09:23.910 --> 00:09:27.540 your interactions, but it actually lowers your cognitive capabilities. 00:09:27.660 --> 00:09:31.365 If you have a telephone in front of you, switched off, then if it's out of 00:09:31.370 --> 00:09:35.579 the room, so it doesn't even need to be turned on, just merely having it present 00:09:35.850 --> 00:09:38.885 distracts us from the moment, and it lowers our ability to think carefully. 00:09:39.439 --> 00:09:41.090 Matt Abrahams: So our phones are making us dumber. 00:09:41.150 --> 00:09:41.870 Christian Wheeler: Absolutely. 00:09:42.170 --> 00:09:42.830 Matt Abrahams: Interesting. 00:09:43.400 --> 00:09:45.140 So I want to dive a little deeper. 00:09:45.140 --> 00:09:48.770 I, 'cause I see being present and non-judgmental sort of together. 00:09:48.770 --> 00:09:52.880 I mean, what are some things we can do to be more present oriented? 00:09:52.880 --> 00:09:55.310 It's one thing to want to be, it's another thing to actually do it. 00:09:55.370 --> 00:09:57.940 Christian Wheeler: Well, a simple thing is put your phone in your bag, turn it off. 00:09:58.160 --> 00:10:01.370 But I think more, we often use our cell phones as a sort 00:10:01.370 --> 00:10:02.920 of security blanket, right? 00:10:02.920 --> 00:10:07.970 The moment we have an unoccupied second of time we have this sense of existential, 00:10:09.479 --> 00:10:11.609 oh, here I am alone with my thoughts. 00:10:11.609 --> 00:10:12.689 What will I do now? 00:10:12.930 --> 00:10:15.270 You can overcome this feeling. 00:10:15.300 --> 00:10:19.530 And in fact, I was alive at a time when we didn't have such devices and 00:10:19.530 --> 00:10:23.189 we just had to wait for the train with nothing to do, and we were okay. 00:10:23.369 --> 00:10:26.010 Matt Abrahams: And you can actually have a conversation with somebody instead of, 00:10:26.390 --> 00:10:27.349 Christian Wheeler: You actually can. 00:10:27.349 --> 00:10:32.150 And for example, that even though many of us have the hypothesis, the lay belief, 00:10:32.329 --> 00:10:35.479 that it would be unpleasant to have a conversation with a stranger at the 00:10:35.479 --> 00:10:37.339 train, research shows quite the opposite. 00:10:37.640 --> 00:10:40.369 That if you strike up a conversation with someone on the train, I think 00:10:40.369 --> 00:10:44.060 they did this on buses in Chicago, you like that conversation not only more 00:10:44.060 --> 00:10:46.280 than you thought that you would, but that it improves your quality of life. 00:10:46.280 --> 00:10:51.439 So having this ability to interact with your environment in an uninhibited way 00:10:51.439 --> 00:10:55.089 from your devices is something that it takes a while to get used to, but 00:10:55.089 --> 00:10:56.439 it can improve your quality of life. 00:10:56.589 --> 00:10:57.219 Matt Abrahams: Excellent. 00:10:57.339 --> 00:11:00.490 And I wanna pick up on a thread that Christian mentioned about not in the 00:11:00.490 --> 00:11:03.790 moment ruminating on what's happening. 00:11:04.120 --> 00:11:07.240 There's a difference, to my mind, between rumination and reflection. 00:11:07.329 --> 00:11:10.329 A lot of us get in our head in the moment when something doesn't go 00:11:10.329 --> 00:11:14.319 right and that prevents us from being in that moment or what comes next. 00:11:14.500 --> 00:11:16.750 But it does make sense after the fact to reflect. 00:11:16.750 --> 00:11:17.630 That's how we learn. 00:11:17.680 --> 00:11:20.860 So that separation between rumination and reflection, I think is important. 00:11:21.040 --> 00:11:24.040 But I want us to spend a few moments talking about AI, artificial 00:11:24.040 --> 00:11:28.360 intelligence, and I'm curious to learn from each of you, how is it 00:11:28.360 --> 00:11:32.349 impacting the work that you do with your students, in the research you do? 00:11:32.560 --> 00:11:34.990 Sarah, do you mind talking about AI first for us? 00:11:35.110 --> 00:11:35.650 Sarah Soule: Sure. 00:11:35.650 --> 00:11:36.459 Happy to. 00:11:36.670 --> 00:11:40.329 I think probably everybody in the room understands that this is impacting all 00:11:40.329 --> 00:11:44.365 of us in ways that we probably even a year ago, didn't think possible. 00:11:44.635 --> 00:11:48.415 But a few things that come to my mind in terms of research, one of the things 00:11:48.415 --> 00:11:52.464 that I've been using it for is to run some of these corporate statements 00:11:52.464 --> 00:11:57.025 that I mentioned through an AI to pick up on the sentiment of the speaker, 00:11:57.055 --> 00:12:01.015 the writer, and to see, and we haven't done much with this yet, but we plan 00:12:01.045 --> 00:12:05.485 to look at the emotionality of these statements and see if we can pick up 00:12:05.545 --> 00:12:09.775 in any way that might impact how people rate these companies in terms of their 00:12:09.775 --> 00:12:11.875 level of corporate social responsibility. 00:12:12.234 --> 00:12:14.635 So that's one way that we're using this in research. 00:12:14.814 --> 00:12:18.775 In teaching, I was thinking about this just the other day, I was preparing 00:12:19.104 --> 00:12:25.140 an exercise for use in a classroom for people to map their social network. 00:12:25.410 --> 00:12:29.470 And so I prepared the exercise and I said, well, I wonder what ChatGPT 00:12:29.489 --> 00:12:31.020 would say about this exercise. 00:12:31.469 --> 00:12:35.910 It gave me some amazing suggestions for reflection questions, things I 00:12:35.910 --> 00:12:39.000 hadn't even thought about, and I've been doing this for many years now. 00:12:39.000 --> 00:12:42.420 So that was humbling, but also very helpful as well. 00:12:42.930 --> 00:12:50.685 I've asked ChatGPT recently to create menus for me of higher protein, vegetarian 00:12:50.685 --> 00:12:56.115 or plant-based diets that meet a certain number of grams of protein per day. 00:12:56.385 --> 00:13:00.345 And I was, again, astounded at how quickly it would spit out these very, I haven't 00:13:00.345 --> 00:13:05.385 tried them yet, but very reasonable sounding kinds of ideas and diets. 00:13:05.385 --> 00:13:10.194 And so I think that this is, while at times terrifying, I think it's also 00:13:10.344 --> 00:13:14.995 been very helpful, in at least those aspects of my life and work these days. 00:13:14.995 --> 00:13:15.324 Matt Abrahams: Excellent. 00:13:15.324 --> 00:13:19.045 So you're using it in research, in your teaching, and even in your personal life. 00:13:19.045 --> 00:13:21.295 Christian, how are you leveraging AI for what you do? 00:13:21.775 --> 00:13:21.954 Christian Wheeler: Yeah. 00:13:21.954 --> 00:13:24.925 I don't use it so much for my teaching for reasons related to what 00:13:24.925 --> 00:13:28.495 we were just discussing, but I do use it quite a bit for my research. 00:13:28.555 --> 00:13:32.055 A new thing that I'm excited about is having participants 00:13:32.115 --> 00:13:34.695 engage in conversations with AI. 00:13:35.025 --> 00:13:38.745 So some of the things that we do, we are interested in having people think about 00:13:38.745 --> 00:13:42.765 their attitudes or their thoughts or their beliefs in a different way than they might 00:13:42.765 --> 00:13:45.285 spontaneously do so, or ordinarily do so. 00:13:45.405 --> 00:13:49.125 And back in the day, what we would do is we would give them instructions to, 00:13:49.155 --> 00:13:50.865 to think about it in a certain way. 00:13:50.895 --> 00:13:55.045 But what you can do now is have people have a conversation with the chatbot. 00:13:55.065 --> 00:13:58.455 They will encourage them over a series of rounds to think about 00:13:58.455 --> 00:13:59.625 something in a particular way. 00:13:59.625 --> 00:14:02.955 We'll ask them follow up questions designed to have them 00:14:02.955 --> 00:14:05.595 have this sort of realization about their thoughts or beliefs. 00:14:05.595 --> 00:14:08.895 And so I think that's an exciting new direction for manipulation 00:14:08.895 --> 00:14:10.215 of psychological constructs. 00:14:10.725 --> 00:14:13.574 Matt Abrahams: So you're using it a lot in research and actually having people 00:14:13.574 --> 00:14:17.415 using the tool to help people reflect on their communication and their thoughts. 00:14:17.415 --> 00:14:17.895 Excellent. 00:14:17.895 --> 00:14:18.285 Very good. 00:14:18.525 --> 00:14:20.444 Jesper, how are you using AI? 00:14:20.775 --> 00:14:23.564 Jesper Sørensen: I would say the extent that I use it, I use it in, in teaching. 00:14:23.715 --> 00:14:26.834 So one of the things that I have always struggled with in teaching is 00:14:26.834 --> 00:14:29.685 examples and ways of illustrating ideas. 00:14:29.995 --> 00:14:33.715 And so I was recently working with some colleagues on a, developing a new 00:14:33.715 --> 00:14:38.155 course where we needed to come up with some kind of hypothetical examples 00:14:38.155 --> 00:14:43.770 to be able to explain certain basic principles in economics and strategy. 00:14:44.070 --> 00:14:47.310 And I'm just terrible at doing that myself, right? 00:14:47.310 --> 00:14:48.660 I just, I cannot do it. 00:14:48.720 --> 00:14:51.720 I can give you all the theory in the world, but being connected 00:14:51.720 --> 00:14:53.280 to the real world is really hard. 00:14:53.280 --> 00:14:55.980 So, that's because I like my phone. 00:14:56.010 --> 00:14:57.630 So stay away, Christian. 00:14:57.900 --> 00:14:58.800 Leave me alone. 00:14:58.920 --> 00:15:02.310 But I did use ChatGPT for this, where I basically wrote a prompt 00:15:02.310 --> 00:15:05.969 saying, okay, you're an expert in this particular topic, and I think in 00:15:05.969 --> 00:15:09.839 economics for example, and then I had like a broad outline of what I wanted. 00:15:09.839 --> 00:15:14.819 I wanted to be a luxury hotel that had an HR management problem and was thinking 00:15:14.819 --> 00:15:17.579 about incentives and all that other stuff. 00:15:17.670 --> 00:15:22.589 The amount of creativity that just popped up like immediately was amazing. 00:15:22.890 --> 00:15:26.880 And I feel more comfortable using it in that way 'cause I do worry 00:15:26.880 --> 00:15:30.840 about these models hallucinating, but these are fictitious examples. 00:15:30.840 --> 00:15:32.730 So hallucination is fine, right? 00:15:32.850 --> 00:15:35.790 But what we want is to make sure that they can be used to illustrate 00:15:35.790 --> 00:15:37.350 the concepts that we're doing. 00:15:37.350 --> 00:15:40.470 And I thought that was actually a really a powerful use case for me. 00:15:41.095 --> 00:15:43.345 Matt Abrahams: I like how you're leveraging its ability to be 00:15:43.345 --> 00:15:45.325 creative, to be creative for you. 00:15:45.745 --> 00:15:50.095 In the work I do around spontaneous speaking, I'll often have my students 00:15:50.095 --> 00:15:54.655 go to an LLM if they're preparing to give a presentation, or some other 00:15:54.655 --> 00:16:00.145 interaction, and have the LLM serve up spontaneous things for them to respond to. 00:16:00.145 --> 00:16:03.385 For example, my students might be giving a presentation and they 00:16:03.385 --> 00:16:07.015 can say, imagine I'm giving a presentation to fellow MBA students. 00:16:07.015 --> 00:16:09.115 What are three questions they might ask? 00:16:09.735 --> 00:16:12.704 And then the students practice answering those questions that are served up. 00:16:12.704 --> 00:16:15.944 So it serves as a way of generating questions so they can practice. 00:16:16.185 --> 00:16:20.865 And talking to someone else on the podcast who is an expert on AI, we had a similar 00:16:20.865 --> 00:16:24.045 conversation, and the recommendation he gave me is, if you don't know how 00:16:24.045 --> 00:16:26.055 to use it, ask it and it can tell you. 00:16:26.295 --> 00:16:29.969 I told ChatGPT that I was interested in doing this particular assignment. 00:16:30.469 --> 00:16:31.605 How could it help me? 00:16:31.605 --> 00:16:34.485 And it gave me lots of interesting ideas about how it could help me. 00:16:34.485 --> 00:16:36.730 And we actually executed on a few of those. 00:16:36.730 --> 00:16:40.150 So sometimes if you don't know how to use AI, maybe AI can help you 00:16:40.360 --> 00:16:41.710 figure out how you could use it. 00:16:42.010 --> 00:16:44.050 So this will be our final question for the panel. 00:16:44.050 --> 00:16:45.190 Christian, let me start with you. 00:16:45.190 --> 00:16:48.160 I'd like for you to share something you're working on that you're currently 00:16:48.160 --> 00:16:51.970 really excited about, and I'll give you some extra credit, imaginary 00:16:51.970 --> 00:16:56.200 extra credit, if you can apply what you're working on to everybody in the 00:16:56.200 --> 00:16:57.700 audience and how they could leverage it. 00:16:58.510 --> 00:17:01.510 Christian Wheeler: A lot of my research projects are a bit in the weeds, so I'm 00:17:01.510 --> 00:17:05.079 gonna give you something very simple, that most of you should be able to apply. 00:17:05.260 --> 00:17:08.110 For a lot of the things that we do, we need to generate a title 00:17:08.110 --> 00:17:11.530 or a heading to what we're doing, and we have a choice to make. 00:17:11.530 --> 00:17:15.160 We could frame that title as a statement, or we could frame 00:17:15.160 --> 00:17:16.690 that title as a question. 00:17:17.294 --> 00:17:22.935 So for example, if I have a newspaper article about green tea, I could say green 00:17:22.935 --> 00:17:25.335 tea has health benefits, question mark. 00:17:25.784 --> 00:17:29.115 Or I could say green tea has health benefits, period. 00:17:29.685 --> 00:17:30.735 Which one is better? 00:17:31.065 --> 00:17:31.725 Question mark. 00:17:31.935 --> 00:17:32.774 I thought so too. 00:17:32.835 --> 00:17:33.495 You're wrong. 00:17:34.995 --> 00:17:35.235 Yeah. 00:17:35.385 --> 00:17:36.375 Isn't that interesting? 00:17:36.375 --> 00:17:38.205 As it turns out, statements work better. 00:17:38.205 --> 00:17:40.185 We've tested this over multiple domains. 00:17:40.185 --> 00:17:44.385 We've tested it with actual newspaper headlines, millions and 00:17:44.385 --> 00:17:45.645 millions of newspaper headlines. 00:17:45.735 --> 00:17:48.705 We've tested it with things like Reddit posts and seeing 00:17:48.705 --> 00:17:50.100 how many up votes they get. 00:17:50.475 --> 00:17:53.024 We've tested it with academic articles and seeing how many 00:17:53.024 --> 00:17:55.195 citations those academic articles get. 00:17:55.385 --> 00:17:59.270 In all of those cases, question marks perform worse than statements. 00:17:59.330 --> 00:18:00.350 The effect isn't huge. 00:18:00.350 --> 00:18:02.540 If it were too big, we would be a little suspicious of it. 00:18:02.810 --> 00:18:05.030 It's a little effect, but it's a reliable effect. 00:18:05.540 --> 00:18:07.760 Matt Abrahams: That was a great answer, exclamation point. 00:18:08.870 --> 00:18:10.130 I don't know how that plays, but good. 00:18:10.440 --> 00:18:12.920 Jesper, what's something you're working on that we can all benefit from? 00:18:13.220 --> 00:18:15.950 Jesper Sørensen: We recently at the GSB, we've been, I've been working 00:18:15.950 --> 00:18:20.120 with some colleagues on developing a new asynchronous course called 00:18:20.120 --> 00:18:24.350 Stanford Business Essentials, and this is a product that's targeted 00:18:24.350 --> 00:18:26.360 at early career professionals. 00:18:26.360 --> 00:18:30.650 So these are basically people who have graduated from university or 00:18:30.650 --> 00:18:32.455 something like that, and they're starting out in their careers. 00:18:33.190 --> 00:18:36.790 And one of the inspirations for this product is I look at my own children 00:18:36.790 --> 00:18:40.510 who have been in this situation and they are just befuddled right by 00:18:40.510 --> 00:18:42.430 what's going on in their organizations. 00:18:42.430 --> 00:18:45.370 Like, why do things happen this way, et cetera, et cetera. 00:18:45.640 --> 00:18:49.575 I think this is a great tool for really getting people up to speed. 00:18:50.370 --> 00:18:54.780 Matt Abrahams: I was recently being interviewed for a Korean outlet because 00:18:54.780 --> 00:18:58.680 my book was just translated into Korean, and the interviewer told me that at the 00:18:58.680 --> 00:19:04.139 university level, they are teaching a class on how to make phone calls because 00:19:04.560 --> 00:19:08.820 students at that age don't know how to actually speak on the phone to somebody. 00:19:09.180 --> 00:19:13.140 And so this notion of providing foundational skills, I don't know 00:19:13.140 --> 00:19:17.100 that we have to get that foundational, but really rings the cord for sure. 00:19:17.370 --> 00:19:18.870 Sarah, how about something you're working on? 00:19:18.870 --> 00:19:20.550 I know you've got a lot going on. 00:19:20.610 --> 00:19:23.430 Sarah Soule: Coming back to the project that I mentioned before about these 00:19:23.430 --> 00:19:26.790 corporate statements, one of the other things that we've been looking at 00:19:26.790 --> 00:19:33.300 is whether or not elements of those statements will induce people to donate 00:19:33.300 --> 00:19:35.610 money to the particular issue at hand. 00:19:35.640 --> 00:19:39.840 Also to write voluntary letters on behalf of the issue at hand. 00:19:39.930 --> 00:19:43.920 And so we wanna try to see if these things can mobilize people to do something. 00:19:43.980 --> 00:19:47.220 So if we have a statement, a fictitious statement about a lumber 00:19:47.220 --> 00:19:52.260 company, and it's just confessed to deforestation, will people who see a 00:19:52.260 --> 00:19:57.435 confession actually be more willing to give money to an environmental 00:19:57.435 --> 00:20:01.095 cause, or write a letter, volunteer to write a letter, a longer letter. 00:20:01.515 --> 00:20:04.515 But one of the things that we've been playing with comes back to something 00:20:04.515 --> 00:20:09.450 that you have taught me, Matt, and that is the what, so what, now what, 00:20:09.659 --> 00:20:11.820 framework that you all now also know. 00:20:12.210 --> 00:20:16.860 And it turns out that framework has been used in studies of getting 00:20:16.860 --> 00:20:21.090 people to mobilize, and it's usually referred to as a kind of collective 00:20:21.090 --> 00:20:26.220 active action frame where you name the problem, you say why it's important, 00:20:26.430 --> 00:20:27.750 and then you mobilize people. 00:20:27.750 --> 00:20:29.460 You ask people to do something. 00:20:29.840 --> 00:20:34.190 So we are looking at that exact same structure and wondering, and looking, 00:20:34.190 --> 00:20:38.780 and we will test, to see whether or not that particular structure is 00:20:38.780 --> 00:20:43.010 more or less likely to induce people to mobilize on behalf of the cause. 00:20:43.070 --> 00:20:45.800 So that's something new, another piece of that project. 00:20:45.920 --> 00:20:47.030 Matt Abrahams: That's really fascinating. 00:20:47.030 --> 00:20:50.720 Persuasion is so interesting, but I like that you're playing with, does the 00:20:50.720 --> 00:20:52.970 structure of the message actually impact? 00:20:53.120 --> 00:20:55.820 One of the things I'm researching and interested in is the role 00:20:55.820 --> 00:20:59.120 of in strategic communication. 00:20:59.120 --> 00:21:02.210 So if you think about it, everything I've talked about while I've been with you 00:21:02.210 --> 00:21:04.760 is about fidelity, accuracy and clarity. 00:21:05.120 --> 00:21:08.120 But often we use ambiguity to achieve our goals. 00:21:08.449 --> 00:21:09.530 Think about this. 00:21:09.590 --> 00:21:14.149 If I cooked a meal for you, I've already told you I'm a lousy cook, and I give 00:21:14.149 --> 00:21:15.949 it to you, and I say, what do you think? 00:21:16.070 --> 00:21:17.060 You could tell me the truth. 00:21:17.060 --> 00:21:19.129 It's awful, but that might hurt my feelings. 00:21:19.129 --> 00:21:22.520 So you might say something wonderfully ambiguous, like, I've never quite 00:21:22.520 --> 00:21:25.949 tried anything like this before, right? 00:21:26.250 --> 00:21:30.210 And so you fulfilled the strategic obligation of responding, but doing so 00:21:30.210 --> 00:21:31.710 in a way that doesn't hurt my feelings. 00:21:31.710 --> 00:21:35.070 So I'm really interested in times where fidelity isn't the goal, where 00:21:35.070 --> 00:21:38.070 we're purposely ambiguous and what that means for our communication. 00:21:38.070 --> 00:21:41.550 So thank you for all of you sharing, not just your insights into your 00:21:41.550 --> 00:21:45.330 research, but also for sharing ideas that we can all deploy. 00:21:45.659 --> 00:21:46.800 So this is the fun part. 00:21:46.800 --> 00:21:50.490 So I'd love to invite those of you who have questions to wait 00:21:50.490 --> 00:21:52.110 for the microphone to show up. 00:21:52.340 --> 00:21:53.900 I see a hand over here. 00:21:54.170 --> 00:21:54.470 Audience Member 1: Hi there. 00:21:54.470 --> 00:21:55.720 Thank you to all the panelists. 00:21:55.720 --> 00:21:57.560 Wonderful to hear all the different insights. 00:21:57.890 --> 00:21:59.030 My name is Gary Struble. 00:21:59.060 --> 00:22:00.560 I run a media company in Namibia. 00:22:00.950 --> 00:22:05.720 And just perhaps linking back into the AI question, you are all quite specific 00:22:05.720 --> 00:22:10.010 about answering it in a way about how AI makes your life easier, but I imagine 00:22:10.010 --> 00:22:14.780 AI is also going to make your lives much more difficult if it isn't already. 00:22:15.200 --> 00:22:18.460 And I perhaps just address the question to anyone on the panel, is from an 00:22:18.460 --> 00:22:23.504 academic perspective, from perhaps even strategizing as to how academia remains 00:22:23.504 --> 00:22:28.514 relevant in how it's currently structured, how are you grappling with the problems 00:22:28.605 --> 00:22:34.155 of AI in grading students and creating the funnels that academia relies upon? 00:22:35.205 --> 00:22:36.945 Christian Wheeler: Well, my challenges are perhaps a little 00:22:36.945 --> 00:22:38.595 different from some of the others. 00:22:38.745 --> 00:22:43.455 Most of the assignments in my class are self-reflections, and I want them to 00:22:43.455 --> 00:22:46.125 think about what that experience was like. 00:22:46.125 --> 00:22:47.415 What did they learn from that? 00:22:47.415 --> 00:22:48.975 How are they going to apply that? 00:22:48.975 --> 00:22:53.985 And perhaps not surprisingly, there's a non-trivial subset of participants or 00:22:53.985 --> 00:22:59.209 students who have AI self-reflect about an experience that the AI did not have. 00:22:59.710 --> 00:23:02.649 So that's a challenge and it's a little discouraging. 00:23:02.649 --> 00:23:06.460 So the, the solution to that is to have them do it in person 00:23:06.460 --> 00:23:07.629 and have them talk about it. 00:23:07.629 --> 00:23:10.929 But I think it is a subset of a more general problem that I think you're 00:23:10.929 --> 00:23:15.520 hinting at, is that people offloading the intellectual work that they might do, 00:23:15.520 --> 00:23:20.500 whether it's introspective intellectual work or creating a new product off the AI. 00:23:20.500 --> 00:23:23.290 And in my case, it's just, um, preventing introspection. 00:23:24.209 --> 00:23:28.830 Sarah Soule: I think another space that we worry a lot about this is in using some 00:23:28.830 --> 00:23:33.929 sort of AI tool to write academic papers, even if the research findings are in 00:23:33.929 --> 00:23:39.149 fact valid and true, and actual research findings, asking an AI to actually 00:23:39.149 --> 00:23:41.520 write the results up is worrisome. 00:23:41.520 --> 00:23:45.120 And so I think those of us who serve as journal editors, who serve 00:23:45.120 --> 00:23:50.669 as reviewers, have to be ever more vigilant to try to figure out ways to 00:23:50.699 --> 00:23:53.010 try to detect this kind of use of AI. 00:23:53.225 --> 00:23:57.365 I don't think anybody would worry too much if anybody used an AI to 00:23:57.545 --> 00:24:01.175 give them feedback on the writing, but to have an AI generate the 00:24:01.175 --> 00:24:03.875 writing is more problematic, I think. 00:24:04.475 --> 00:24:06.665 Jesper Sørensen: Yeah, I would just say that I think it's definitely 00:24:06.665 --> 00:24:11.445 a problem in the context of giving exams and so on and so forth, as 00:24:11.445 --> 00:24:13.305 Christian was referring to as well. 00:24:13.395 --> 00:24:17.865 I think professors and then teachers are still, I think, struggling with the nature 00:24:17.865 --> 00:24:21.675 of the problem, which I think actually has less to do with the technology per se 00:24:21.975 --> 00:24:26.565 than with the equilibrium we reach with respect to what an exam serves to do. 00:24:26.805 --> 00:24:31.410 And I actually think it's gonna require faculty and instructors everywhere to 00:24:31.410 --> 00:24:34.170 think more deeply about what they're trying to accomplish and make those 00:24:34.170 --> 00:24:36.510 kinds of assessments more meaningful. 00:24:36.780 --> 00:24:40.260 Because I think a lot of students, they end up using these kinds of tools to 00:24:40.260 --> 00:24:45.510 answer questions because they are not actually interested in the introspection 00:24:45.510 --> 00:24:46.820 or the learning or the development. 00:24:46.980 --> 00:24:50.615 They know that would be something they would get out of it if they wrote the 00:24:50.615 --> 00:24:55.445 work themselves, but they think that the whole exercise is just pointless. 00:24:55.504 --> 00:24:59.495 But that's a flaw on our part, and that's what we have to address, right? 00:24:59.495 --> 00:25:02.195 So there are some fixes that are around. 00:25:02.585 --> 00:25:06.814 Having people take exams by hand and in person and so on and so forth. 00:25:06.814 --> 00:25:08.735 And I think that's also part of it. 00:25:08.825 --> 00:25:11.135 But at the end of the day, I also think we need to really think about 00:25:11.135 --> 00:25:13.655 what assessment is and what we're trying to accomplish with assessment. 00:25:14.370 --> 00:25:17.010 Matt Abrahams: One of the things my co-teacher and I are thinking about 00:25:17.010 --> 00:25:21.660 doing is using the communication we have with AI as a way of reinforcing some 00:25:21.660 --> 00:25:23.160 of the communication skills we have. 00:25:23.280 --> 00:25:25.560 When you're typing in a prompt, you're writing a message. 00:25:25.620 --> 00:25:28.750 Now the audience happens to be an LLM instead of a person. 00:25:29.110 --> 00:25:31.740 But we can actually reinforce some of the ideas that we're trying to 00:25:31.740 --> 00:25:34.350 talk about when we actually have human to human communication. 00:25:34.350 --> 00:25:39.810 So actually using the way we interact with AI as another avenue to teach some of the 00:25:39.810 --> 00:25:41.580 skills is something we're toying with. 00:25:41.970 --> 00:25:44.970 Audience Member 2: Um, my name is Tana Tutu from Ethiopia. 00:25:44.970 --> 00:25:50.010 I'm in manufacturing, but by training I'm a psychologist, but I don't practice. 00:25:50.040 --> 00:25:54.300 But instead of saying I don't practice, I practice because I live with people. 00:25:54.300 --> 00:25:55.824 So I deal with people. 00:25:56.550 --> 00:25:58.710 So my question is regarding AI. 00:25:58.980 --> 00:26:03.120 I know it's going to take over some of the jobs and positions. 00:26:03.660 --> 00:26:09.330 And regarding having a chat with a chatbot, when you give it a prompt, 00:26:09.420 --> 00:26:13.350 you have to give it, you are a psychologist, a socialist like that. 00:26:13.635 --> 00:26:19.455 So for somebody who's lost their parents, can you just give them a prompt by, you 00:26:19.455 --> 00:26:24.555 are my mom, so you'll be speaking to me like this, speaking to me like that. 00:26:24.555 --> 00:26:29.475 So in a positive way, I see that in overcoming grief. 00:26:30.314 --> 00:26:31.845 That's one of the questions. 00:26:31.845 --> 00:26:34.185 I want you to give us a view. 00:26:34.485 --> 00:26:39.164 The other one is, in having a discussion with AI, like with the chatbot, 00:26:39.495 --> 00:26:42.195 how safe is our secrets with AI? 00:26:42.824 --> 00:26:44.925 You know, with humans, you're scared. 00:26:44.955 --> 00:26:46.215 Oh, I can trust that. 00:26:46.485 --> 00:26:49.395 And then there will be another version of it, which is gossip. 00:26:49.635 --> 00:26:52.895 So just give us few dimensions in that respect. 00:26:52.895 --> 00:26:53.345 Thank you. 00:26:53.745 --> 00:26:56.625 Matt Abrahams: So I don't know how to answer either of those questions except 00:26:56.625 --> 00:27:02.115 to say that my sense is as somebody who is always very concerned about privacy, 00:27:02.504 --> 00:27:06.555 that I would be very concerned sharing very personal information with it. 00:27:06.795 --> 00:27:11.865 That said, I certainly know that we are far deficient in the number of 00:27:11.865 --> 00:27:13.754 therapists that we have in the world. 00:27:14.230 --> 00:27:18.060 And people have lots of serious mental illness and challenges, and 00:27:18.060 --> 00:27:22.050 there might be an opportunity for AI to help some people in some ways. 00:27:22.050 --> 00:27:24.090 So we have to think about what that means. 00:27:24.510 --> 00:27:29.490 I can tell you from my perspective, I think there's great opportunity to share 00:27:29.490 --> 00:27:33.180 information through LLMs and chatbots. 00:27:33.180 --> 00:27:37.865 So for example, we're in the process of creating a chatbot for the podcast where 00:27:37.865 --> 00:27:41.885 you can go and you can ask a question just like you did, and it will scour all 00:27:41.885 --> 00:27:44.195 of the wonderful guests that we have. 00:27:44.465 --> 00:27:48.335 And it will say that question is very similar to something Jesper said in this 00:27:48.335 --> 00:27:50.015 episode, you might wanna listen to it. 00:27:50.015 --> 00:27:54.010 So I see it as an a mechanism to help people learn as well,. 00:27:54.040 --> 00:27:55.780 But I think it's really early days. 00:27:55.810 --> 00:27:58.389 I don't know if others of you have opinions on this, but I 00:27:58.389 --> 00:28:00.879 don't know how to answer those questions except to say they're good 00:28:00.879 --> 00:28:02.260 questions and we should look at it. 00:28:02.500 --> 00:28:03.850 Christian Wheeler: Yeah, I don't have a ton to add. 00:28:03.850 --> 00:28:08.290 I mean, there's research on, you know, there are a number of AI companions now 00:28:08.290 --> 00:28:10.419 that can take on various characteristics. 00:28:10.810 --> 00:28:13.959 The research shows that people get very emotionally attached 00:28:14.020 --> 00:28:15.669 to these AI companions. 00:28:15.730 --> 00:28:19.090 Whether that's a good thing or a bad thing, I don't know. 00:28:19.150 --> 00:28:23.020 I don't have research on that particular aspect, but for me personally, I will 00:28:23.020 --> 00:28:25.120 take a person over a screen any day. 00:28:25.570 --> 00:28:27.860 Audience Member 3: Uh, my name is Leslie Meringue from Glytime 00:28:27.880 --> 00:28:31.630 Time Foods, Zimbabwe, and we run the Kelloggs of Africa. 00:28:31.960 --> 00:28:35.860 But now my question is, if we are competing neck to neck with that 00:28:35.950 --> 00:28:39.790 particular business in a particular region, and you compare yourself with 00:28:39.790 --> 00:28:43.330 that particular business, does it not come with legal connotations to it? 00:28:43.935 --> 00:28:45.675 Jesper Sørensen: To legal connotations? 00:28:45.675 --> 00:28:46.365 Is that what you said? 00:28:46.514 --> 00:28:47.085 Yes. 00:28:47.145 --> 00:28:47.685 Yes. 00:28:47.774 --> 00:28:52.575 I would not encourage the use of analogies like that in a marketing context. 00:28:52.665 --> 00:28:55.034 Um, and that's not really what we were focused on, right? 00:28:55.034 --> 00:29:00.045 So we were focused on how do you concisely and effectively communicate what your 00:29:00.105 --> 00:29:02.085 company is about to somebody else. 00:29:02.085 --> 00:29:03.675 Now you need to be careful. 00:29:03.675 --> 00:29:08.325 So for example, I was once teaching some of these ideas to executives and, and they 00:29:08.325 --> 00:29:10.695 were a group of Australian executives. 00:29:10.755 --> 00:29:14.025 And so I told 'em, I asked them about a company that's based in Silicon 00:29:14.025 --> 00:29:16.665 Valley called Rover, Rover.com. 00:29:16.665 --> 00:29:19.935 And Rover is the Uber of dog walking. 00:29:20.115 --> 00:29:22.465 It's a very Silicon Valley idea. 00:29:22.505 --> 00:29:26.320 And so I asked them what, like when you hear the phrase the Uber of dog walk, 00:29:26.320 --> 00:29:29.800 none of them were familiar with this company, what does that mean to you? 00:29:30.070 --> 00:29:34.120 And one of the first answers was, oh, that means that I can get 00:29:34.120 --> 00:29:36.629 a dog walker on demand, okay? 00:29:36.629 --> 00:29:39.180 Because when you think about an Uber, one of the characteristics 00:29:39.180 --> 00:29:43.050 of an Uber is you pull out your app and you say, I want a car right now. 00:29:43.050 --> 00:29:43.860 And it shows up. 00:29:44.100 --> 00:29:45.990 That's not at all what Rover does, right? 00:29:45.990 --> 00:29:48.780 The analogy is about it's gig working, right? 00:29:48.780 --> 00:29:52.410 So they have a bunch of people who have too much time on their hand and not enough 00:29:52.410 --> 00:29:56.275 money, and a bunch of people who have too much money and too many dogs, right? 00:29:56.365 --> 00:29:59.335 And they're, they're trying to get them aligned with each other. 00:29:59.395 --> 00:30:01.315 And that's also what Uber is like. 00:30:01.315 --> 00:30:04.165 But again, so the reason you need to be careful with analogies is 00:30:04.165 --> 00:30:07.825 there can be that kind of miss because people have their own 00:30:07.825 --> 00:30:09.985 mental models of what's going on. 00:30:10.165 --> 00:30:14.264 So when you compare yourself to another company you need to be careful about what 00:30:14.264 --> 00:30:18.014 it is your audience is actually hearing when you're making that comparison. 00:30:18.014 --> 00:30:20.835 'Cause otherwise it can really lead to misunderstandings. 00:30:21.044 --> 00:30:23.115 Matt Abrahams: Thank you for being part of our live audience. 00:30:23.264 --> 00:30:27.254 And with that I end, I'll ask you to thank our amazing guests and I appreciate 00:30:27.254 --> 00:30:28.574 the questions that you all ask. 00:30:28.574 --> 00:30:29.475 Thank you very much. 00:30:32.955 --> 00:30:37.064 Thank you for joining us for this special Think Fast Talk Smart podcast 00:30:37.165 --> 00:30:42.465 episode, recorded live as part of the biannual Stanford SEED Conference. 00:30:42.960 --> 00:30:47.639 To hear another live episode like this, please listen to episode 194. 00:30:48.360 --> 00:30:53.670 This episode was produced by Katherine Reed, Ryan Campos, and me Matt Abrahams. 00:30:54.090 --> 00:30:55.830 Our music is from Floyd Wonder. 00:30:56.190 --> 00:30:58.410 With thanks to Podium Podcast Company. 00:30:58.825 --> 00:31:02.335 Please find us on YouTube and wherever you get your podcasts. 00:31:02.515 --> 00:31:04.675 Be sure to subscribe and rate us. 00:31:04.885 --> 00:31:06.985 Also follow us on LinkedIn and Instagram. 00:31:07.225 --> 00:31:11.635 And check out FasterSmarter.io for deep dive videos, English language 00:31:11.635 --> 00:31:13.975 learning content and our newsletter. 00:31:14.545 --> 00:31:17.845 Please consider our premium offering with extended Deep Thinks episodes 00:31:18.900 --> 00:31:24.735 AMAs, Ask Matt Anything, and much more at FasterSmarter.io/premium.