Tim Houlihan  0:08  
Welcome to Behavioral Grooves. My name is Tim Houlihan, and I'm here with longtime collaborator Kurt Nelson. Hello. And we want to talk about something you probably haven't heard much about, and that's AI. What rock have you been under? Right, right. Okay, so what we're doing is we're going to do two special grooves sessions on AI's impact on the human side of work, and this is the so this is the first of two episodes to cover that. We know that every organization is racing towards AI, right? There are all kinds of problems. I just read an article yesterday about how billions are being spent, and still like 40% of workforces are not using it. It's crazy, right? What's happening in AI? It's happening so fast. But the efficiency gains are real.

Kurt Nelson  0:57  
Sometimes, yeah, we can't

Tim Houlihan  0:59  
we can't deny that, and and so are the hidden costs to collaboration and connection and cognitive capacity. These are real problems that we have to have to talk about. And this episode, we're going to explore that research and how AI is quietly eroding teamwork. Ouch! Hard to think about that. But but Kurt, you have done a lot of research on this, and so we want to share your insights about this. About and maybe gives leaders some kind of science-backed tools to help them get ahead before it damages their culture and their team's performance, right? Yeah. So, so let's start. Let's just start right off the bat, with just when we're talking about AI with leaders, the the most common focus is on productivity and efficiency. But what do you think they're missing?

Kurt Nelson  1:53  
It's exactly what you talked about at the forefront of this. Is there are these hidden costs that are from the human perspective, the employee perspective, in how we think about AI, how that thinking then parlays into our behavior and the overall adoption. So there was some research out that showed, hey, almost 30-2% This was from KPMG, I believe. 30-2% of people are actively sabotaging AI's implementation, and and actively isn't meaning like they're going in and destroying the computers, but they are potentially doing things like, "Hey, I'm supposed to be using this. I I will act like I am, but not really. I will not take the output that is done. I will use my own component. I will negatively talk about those things behind people's back, so there is a cost, and some of that cost from the research. And there's some research is really interesting because AI's implementation into organizations, if we think about it, has only really been widespread for three years now, three or four years, and the research is just coming out about some of this, and so where the research typically falls is kind of around two two areas, and one is around team and collaboration and culture, those aspects of it, and there's another aspect about cognition and critical thinking and trust, and so there's a number of different factors, and we'll talk about both of these. I think overall, but my team has been working on this, and we are working with clients on how do you take this insight and make sure that if you are putting an AI system in, and you want your team to use it, that you're not having these hidden costs just reduce the efficiency, or actually even sabotage the implementation of all of this, because AI isn't going anywhere,

Tim Houlihan  4:17  
right? Except except deeper into our lives,

Kurt Nelson  4:21  
except for deeper, you know. There's this, I think, false belief, and you can agree or not disagree or disagree with me, that hey, the faster we can implement AI, then we will be ahead of our our competitors. And the fact of the matter is, is that your competitors are implementing AI as well. Yeah, it isn't about who gets there first. It's about who gets there better with more long-term sustainability in how they're using AI and that they're using AI right because some of these factors have long-term consequences. That aren't seen in the short term, that will negatively impact you as a company moving forward, and the companies that do this well will be the ones who don't don't have to pay that cost and are going to be much better off in the long term.

Tim Houlihan  5:16  
Well, let's talk about something that you you just brought up, and that is about collaboration and connectivity. From a behavioral science perspective, we know that we humans are really great in the collaboration field. It's our it's a natural instinct for us to collaborate, to connect with others. What's the impact that AI is having on this connectivity issue?

Kurt Nelson  5:37  
I'll give you a hypothetical. Okay. Five years ago, I'm on a sales team. We have eight, nine people on the team. I run into a issue with a client. Client is giving this pushback on what we're trying to sell, and I've never run into this before.

Tim Houlihan  5:59  
Okay.

Kurt Nelson  6:00  
What? How would I typically respond? How would you, if you were in that situation, Tim? How would you typically respond five years ago?

Tim Houlihan  6:06  
Well, a few years ago, that that would have been pick up the phone, maybe call a close colleague, send an email to a bunch of more senior people, and solicit their advice. Right? Yeah. Maybe maybe talk to my manager even.

Kurt Nelson  6:18  
Talk to your manager. Maybe it comes up at a team meeting, and and as a team you discuss it, or as you said, send out an email and go, hey guys, I'm running into this issue. Has anybody ever experienced this? And what was the result? Is all of a sudden you'd get a couple people responding back to you, or as you said, you'd call somebody up, you talk about it, and as that conversation happens, you're building connections with those other members of your team. Yes, you are a understanding. Oh, this person has some really good insights, and oh, they're funny, or they bring in this other aspect of we're talking about some other aspect of our lives where we have a connection. That's how collaboration and connections work. We build those interpersonal relationships. Fast forward to today, and instead of doing what we just talked about, it's easier in many instances, to just type in to whatever AI agent that you're using, "Hey, I'm running into this issue with this client. How do I handle this? And AI pops out a fantastic, wonderful solution.

Tim Houlihan  7:38  
Right.

Kurt Nelson  7:39  
You know.

Tim Houlihan  7:39  
Right. That might

Kurt Nelson  7:40  
be absolutely 100% right. Yeah,

Tim Houlihan  7:45  
yeah, and it's it's going to come with a lot less friction too, right? Because a lot. I don't have to ask anybody. I don't have to bug anybody. I don't stop them in their day.

Kurt Nelson  7:52  
It's going to be much faster, right? You're going to have it immediately, and then you can probably use that to prompt and like get some different pieces, which is great. That's one of the things that we like about AI is that it's fast. That outside of hallucinations and and work slap and some other things that we'll talk about, that there it usually gives some pretty good information, particularly if I'm knowledgeable enough about my own business and my you know the the person that I'm selling to to go oh this works but this won't work but I can take this and that's a really good way of thinking about this and now I'm I'm going deeper into that thought it's it's helping me kind of think through this issue on my own yeah that's fantastic what we lose is that connection to the other team members, and this isn't just happening on sales teams. This is happening on teams in general. This is happening from a mentoring component. When we think about, if I look back at my, the way that I kind of expanded into the world that I am, it's because I had mentors. I had people that took me under their their wing because I was asking them questions, or they saw something in me, and we just worked together, and we had that collaboration. Well, if I can just now go to AI when I have questions about those, there's less opportunities to mentor, or my feeling that I need to be mentored. There's also that connection with your boss. You're not going to your your boss with all of those different pieces, and that has a negative impact on a couple different things. One of them is culture. How does the organization think about this thing? Right. It also has impacts on on how we, you know, when we do face something that is not. And unique. How do we come up to that? It has impacts on motivation. We know, Tim. You and I know we're not just motivated by rewards and goals. We're motivated by the sense of belonging, of feeling recognized, and of having a shared purpose. And those elements decay.

Tim Houlihan  10:24  
Yeah. Okay. So you brought up culture. I think that this is this is an important connective tissue here to to the collaboration thing. It it certainly seems like with a more individualistically driven world that culture it could absolutely erode, right?

Kurt Nelson  10:45  
Yeah. So I use the term cultural fitting. That's it's this idea that cultures, when they're really well, are kind of thick, and they kind of they they're this guide to how we work here and the expected behaviors that we have, and and they are around us all of the time. And culture is shifting when we're using AI because of some of the things that I talked about. And this isn't overnight. This isn't like oh, we put in AI and tomorrow our culture is all gone. Yeah, we have a different culture. No, that's not how this works. Same thing about the teamwork and collaboration. You start using AI, that teamwork doesn't automatically decrease overnight or within a week or within a month or even within a year. We have those histories that we have, but new people get brought on, people move, and all of a sudden, some of that happens. Same thing with culture, right? We culture is shaped by the conversations that we have, and again, if AI is taking away those collaborative conversations, we're missing some of that, right? So the problem solving, mentoring, the

Tim Houlihan  12:09  
the the social, the the fact that you and I are both the soccer coaches of our of our kids' teams, those those that sociality. If I understand you right, that that when those things start to erode, we actually lose our ability to connect with the rest of the organization. Yeah, and

Kurt Nelson  12:27  
part of that is because there's less human friction. There's you don't get to rub up against those other people. Yeah, and at first that seems great, right? It's like I get the answer faster. I don't have to, you know, sit through these meetings where we're debating these ideas or doing all this stuff, but we're not learning. Where friction is where we learn. Friction is where we understand what is rewarded within the organization, what isn't rewarded in the organization. And I'm not saying that goes away. That that still happens, and there's a part of this that is still really kind of key to this, but that culture thins, so it doesn't have the same component, at least from the the work and the research that that we've seen. That's right. A new hire isn't going to get that same sense that hey, people who have been here for 510, 15 years have right. So because AI isn't going to inform them this is how your culture behaves, right? Yeah, and and you take that into account with the fact that many organizations are are using hybrid work right after COVID, and we don't have the in person meeting of people, which is bringing some of this culture thinning to place already, and you're just layering that on top. And so, you know, for this, for for both the the, well, let's let's talk about another aspect of this, which is work slop, because that kind of ties into some of this as well,

Tim Houlihan  14:12  
right? Because you you use the term work slop, you also refer to hallucinations, and they're different, right?

Kurt Nelson  14:17  
Yes. So so can

Tim Houlihan  14:19  
you just differentiate between a hallucination and work slop,

Kurt Nelson  14:23  
so AI AI hallucinates. Basically, it says stuff that isn't true. It makes things up.

Tim Houlihan  14:30  
Makes it up. Yeah, it's

Kurt Nelson  14:31  
a lying machine. It's not lying on purpose, but it basically hallucinates. And there's lots of research on this, and that can be very damaging for an organization. There's a number of times where, like, lawyers who have used AI and they cite cases that are not real. They never happened, and they get disbarred.

Tim Houlihan  14:56  
Right.

Kurt Nelson  14:57  
So that's a hallucination. They're just it's it's where. AI is coming up with a solution that isn't real.

Tim Houlihan  15:05  
What is work slop? Because I think this is maybe the the bigger issue.

Kurt Nelson  15:09  
This is the big well. This is a big issue.

Tim Houlihan  15:12  
Organizations

Kurt Nelson  15:13  
for organizations. Workslop is this idea that now when people. people within an organization, I don't need to have an expertise or a skill set or knowledge about a topic or a project or whatever it is that I'm doing, and I can just go into AI and have it produce something for me. And let me give you an example. Yeah, working with a client, learning and development team within this group, new employee brought on because they're good with platforms. They're not necessarily a learning and development specialist. Brought into the global organization in Japan, right? Large global organization. One of the things that's on their to-do list, way down on the to-do list, is we need to come up with global operating procedures for learning and development (GOP) Right? There is a standard operating procedure that's designed already for the U.S. 90-plus page document that that's there. New employee takes that 90-page U.S. SOP, throws it into AI, and says, "Create global operating procedures from this. AI does that. It comes back instantly. Comes back. It looks great. It sounds fantastic. All of these wonderful things because it's fluent and various different pieces, but it's not right. It doesn't work. It's not that it hallucinates, although hallucination might be part of it, but it doesn't take in the nuances. It doesn't have that expertise when they looked, this 90-page output that came out of it as well, right? They said, "Look, we got this done. They they handed it up to their boss and their boss's boss, and they didn't really think about it long enough, and they got got raised up to the executive team. By the time it got up to the executive team, they're excited about it. They're going well. Let's implement this. Well, now it comes back down to the global learning and development people, and they look at it and they realize this is all wrong. We can't implement this. This is you know the nuances in Germany and it are different in France, and the AI didn't address that. That's like China is not a homogenous component. There are multiple things that we have to think about within that. There's all sorts of things that are wrong because it's complex. It's nuanced. There's a bunch of inner knowledge that people have that wasn't brought into this, and the junior person didn't know enough to think about that, or to ask those questions, or to have that, and therefore, all of a sudden, you know, there's additional work. Now they have a number of people that have to be put on this. This got expedited by senior leadership. Somebody had to put their tail between their legs and go back up to senior leadership and say, "Oh, by the way, when we said this was done, it really wasn't. This was, you know, AI messed up, and now we can't implement it. So there's a lot of issues with that, and that happens quite often. That people are looking to help. They're they're not being malicious by this, or they're looking to say, "How can I advance if I just like bring in this new again using training? Like, oh, we've talked about having we need a training, you know, program on X. I might not know anything about X, but I can prompt AI, and all of a sudden, here's a here's the deck for it. Here's the training outline. Here's all the stuff, but it's not right, or it's not to the degree that it needs to be, and therefore now you're bringing additional people in to review it, revise it, change it, modify it, and as you and I know, sometimes it's much harder to update something that's wrong than to start from the very beginning and create something fresh. And so you're ending up adding work into the organization, slowing things down, creating tension within people. This is the other piece. Now those people who have to go back in and fix that issue, what do they think about that person, that junior person who just had this?

Tim Houlihan  19:53  
Right. Do

Kurt Nelson  19:53  
you trust that person?

Tim Houlihan  19:54  
Yeah, their reputation is going to be short-circuited at best. Yeah.

Kurt Nelson  20:00  
Or that executive who designed, hey, I got the five courses here that we've been talking about for years, and I was able to make them in minutes. And you're going, yeah, but they're not right. And you know, do you do you respect that person like you thought you did? So there are a number of those factors that come into play when we think about work slop. So

Tim Houlihan  20:22  
yeah, for me, this cascading effect of building extra work in for other people in the organization is the most insidious part of this. Is because we're using AI to save time, improve efficiency, improve productivity, and yet the opposite can happen with, especially in this work slop environment. Am I am I getting that right?

Kurt Nelson  20:46  
You are 100% correct there, and I think there is those cascading components. Now, as AI gets better and better, work slap will probably decrease because the output will be be better, but it's never going to be 100% given the nuances and context of an organization and all of those factors. And you want to make sure that the people who are building things with AI have the knowledge and the skill set to be able to look at the output with a discerning eye, with the knowledge to be able to go, that's good, but it's not quite right, or that's a hallucination. That is not anything that is even there. Right. Or I like this, but there's another way of thinking about this that might be better for this, and those are the things that often get lost, as you talked about, when people who are putting this in place don't have the skills, they don't have the knowledge, they don't have that expertise, and so they're just pushing stuff out, and that stuff isn't is going to cause again additional work down the line, or not meet the needs of the organization in the ways that it could.

Tim Houlihan  22:10  
Yeah. Okay. So we've we've talked about the negative impacts on cultural and and and cultural thinning. Yeah. The decay in in the collaboration aspects, and then and then work slop. Is there anything that you want to add about about any of those three topics that we've talked about, Kurt?

Kurt Nelson  22:29  
I think here's the important thing: is that all of those can be mitigated. All of those costs can be mitigated if an organization is purposeful in how they approach this, but you can mitigate collaboration things by teaching managers how do you set things up so that you have more opportunities to interact. These, and this isn't just an AI thing. This should be a like a you and I have talked organizational

Tim Houlihan  23:08  
development thing.

Kurt Nelson  23:09  
Organizational development. We know that teams that are closer together share more information. Are you know help each other out in those situations. It's better organisationally across the board. Well, you can develop things like instead of at your team meeting, you can say you can talk about the AI prompts that you've used and how they've worked. So then you can get people discussing like, yeah, I used AI in this situation about this issue that I had with a client. You can build that in as part of the process. You can create connection points with team members. Team building, like I used to do, you know, a long time ago, is in a formal way. But even informal team building and making sure that your team members are connecting on a interpersonal basis as well as just a technical. This is the you know job that we're doing, but get to know people as individuals. You can design those things in, and we have come in and help companies do that. Same thing on cultural thinning. There are things that you can do as an organization. You can build those in purposely, design that, making sure that you are focused in on those aspects of the culture in order to make sure that that fitting doesn't happen.

Tim Houlihan  24:44  
You're talking about changing the environment in which people are doing their work,

Kurt Nelson  24:48  
setting up the environment to support this, making sure that leaders communicate things in a way that. Drive that understanding of culture, setting up connectivity points within the organization, so that that culture, the desired culture that you want, gets reinforced. One of the things that companies are not doing that I think they need to be doing, and we can, you know, we've helped companies do this, right? Is set up an AI philosophy and governance system, not on the technical side of things,

Tim Houlihan  25:33  
right?

Kurt Nelson  25:34  
An AI philosophy and how people need to use this, and various different things. So there's a whole bunch of those factors that come into play, but if you put in an AI philosophy in a governance, you can curtail that. It's not saying you're going to eliminate all of the costs, but you can be purposeful about it and design it so that doesn't happen.

Tim Houlihan  25:57  
Absolutely. So, so in our next episode, we want to get to trust? We want to get to how some of these, how the leaders and the managers have to work with this. The the kinds of strains that are on on managers, as well as the the way leaders can actually sort of navigate their way through those those things. I I think those are going to be some pretty big things, and and we're going to get to the the cognitive surrender issues. I think in the next episode as well because that's yeah that's absolutely critical to it. Anything else that we need to cover right now?

Kurt Nelson  26:31  
I'm sure there's tons of things

Tim Houlihan  26:33  
we could cover right now, but no, because it keeps changing every day.

Kurt Nelson  26:37  
It does keep changing, and so I just just for people who are listening, if this is an issue, feel free to reach out because we do have a number of different solutions around this. Everything from just working with your executives to inform them of this, or doing a keynote to your key leaders within the organization, to helping to doing a an AI human diagnosis, understanding what are those issues that are happening within your organization, and what is unique to you as as your company, to the philosophy and governance piece, to putting together training workshops that for your frontline managers, so they understand what they can do and how they can ensure that collaboration doesn't decrease. That they can focus in on making sure that work slop isn't happening. That they can preclude that cultural thinning and a number of other factors that come into play.

Tim Houlihan  27:45  
Absolutely. So, grooves, we hope you use these fantastic ideas that Kurt has just shared with us today, and use them this week as you go out to help you find your groove.

Transcribed by https://otter.ai