314. Why do we all love an underdog? PLUS! Responsible A.I., Ford's human rehiring spree AND the psychology of disagreements.

0:02Coming up this week in work. I know. You've heard of companies prioritising AI. But have you heard of any prioritising responsible AI? What the hell are you talking about? Well, apparently, according to Jake Chuba last week, that's what I sound like. Yeah, but don't forget you did a slightly Cockney Brummie accent as well. Yeah, it was kind of giving somewhere between Dick Van Dyke and... Dick. I don't know, Emily Blunt? Yes, that's a really good... Dick Van Blunt. Yes. Anyway, we're doing the intro, Leanne. Sorry, as I was saying, I know you've heard of companies prioritising AI, but have you heard of any prioritising responsible AI?

0:42Well, NatWest Bank are on a mission to equip their employees with the skills and confidence to use AI responsibly, all 60 ,000 of them. Is this just a PR piece or is it a step towards a more human approach to AI? Well, also this week, talking of AI, yesterday Ford backtracked on their AI rollout and have gone on a hiring spree for humans. Who would have thought it? I'll explain exactly what happened and why their AI-ification, is that a word? Why their AI rollout didn't work. And in truth or lie, we all love an underdog. Or do we? Since a certain US reporter had a pop at Bosnian Herzegovina, it feels like the rest of the world are rooting for them.

1:22Come on, Bozzy! So I'll be walking you through the science of why you're feeling this way very shortly. And in the workplace surgery, one listener asks, What do you wish someone had told you before you became a manager? They've been promoted without any training or asking us for some words of wisdom. This is Truth, Lies and Work, the award-winning podcast where behavioural science meets workplace culture. Brought to you by the HubSpot Podcast Network, the audio destination for business professionals. My name is Al. I'm a business owner and I'm here to ask the questions you might be afraid to. And my name is Leanne.

1:52I'm a child occupational psychologist and I'm here to answer them. And together we'll help you simplify the science of work and build amazing workplace cultures. Oh, we've got a new intro, Sean. We will be talking about Oblong next week because Sean O'Shea said, what the hell is Oblong? That's our company. We should be promoting it. Anyway, we'll be right back after a quick word from our talking promotion. Nice link. After a quick word from our lovely sponsor. You've all heard HubSpot though, haven't you? See you in a second. Hello, hello. Welcome back, Lee. Al. I saw something today. Now, those people who are regular listeners know that we live in Bosnia and Herzegovina.

2:24Aye. We specifically live in Herzegovina. I do. and I saw a great post. Now I don't know if it was a parody. I hope it wasn't. An American who tweeted I think. Oh it won't be. It won't be. I said basically it's not fair. Why does the USA have to play two countries when everyone else only has to play one? Why are we playing Bosnia and Herzegovina? Outrageous. What I love is that someone replied and said Bosnia in here? Let me just explain what happens. Bosnia plays the first half then we saw being Herzegovina for the second half. obviously hilarity ensued and this person got even more indignant and going, we USA!

3:00USA! USA! Yeah, so anyway, what else are you seeing? Yeah, I mean, the Americans aren't doing themselves many favours at the World Cup at the minute, are they? I mean, is it true that they basically deported Iran the minute they got knocked out? Did they? I don't know. I feel like I heard this. I don't know if it's true. Let's pretend it's true. This is an official, we're a proper podcast. Let's not pretend it's true. And of course, Abigail Velez, was that her name? the reporter who was like I don't even know where Bosnia is I couldn't point out a map it's like that says so much more about you than it does about Bosnia and Herzegovina yeah I did have a good bit of banter with Jake about that actually I'm not sure if you saw the map that he created for me you put it in the comment section of my post yesterday basically with the USA in the middle of the world map and explaining who everybody else is so basically the UK is where it rains, Canada is where it's cold and they see it a lot Australia interesting but very far away yeah yes anyway sounds like a lot of fun it's a lot of fun I think it's a visual you need to probably go see it really but the World Cup I didn't understand why I was forcing him to watch football since Sunday because it's a knockout now now I enjoy watching football I just know nothing about it but usually once he gets into knockouts you're like locked in yeah yeah so on Sunday when I was actually watching football you were like yeah if you want I was like oh no Japan Brazil or something you were like last night yeah and you were like I'm not fussed and then I only realised this morning that I'd been reading the Google tables wrong so Japan's out Germany's out Japan is out Germany's out shock shock shock and the Netherlands I watched both penalty shoots out this morning and they were really bad both of them were really bad like they might as well have just flipped a coin because nobody could put the ball in that net really yeah but interesting I did feel for Japan though I thought that they played really well against Brazil and Brazil just nicked it in the last minute, didn't they?

4:57Sorry, sorry, US listeners. It's not necessarily that we're like, we don't, we do like USA. We're like Americans. We're like a lot of Americans. USA, USA. It's just that Bosnia is a proper Goliath. Well, this is where we live as well. It's not just a random team we decide to support over America. It's where we live and where we call home. So, of course, we are supporting Bosnia-Herzegovina. and if the USA beats them well we wish you well we wish you well yes so Lee you have time for your favourite time of the week yeah let's do it cue the jingle it's my line honestly we're waiting one week son of a bitch stole my line did you know that was improvised I was just going to say that it was totally improvised yeah yeah brilliant oh we miss him we miss him we do what a guy okay so Lee what have you seen yeah so I saw this from Nat West how long do you reckon it would take to train every single person in your business about how to use AI responsibly.

5:53And apparently NatWest reckon that they're going to train 60 ,000 of their staff put through this mandatory AI and data training in about two to three months. Right, okay, good luck, good luck. It's ambitious, but I admire it. So yeah, this builds on something NatWest Group published a bit earlier before it hit the news. It was actually in Ivy's newsletter that I saw this. If you are interested in AI but find it massively intimidating, definitely go and follow Ivy Jobs. They are AI trainers and they're both founders have been on the show before. They're now funded by Stephen Bartlett but don't let that put you off.

6:29They're incredible at what they do. So go and have a look at their website. They do lots of free courses and sign up for their newsletter. Yes, a case study was done. A bespoke AI and data ethics program that NatWest ran for their leaders in partnership with the University of Edinburgh. That one was multi-day on campus led by Professor Shannon Valor. with academics from philosophy, informatics, and the law school all teaching the same room of people. So, leadership layer kind of got a deep version, and now they're rolling out the principles to everybody. So, Nat West, head of responsible AI. That's a job title I'd not heard of before.

7:05Dr. Paul Donga? How do you pronounce that? D-O-N-G-H-A. Trying so hard not to be puerile and stupid here. So, yes. So, Dr. Donga. Donga. says the goal is to equip colleagues with the skills and confidence to use AI responsibly as it becomes embedded in how the bank runs day to day. And this is a stat that actually I thought was quite interesting. Beyond the mandatory modules, over 50 % of staff have opted into extra training. I'm sure L &D practitioners listening will go, What? Really? Wow. Yeah, what do we think? Thoughts? Yeah, that makes sense. I think obviously when you're talking about dealing with something as important as money, then you do need responsibility.

7:52If you are like a building firm using AI just to make things a bit easy, you probably don't need to worry too much about the responsibility side of it because that's probably managed by a human. But this makes total sense. I am surprised there's 50 % opt-in. I'm wondering if that's been massaged a little bit. We don't have all of the figures. like maybe it's like oh you get a bonus because i think a lot of people generally more people are skeptical about about companies rolling out ai so i'd be i'd be surprised if it's just like 50 % of people just went yeah cool that sounds amazing but maybe it is hopefully it is well done yeah well that's how they presented it and on the flip side of that you could say that people are afraid of ai so actually the more i learn about it and the more i engage in this training maybe i'll be more skilled and if there is a round of layoffs i've done some training that other people haven't But Sun Tzu, where you know your enemy.

8:41Yeah. Exactly. Enemy close. Friends close. Enemy's closer and all that. Yeah. Yes. I thought it was interesting because there isn't really any regulation around this right now, even in the financial services sector. And we did see this at the beginning of the year, didn't we? We reported that there's going to be lots of behavioral scientists and psychologists being recruited into financial institutions to talk about ethical and responsible AI. So I think choosing to train the entire workforce, making it mandatory, and just in a few months, I think that's a really good sign from NatWest kind of saying this is how seriously we're taking it.

9:20And most importantly, how serious we're taking ethical AI and responsible AI. You had a story about NatWest about this time last year that we both liked. Was that something about how they were hiring people? I can't remember now. I have no clue. No. Genuinely no clue. I think my point is this isn't the first positive story like this we've seen from NatWest. So, yeah, I think it's really, really interesting. So they're going to be looking at how people actually use AI day to day, what they trust it with, what they double check, what they assume it's got right. And it kind of goes beyond that technology question, doesn't it?

10:01It's that behavior question. As much as you say if you're a builder using it, it might not be as relevant. I think in terms of accuracy is the real question, isn't it? Because we all over-rely on it a little bit and just assume it's right. I'm going to hold my hand up on that as well sometimes. So, yeah, I thought it was interesting. I like it. I will leave a link in the show notes. Our thoughts. No, done that. What do you see? It's going so well. Well, talking of AI, and just so you know, I don't see what Leanne's going to say before, and she doesn't see what I'm going to say before, so we'll make our notes separately.

10:35So I saw an article on AI about Ford. So basically, Ford is, as you know, Ford cars. It's a major employee who, for the first time, admitting on the record that AI didn't work the way they hoped. So here's what happened. So Ford have just confirmed they rehired more than 300 veteran quality inspectors. Rehired. So this happened after its AI-driven quality checks. I'm reading from this. Failed to match what those human engineers were actually doing. This came out in Bloomberg. but I saw it on BBC. And Ford's VP of Vehicle Hardware Engineering is a guy called Charles Poon, which, again, I've got to stop being so childish.

11:14So basically, this is what he said. Artificial intelligence is a fantastic tool, but it's only as good as the information you use to train it. Well, we all know that. However, what happened was, we said, and let me just double check, because it was about this time last year, Jim Farley said, AI will leave a lot of white... Jim Farley is the CEO. will leave a lot of white-collar people behind. And then in October, then Kumar Golotra, who is the COO, was boasting about deploying AI across the entire industrial system. However, and this is what they did. They did 900 AI cameras, so they catch quality issues.

11:50They did all kinds of AI stuff. And today, they're like, shit, didn't really work. I have to go and get those knowledge workers back. And I think they've rehired them, and I'm hoping they've rehired them on a bit of a salary call. I hope so. What are your thoughts? I enjoyed how you said COO. I thought you were surprised by that second O. COO. I was thinking CFO. COO. That's a great C, sweetheart. COO. He's the Chief Officer of Realisation. COO. Oh, I see now. What's funny? What did you say? I'm going to be thinking while you're talking about all this.

12:40Pull it together.

12:45Once every 90 days, I make Leanne laugh so hard that she cries. And honestly, it is my proudest, proudest moment of the year. And we've hit number one in management. We've hit number two in business.

13:02Cheap opposite realisation and some of your best work I'm trying to party I'm trying to pull it together I enjoy that Ford have taken accountability for their mistake and been transparent and embarrassed it I hope those people they were hired got a nice little pay bump and lots of other benefits they didn't have before I think it's just a good lesson if you rush this stuff then it's probably not going to work out very well so yeah good for you Ford I guess in the end yeah I guess I think the key thing what Ford have done is they got rid of all of their knowledge basically so they said right we're getting all these veteran engineers we'll get rid of them and replace them with AI but they forgot to ask the knowledge engineer and the knowledge workers what they knew and that's what they should have done now what we might be reporting in a year's time I'm saying they've just got rid of 300 of the new rehires.

14:01Because they've rehired them to retrain their AI. Yeah, so if I was going back, I'd be a bit like, hmm. Don't be suspicious. Don't be suspicious. If you haven't seen Parks and Recreation, that's one of the best lines. Leanne said that to me when I was buying her a birthday present in the supermarket. And she was shopping one side and I found her Hendrix Gin, which you don't often get in this country. I'm like, oh my God, this is going to make a perfect birthday present. So I started to walk past her and she came out of an aisle. so I just basically put my head looked the other way popped it under my coat and walked away and just ran you didn't just walk you ran and Leanne's like don't be suspicious don't be suspicious anyway Leigh what else you see me this week my love when you disagree with somebody at work do you think it's better to say it out loud or type it out oh oh I can see pros and cons for each which one would you therefore just would feel more comfortable for you I think I would rather type it out, send a message.

14:59However, I'd be worried that be received as someone who, as me not having the balls to look you in the eye and say, you laugh too much in the pod. Yeah, I think most of us would feel more comfortable typing it out. But according to new research I saw this week in the BPS Research Digest, that is the wrong call. Yeah, so it's written up by a wonderful person called Emily Reynolds, covering a study published in Nature Communications. One of my favourite periodicals, that. Me too. This was from a group of researchers at Imperial College London and UC Berkeley. Have you ever noticed how Berkeley is spelt?

15:37I wonder that Americans allowed that. We would call it Berkeley. Well, I don't know if we'd call it Berkeley. It's more that it's like, it's Berkeley. Oh. But they say Berkeley. But usually they're quite particular, aren't they? About too many, like, anyway, bashing Americans. Aluminum. Jake's not here. Sorry. Habits. We look forward to your letters. So anyway, yes, the researchers wanted to know whether talking versus typing actually changes how an argument goes. So they ran a series of experiments with more than 1 ,500 people over 1 ,800 conversations. Each person disagreed with a stranger about a hot political topic, some out loud, some in writing.

16:16and afterwards everyone said how well they understood the other person, how much conflict they felt and what they thought of that person. And the result held up every single time. So the people who talked out understood each other better, they felt less conflict, they liked the other person more and were more likely to end up seeing things the same way compared to the people who wrote it down. So the researchers found that talking naturally brings out more I hear you language and more of those kind of little signals that you're listening and you're open to that person's point of view you're not kind of just talking and waiting for your term your turn now oddly when people did use that type of language in writing it did work and it did work actually better than speak or speech but they just didn't really use it very much nobody kind of bothered to type because it felt so forced i hear you i'm listening type things but the thing that was really interesting in the final study the researchers asked people how they'd choose to handle a disagreement.

17:18So despite speaking producing the better outcome every time, 84 % said they'd choose to write. And only 7 % said they'd speak and 9 % said video call. So people think writing is safer because it gives you time to craft the perfect response. But the evidence basically says that instinct is largely wrong. Thoughts? Yeah. Do you know what? It's down to the whole fear of public speaking. people would rather die than go on stage literally um and so i think it's almost like a fear of conflict it is if you write it you can spend 40 minutes like i'm typing i'm typing you i don't know why i'm doing that if you write it you spend 40 minutes moving it around read reading it um i get someone else to read it i do this quite often with you i'll like i've written this email it's a little bit of a contentious issue can you reread it whereas when you're saying it you're just saying it so i can totally see how your preference would be to write it but actually i can also see why by saying it will make much sense.

18:17Because you get, you know, words are like 8 % and 92 % is non-verbal, that kind of thing. So yeah, I can definitely see that. I think you get that immediate feedback when you talk to somebody as well, don't you? So if you are being a bit blunt, you might see that eyes of the ray, the rays of the eyebrows, or the eyes widening, might make you rain it back a little bit. Because you say people typically just don't want that level of confrontation just for the sake of it. I mean, yeah, I thought it was really interesting. I mean, it is, of course, worth saying. there are some limitations these were strangers they were disagreeing about politics it was relatively low stakes so if it was kind of a heated disagreement with a colleague we might see very different results but I thought was very interesting is a gap between what people choose and what actually works so if you are a business owner I think that's a takeaway next time you're kind of tempted to to kind of send off a carefully worded email or a slack message to try and resolve of a disagreement, it might be better to just pick up the phone or jump on a Zoom call and have that conversation face to face.

19:20Interesting. Interesting. So if you are interested in this kind of thing, then we've got a workshop coming up in August, no, September time with the Right Honourable Steve Baker MP, the guy who pushed through Brexit, which ruined our lives. And I told him this. But what's interesting is he now has an agency with a guy called Paul and their sole thing is how to have disagreements at work. So we will definitely put in this research to him and see what he says. And also, we're going to have a real-life argument about Brexit. That would be interesting. I'm not, am I? No, I am. Okay, that's good. It'll escalate with you.

19:58Anyway, right, all the links are in the show notes. As ever, we will take a short break. And then after we come back with our Truth or Lie, we'll see you in a second. Welcome back, welcome back. it is time for our truth or lie. This is the time of the week where we take a popular workplace idea and ask if the research actually backs it up. Today's question, do we all love an underdog? Now, there's something about a person or a team or someone like the David and Goliath, if you're facing impossible odds, we root for them, we want them to win. But is it actually true? Is rooting for the underdog a universal human impulse or is it a story we tell ourselves that doesn't quite hold up.

20:38Leanne has done the research and she's here to tell you what happens. No, what happened? I love an undog unless they're playing against Liverpool. Fair. Which I think maybe gives you a hint as to where this is going. Of course, I need to start with what we mean by an underdog. So in the research, it does have a pretty specific definition. It's somebody who's got more chance of losing or the odds are stacked against them to begin with because there is less money, lower expectations, it's not their fault. So I think kind of the USA Bosnia is a good example here. Bosnia and Herzegovina are most certainly the underdogs.

21:18So do we root for the underdog? This is the evidence for, and the short answer is basically yes. It shows up in a lot of places. So the most direct evidence comes from Vandello, Goldschmied and Richards who ran four studies in 2007. In one, 71 people were showing Olympic swimming matchups between countries with very different medal histories and they consistently backed the nation with fewer medals. The same country gained a lot of support depending purely on who they were up against. So when Belgium was playing Sweden, Belgium was popular because Sweden has won more medals in the past. But when Belgium was against Slovenia, Slovenia, because Slovenia had won less medals, they were the favourite.

22:04So the same country gained a lot of support depending purely on who was up against. So Belgium was popular when facing Sweden, but lost support when facing Slovenia. In another study, 60 people read about the Israeli-Palestinian conflict, but were shown different maps. So one made Israel look large on the page. The other showed the whole Middle East, making Israel look tiny. When Israel looked large, 70 % of people called Palestinians the underdog and sympathized with them. When the map zoomed out and Israel looked small, 62 % switched. They called Israelis the underdog instead. So the same conflict, same facts, different map was the only thing.

22:48So that visualization of an underdog made people switch. Then we have a 2021 study, which is a bit more recent, that tested feelings people naturally tune into when they watch two people interact, one powerful, one not. And across six experiments with around 800 non-participants, people consistently focused on the emotions of the less powerful person. So in one experiment, 84 % of participants tuned into the powerless person's feelings and only 13 % focused on the powerful one. Interesting. and it crosses cultures. So Goldschmidt, Gailey and Keith tested underdog support across Israel, Japan and China.

23:30This was in 2018. And all three countries, of course, have very different values, different political systems. But in all three, people backed the underdog over the favourite. The size of the effect was almost identical regardless of where participants were from. It is interesting. So the obvious question is, well, why? Why do we do it? Well, Vardello's team got to this in one of their studies. They had 57 people watch the exact same basketball game, but told half of them one team was the heavy favourite. And the people who thought they were watching the underdog rated them as trying harder, even though they were watching identical footage.

24:09So it's that perception of effort, not talent, that made people like them more. And that connects with something that psychologists called effort heuristic. So it's basically a tendency for us to value things more when we think it produces, it's required a greater effort to produce. Yeah, Phil from Nudge has talked about this a lot. Open kitchens is another example. A piece of art that someone said, hey, I made this and everyone said, that's a bit of shit. And then someone else said, someone took 10 ,000 hours to make it and they said, brilliant. Exactly, yeah. So it's not just a case of admiring the underdog.

24:44we just instinctively read that struggle as proof that they're giving more effort and that makes us value them more. Cool. So that sounds to me like, as far as I'm concerned, I'm like, right, the underdog exists, the underdog syndrome exists. Yeah, it's never quite that simple though, is it? No. It's never quite that straightforward. There's nuance. So because the research, yes, shows we root for the underdog, it does also show that support has pretty clear limits. So it's not just always the underdog. So there was a researcher called Kim who ran a study in 2008 where they asked people to choose between a large established water testing company and a small two-year-old startup.

25:26When the stakes were low, for example, testing water for hardness, people backed the underdog startup. But when the stakes were high and they were testing for mercury toxicity, they abandoned the underdog and chose the big company. same underdog same disadvantage but there was higher personal consequences so when it really mattered the emotional pull of the underdog wasn't enough that makes sense yeah i mean the same study also found something interesting about art so people rooted for the underdog artist but when they looked at the exact same painting they rated it as lower quality when they thought it was by the underdog than when they thought it was made by an established artist.

26:10So we want the underdog to win, but we don't fully trust their output. Yeah, I suppose, right. So it would be like self-taught plasterer, for example, against a huge plastering conglomerate. Then you'd be like, oh, you know, I want the, I'm going to choose Bill because he's self-taught. He's just taught himself from the very beginning. Self-taught brain surgeon, not so much. Not so much. Higher stakes. Yes, higher stakes. Exactly. The same team also tested how automatic this response is, which I thought was interesting. So they showed 60 people a 15 second animation of grey circles moving across the screen.

26:48One circle struggled to climb a hill. Another moved very freely. Participants consistently rooted for the struggling circle. Despite it basically just being a dot. It's got no backstory. It's got no face, no context, no consequences. is the underdog instinct kicked in anywhere which I think tells us something quite important that it isn't a considered moral judgment this supporting the underdog it's fast it's automatic and automatic reactions tend to be fairly shallow so the research I thought described this in a really beautiful way it's a mile wide an inch deep so that shallowness shows up in a number of different studies as well.

27:30So a study from 2020, underdog support was real, but the gap between the underdog and top dog support only widened significantly when the financial disparity between the two sides was made very, very visible. So when the resource gap was smaller, it was still unfair. It was just less obvious than support for the underdog drop. So if, for example, you're watching Manchester City playing Chester, you're going to go, oh, well, Chester, come on, you know massive disparity here if you're watching city play united that's probably a bad example because there's different reasons for that isn't it city versus bournemouth yeah then it's not maybe they're both premier league sides it's not quite as much of an undog etc etc does that make sense makes sense yeah and finally a cross-cultural study found that support isn't unconditional so when a team had low expectations but it turned out they also had more money than their opponent people withdrew their support entirely so the underdog effect depends basically on the disadvantage of being genuine and undeserved so once people sense that disadvantage isn't real then the sympathy kind of goes with it it's a bit like you know when um when a struggling artist and he gets a big um a big record deal and then some people go oh sell out sell out yeah they were totally yeah totally exactly great example great example thank you very much great example So what are these studies telling us?

28:53Why does underdog come with all these conditions? Well, the researchers basically offered two explanations. The first is that underdog support isn't really about the underdog at all. It's about fairness. So as humans, we back the disadvantaged party because it feels like the right way to restore balance. And the moment the disadvantaged stops feeling undeserved, the kind of motivation goes with it. it was never really about them. It was about us trying to correct unfairness. The second explanation is that this is a shallow preference, not a deep one. So it was described by Kim and colleagues as a moral patina.