205. Say What Sucks: Catalyzing Organizational Change with AI
41:05 Share

Searching for a job shouldn't feel like a full-time job, but for many of us, it does. Between figuring out if you're qualified and wondering if anyone else is reading your application, the process can feel broken. LinkedIn is changing that, helping you explore better opportunities without the usual frustration.

LinkedIn is reimagining the job search experience to help you confidently navigate your next move. I use LinkedIn regularly, and I've seen how its personalized insights can help all of us discover opportunities aligned with our skills and career goals.

This year, LinkedIn is rolling out new AI-powered job search tools that go beyond keyword matching. They actually understand your intent and suggest roles tailored to your strengths. One feature I'm excited about is Job Match, which shows how well a job aligns with your background and gives you actionable next steps. Dare to discover what's next. Explore your job potential at linkedin.com/jobs. That's linkedin.com/jobs.

When it comes to AI, catharsis catalyzes change.

My name is Matt Abrahams, and I teach strategic communication at Stanford Graduate School of Business. Welcome to this special live episode of Think Fast, Talk Smart, the podcast, recorded at South by Southwest. Many of us know about AI, and some of us even use it. But how do you bring AI to your organization and make it have a positive, productive impact?

This is what I have been curious about for a long time. So when my friend and two-time former guest Jeremy Utley asked me to facilitate a panel to discuss his AI implementation work with the MBA's Portland Trailblazers, I jumped at the chance to speak with Jeremy, Krista, and David. And I have to say, it was a slam dunk for AI best practices and learnings we all can implement.

So without further ado, let's listen in to our conversation on the South by Southwest stage.

Before we get started, I wanted to remind everyone that we have amazing resources on FasterSmarter.io, including recommended books, deep dives into topics such as anxiety management and spontaneous speaking, and content for English language learners for every single episode. Check out FasterSmarter.io slash resources. And to learn even more, you can subscribe to Premium at FasterSmarter.io slash premium.

Well, good afternoon. My name is Matt Abrahams. I teach strategic communication at Stanford's Graduate School of Business. I host a podcast called Think Fast, Talk Smart. It's all about communication skills. We're very excited today to talk about the particular application of AI within a business. This business is in the world of professional sports, the NBA team, the Trailblazers. And I am honored to be on stage with these

wonderful guests who have actually employed and deployed AI. We're here to share their experiences and best practices with all of you. With that, I thought we'd just start with quick introductions. I'll start with Jeremy, farthest away from me. Go ahead and introduce yourself briefly and share how you're connected to the team.

Hey everyone, I'm Jeremy Utley. I am an adjunct professor at Stanford. Been there since 2009 or so, teaching mostly design thinking, innovation, creativity, entrepreneurship courses. And then my world, much like many of you, was rocked a couple years ago when Chad GPT came out. Unlike most of you probably, I had just written a book about creativity

And a month later, this tool that's amazing for creativity came out. And I look in the index of my own book, the AI is not even in it. So I strapped myself into the front row of the classroom as a student. And I spent the last couple of years trying to get as close as I can to people like David and Krista who are doing this stuff in the real world and learning from them as much as I can and then sharing as publicly as I can. And so it's really fun to be here with all of you. And I thank Matt and Krista and David for allowing me to join the conversation.

Thank you. David, how about you? Yeah, I'm the VP of Digital Innovation with the Trailblazers and the Rose Quarter. We're a dual business for concert events and basketball. My team has run digital products, app, web, and arena digital products as well, and then digital marketing. And then recently we are taking on the strategy and implementing the strategy around Gen AI for our company. Thank you. Krista?

I'm Krista Stout. I oversee strategy and innovation for the Trailblazers. I've been there about 11 years and get to work with David and Jeremy on our AI strategy and implementation.

Excellent. Thank you. So many of us know how basketball is played and we see what your team does, but a lot of us might not understand the business behind basketball. Krista, could you spend a few minutes talking about the business side of what you do? And then Jeremy, I'd love to have you share how the Trailblazers came to Explore AI and get this whole project started. So we'll start with you.

Yeah, for sure. So our business is set up into three areas. The first area is the one that people probably think of when they think of the Trailblazers, which is our basketball operations. So that's players, health and performance, coaches, etc. It's one element of the business. Another element is actually our venue operations. So the people that put on the concerts and events and all that. And then we have our business operations, which for us is about 300 people per

And that is run in the same way that probably any organization you've ever worked at is run. HR, finance, sales, marketing, et cetera. So for today, we thought we'd focus on that part of the business operations because hopefully it's most relevant to you all as well in your AI journeys. And my boss, who's our president of business operations in August of 2023, took the time to go to this AI training and he came back and he was like, Krista, we got to figure out how to implement AI across our organization. Like it's the future, we have to figure it out.

And at the time, David and I had been, because we're in charge of figuring out emerging tech, we had been talking to Jeremy about a whole other thing. And we pivoted and we were like, so Jeremy, can you help us figure out how to implement and create a strategy for AI instead? He was like, yeah, let's do it. So David, AI came about as a result of your boss essentially saying this is an important thing to do. What were some of the burning questions that you guys had that brought you to Jeremy?

It's an emerging tech. It was something we were excited about, but there was no professionals on staff that understood how to take this and run with it necessarily because it's brand new. Machine learning, obviously, is something that's been around for a little while, but this from a generative AI side and being an accessible technology was new to us. So who's going to take that and run with it? We looked internally. We found people who could turn ideas into action, but we needed someone to be the barrier breaker for us. And the way I refer to that is someone who says,

We could show that it's accessible, that it's something that's fun, it's something really impactful, remove some of the fear from our staff around something new, and then just tear it away brick by brick, the wall that people might have put up around, "I can't do this, it's too challenging," and "I'm scared of it," like all these sort of things. And so how can we remove those, whether that's within one-on-one sessions, group learning sessions, sharing how you use AI personally was a big one for us. So that's the thing that kind of jump-started everything.

I want to come back to the change management piece of this because it's definitely challenging and I'd love to hear what you all did. But Jeremy, what excited you about this part of the relationship that you had with the team? These guys are the ballgame for me because we had already been collaborating, as Krista mentioned, exploring different kind of technologies or businesses or we were doing all sorts of stuff because we kind of had a kindred spirit in terms of our

our willingness to experiment, try new things. And I came to appreciate and admire the way they were approaching experimentation in the business. As Matt, you and I actually talked about this on another episode of the podcast, and we talked about that research there. But I had been privileged to be a part of this research program where a partner and I were basically studying how does generative AI impact creativity? And we found some kind of counterintuitive stuff, but I would say armed with those observations and insights about

how do normal people get the most leverage out of this technology? But there were a lot of ideas I had. When they came to me saying, "Hey, is there something here with AI?" To me, I saw an incredible almost sandbox and opportunity to collaborate with folks who I could test some of my ideas with. Because I already knew them and I knew they were the kind of people I wanted to work with, and because I had a bunch of ideas that I had studied in the lab, so to speak, it felt like the perfect opportunity to test some of those hypotheses.

Excellent. And David, I'm curious, how did you identify the problems first to apply AI to and how did you prioritize those? Yeah, I think after we had done some learning sessions with Jeremy and try to get the baseline up with Gen AI knowledge for our staff in general, we started to reach out to individual departments, well, all departments actually, and invite them to a lunch and launch is what we coined these type of practices. Lunch and launch. Lunch and launch. I like it, yeah. We would get as many people from the department as we could to get into a room with us

And the whole concept was to share with them, let's talk about your systems. And within your systems, what pain points do you have? Let's identify those pain points and let's start with one and say, this is the one we have to make it work. We want 10 pain points. We want 20 pain points. Because who knows, like the way AI works, we can solve for one and then potentially have a list of a ton more we can tackle next. Let's find that first one.

So we'd go through and we let people talk cathartically about their job and what bothered them and what could be done better. They could spend more time doing something else. And then we would take that, we would assign a co-pilot from that department to assist us throughout the process with the strategy in mind that keep these people as close to the build as possible because they're closer to the problems than we'll ever be.

So with them in tow and in helping us out, we would build utilizing AI to either build a software, build a Slack integration, build a simple GPT, that sort of stuff, and then pitch it back to them and then get their response. And then from there, we would then a level of measurements. We can see how it's performing under a couple of filters of, is this feasible and sustainable long-term?

Does it impact business efficiency, revenue, or fan engagement? Are they going to adopt it? And what checkpoints can we put in place so that there is adoption? So lunch and launches have been super powerful. We plan to do two builds per department for every department. Can I say one thing about lunch and launch? There's a clear role of branding. And Krista is a marketer. She understands that. I can't overestimate the importance of branding. And I've got an AI-focused podcast where I talk to AI leaders in different organizations. I think

I think about JJ Zhuang, who's the head of AI at Instacart. Their internal effort, they call the carrot AI team because the carrot is their mascot, okay? I talked to Bryce Shalamo, the head of AI at Moderna. They call their internal team the GCAT, which are the core components of DNA, but the generative AI champions team.

But I think there's a role of even thinking about the effort as branding it. It gives it a sense of credibility. Oh, you've heard of the launch and launch? You haven't been to a launch and launch? There's something to that that I think when you talk about change management, thinking about branding stuff is actually a meaningful part of it. And it sounds to me that not only did you take the time to brand it, but you started with people identifying their pain points. So all of a sudden, you're not coming in and saying, we're going to use this new thing to fix things. You had them share what

their concerns are, where their challenges are, and then had them thinking about how I can help. And what's very clear in the literature on influence and persuasion is that when you get people to buy into the problem early, they're much more likely to adopt and follow through. So I think that's beautiful that you did that. In addition, it sounds like you also came up with very clear criteria for what success would look like. So you'd build some things and then you had some clear criteria, which I think is a good bit of advice for everyone.

Did you do anything in particular to help prioritize which things you focused on first? We managed the project from a top level, right? We were running all these different lunch and launches, so we saw what was coming in and then we could use the same filters without everyone involved and say,

This one's going to really impact our revenue. This one's really going to improve customer experience. And then communicating that back to stakeholders. And we had a co-pilot along with us too, so we can utilize them to spread that message within their group. Like, it looks like the timeline for your build is in two weeks. Hope you're all excited. So like just kind of managing that, the expectations around this, because everyone was excited. But from an overall top down, what's most impactful for the business, that was something that we would match.

I love that somebody teaches strategic communication, that you were including communication throughout the process. That's really important to bring people along and to keep it going long after you've created that particular solution. I'm hoping each of you can give us a concrete example of something you did that is impacting the business. Chris, did you mind starting with something that you saw really impact the business?

Yeah, for sure. So the builds that David's talking about, we have about 35 of those that we've done across the company. And my favorite one currently is called David Detractor and his counterpart, Kelly Kindness. So I assume that you all, like us, send surveys to your stakeholders, to your customers to get qualitative and quantitative feedback.

We were doing a really good job getting and visualizing the quantitative feedback so that we could learn from it and implement it. But the qualitative feedback was much more complicated. So we dug into it during one of the luncheon launches and learned that a couple of different people were spending a combined almost 40 hours a week. So almost an entire full-time employee's worth of time just digging into the qualitative responses from our post-event surveys.

We have like millions of responses over the course of the year. Not all of them have qualitative responses, but people go into the system, read it, decide if they needed to send it to someone else or not, decide if they should respond, get approval to do a make good if they had a bad experience, et cetera. It was just like a lot of mundane work. And so David,

I guess, did you name this after yourself, David Detractor? It was the first one, so I think it was the easiest one for me. David built an alliterative tool, David Detractor, that ingested all of the post-event verbatims, filtered out the ones that we didn't need to respond to. When people were like, boo, you're like, okay, I'm not responding to that. But if people have a specific thing that we need to respond to, that would actually go to a specific Slack channel

where people that were relevant to that Slack channel, and I'll give an example in a second, could read it, put a specific emoji on it that then creates a draft in their Outlook outbox to send to that person. So before it took 40 hours of people's time, now it takes seconds and two clicks.

So my favorite example of this recently is actually this person who came to a game and really wanted a vegan hot dog, but the hot dog bundle didn't include vegan hot dogs. So she has this detractor feedback. It surfaces automatically to our head of F&B who reads it and is amazing. And she's like, "Hey, we should include the vegan hot dog in the hot dog bundle." So she makes a change in the operations of the business, hits one click, responds to the person, gives her an F&B credit to come back and to get a free hot dog.

and this person now hears back from us right away, right? So it works really well on the detractor side, as you can see, but it also works on the promoter side where we can surface really quickly

any positive experience someone has at a game, we service actually across the whole company, which is really nice because if you work somewhere and you have no idea how the experience is, it's really rewarding to see these and read about these positive examples that people have across the company. And on top of that, they're often like the warmest leads we could possibly have, like their hand raisers. Someone literally said the other day, I had the best time at the game. I want to come every week. And so we're like, hey, sales team, you want to call her? She seems interested. So

To close on this example, it drives revenue because it services warm leads. It improves our customer experience because people hear back from us. And it improves efficiency. So we basically cut out one FTE's worth of mundane tasks as part of this process.

It's a great example of how it was able to help you. F&B food and beverage. Sure, yes. Making sure everybody's following along. David, please, what's one of the things that you're proud of or impressed by? One other addition to the Kelly Kindness piece. One, name your bots. That's also a part of our branding strategy. People can refer to them easily. It's a pro tip. Name your bots and give them a human personality. Correct. I think a lot of talk around Gen AI is one of those concerns around disconnecting human to human connection because of

The use of these tools, this is a perfect example of how it's actually increased human to human connection. I think on the Kelly Kindness in particular, we're acknowledging folks who had really good experiences trying to solidify like a core memory, a core moment for them and build fandom. We'll admit there's areas where that was happening and very small scale, but now the scale for that is like immense because of this tool.

And so I really like to call that one out in particular. I recently was interviewing somebody else and they were talking about how they have built into their system whenever an employee calls in sick, they have a bot that automatically will send them chicken soup for their house.

And it's a way of showing kindness and showing that they care and that somebody is monitoring that you're not at work that day. So it does a whole bunch of things. So this ability to drive connection, I think, is important. Is there another tool that you are pleased about? Yeah, another named one, Billy Brand. So Billy Brand is trained on our Complete Content Style Guide, our Brand Guideline book, some history about the Portland Trailblazers.

Anything else that our brand team, PR team deems appropriate to be within there. And the main pain point we're trying to solve here is try to avoid these revisions. And I think that's one thing that this thing was trying to solve for is if you're having copy, if you're having creative, if you're having anything that's supposed to be public facing, let's have a tool that we can load it up. It can audit all that stuff for any brand things that are not aligned. Give that feedback, give suggestions on how to fix it. And so we're trying to cut revisions from

six, seven, down to one, two, three. Can we get it down so we're not spending time doing that stuff, moving these campaigns forward and moving the best campaigns forward? So Billy Brand has been really impressive. Just to make sure I understand the kind of economic impact, you're saying a typical, say, somebody in marketing or insert department here, they're getting feedback on how to align more with the brand voice, maybe six or seven times in some cases, and this tool is taking it down to two or three times. Yeah.

That's cool. We're all flipping upwards of 10 or 15 times. Wow. Six or seven. It's generous, conservative. The best for adoption purposes, it's still, this is one of the first ones I built. So I actually built it without someone helping me with it who was closest to the pain point. So it actually went through a couple different systems. One, it was like, it previously used to be within Slack, but people were like, I don't want to use it in Slack because then everyone can see that I don't know how to

do this or that. I was like, great point. So it's now a web-based software that we're rolling out to folks. And I had one member of my team who complained that he ran out of tokens the other day. And I was like, that's the best problem I could ask for. So I'm sure it's 100% brand compliant, that website. Absolutely. Excellent. Very good. Jeremy, what's one of the tools that you're excited about?

I'll give a non-Trailblazers example, if that's okay, just to broaden the aperture a little bit. But I've had the privilege of working with a bunch of, this is going to sound crazy, but a bunch of park rangers in the National Park Service, which is just super cool. They reached out somehow. I don't even know how. Hey, all of our backcountry rangers and facilities folks want to learn how to use this tool. Can you help? I was like, totally.

And we did some kind of basic foundational training. And one of the things that we focus on, similar to the conversation about pain points that Matt's drawing out of David here is what sucks about your job? What takes way more time than it should? And a really great kind of stem for finding opportunities is to finish the sentence, it sucks that dot, dot, dot. And we had people just think about that. What sucks?

And one of the folks in this group, it was a group of about 60 folks, he said, it sucks every time I've got to replace the carpet in the lodge. He worked at Yellowstone or Yosemite, something like that. Every time I got to replace the carpet tiles, I've got to fill out 10 pages of federal funding requests that include OSHA requirements and ANSI standards and historical heritage site removal preservation. I'm like,

a backcountry ranger, man. I don't know the answer to this stuff. So he built in 45 minutes a tool that could reference all of the relevant databases of information, including expense information, all that stuff, and would take a crack, a first pass at drafting the document for him. And it took him, he said...

whenever he said the last time he had to replace a carpet tile, it took three days to fill out the paperwork. This thing took 15 minutes. It took him 45 minutes to build it. So call it an hour in total, but three days minus an hour is not bad. But here's the really cool thing. When you codify these workflows and simple kind of shareable tools, the individual who builds it

Gets the benefit, but then anybody else for whom it's relevant also gets a benefit So in his case seems Adam I tried to tag him on LinkedIn to shout him out homeboy doesn't even have a LinkedIn account all right, so it's like this is someone who literally has no presence on the web no tech experience and Someone shared his tool. There are 450 parks across the US where there's a role like his and

The National Park Service is estimating that tool is going to save the National Park Service 7,000 days of human labor this year. Just his tool.

That's amazing. But the shareability of it, right? The fact that Billy Brand is shareable, it would be useful maybe just to you, right? But now it's useful to anybody who's trying to create brand-aligned communication. Which is everybody. Which is everybody, right? Yeah. So I'm hearing a couple things to take away from this. First and foremost, these tools can build connection, not reduce connection. Two, it's really important to think about where you place the tools.

and how you involve people in the process. Three, the shareability of these things is really important. And so I love the specific examples, but the lessons that we learn, I think, are really important. AI for non-technical people can be intimidating, maybe not for one park ranger, but for many people it can be. Krista, how did you help make AI accessible to your less technical folks?

Yeah, I think you said it as part of your recap. It was for us, it was figuring out how to connect with people and bring people into the process. So when my boss, Dwayne, said, hey, you got to figure out how to like have everyone at our company using AI. 18 months ago, I was like, I don't know how to do that. So the first thing I did was on a full team call, I just asked if anyone that was using AI wanted to come talk to me or be part of a

about how they're using it, what they liked about it, what they don't like about it. Let's just talk and then go from there. And so it's like tapping into people who were like, we had 35 people who were already using AI 18 months ago, who were then excited to share what they were using it for and whose job is not to create strategy for new tech across a business, but they got to be a part of that and got to help shape it.

I'm curious, Jeremy, what have you seen beyond the Trailblazers that helps organizations bring AI beyond just this as an IT initiative? How do you bring it to everybody in the organization? I don't think we have to look very far beyond the Trailblazers. I think they're like a great case study of creating space for folks.

creating venues and venues for sharing and for celebrating mechanisms for learning incentives where we can get into all that stuff as well. Right. But there's a bunch of pieces there. I think one simple thing that has really helped a lot of folks I talk to is when they say, I don't know how it's relevant to my job.

The kind of meta hack, which feels like a Yoda-ism but is not, is you can use AI to use AI. The basic idea is if you're not sure how AI can impact your work, you can actually pull up Chad or Claude or Jim and I. I'm not model agnostic, but I'm not hyping a model here. But you can pull up any of them and say, hey,

I have no idea how to use AI in my work. Would you act like a, insert LLM here, chat expert or cloud expert or Gemini expert or Grok expert, act like a Grok expert and interview me about my job so that you could recommend three to five obvious and maybe totally non-obvious ways I could use AI.

And you know what? It'll totally do that. The biggest thing is actually getting your imagination sparked. And most failure to use is actually a failure of imagination. And part of the value of mechanisms and forums like Lunch and Launch and like gathering people together is it helps broadcast and showcase a bunch of things that people go,

I would have never thought to do that. And basically what you want to do is create these forums where you kind of give these forehead slaps. I can't believe I've never thought to do that, right? And if you can provide enough of those moments and then celebrate how people are trying stuff, it's just kind of a snowball effect. I would say just to build on that, but one of the things that Jeremy unlocked for us that also helped with the change management was tapping into personally relevant experiences.

AI examples. And so rather than starting with how can AI help me do my job better, which is great and helpful, we started with prompts that really tapped into personal issues that people had. So not work appropriate personal issues. So for example, think of a decision that you have to make in your life. It's a hard decision. And then

Jeremy had a whole specific prompt that we just like copy and pasted into ChatGPT where we described the decision we had to make and some of the challenges and asked ChatGPT to interview us to get more context and then to work through that challenge with us. So I did it about my daughter starting kindergarten, schools and whatever. And it was so helpful, right? Like the kind of advice that I wouldn't, I just never would have thought that an AI tool could help me with that then sparked a million more ideas around how it could benefit me at work.

That was actually inspired by a life experience. We ended up doing a series of emails or Slack something, right? Where we basically said, here's a use case, here's a prompt you can copy paste, and here's a video of a professional nerd in California doing the thing, right? And the only reason we even did that, by the way, is because I had this true story about my, in my personal life. I'm riding with my grandma in the car one day. She lives in Oklahoma. She's mid nineties. I love her. I love you, granny, if you're listening to this. And we're driving the car. She's like, hey, what is this chat thing that you're working on? And I'm

Think about how do you answer that question? Your 95 year old grandma asks you, what is Gen AI? You know what I'm like, what's an emotional question you'd ask Faye Ann, who's her neighbor, that she'd ask Faye Ann about? And she goes, I thought this was technology. I go, just bear with me just for a second. She's, we don't even know how to think about assisted living. And I said, let's invite my friend Chad Gpt to the conversation.

And I'm driving, she's in the passenger seat. I just said, hey, my grandma just asked me about assisted living. I don't even know what framework to reference. I have literally no idea how to think about this. Before you give us any advice, would you ask her three or four questions so that you can customize your advice to her? And I said, sure. Have there been any changes to her mobility recently? And I hand her my phone and she's like, this is amazing. And I said, the tech's pretty cool, right? She goes, not the tech.

I've never thought about assisted living like this. Two days later, I get my favorite text message ever from Granny. Jeremy, we're out of cream of mushroom soup for the green bean casserole. Do you think your chat thing could help?

To Krista's point, the reason she thought of the work application was because she had this personal experience. And that made me realize, and then I just, all these random things in my life, we happen to have this opportunity. I said, hey, instead of starting with work, let's start personal. Let's give the first prompts personal so that people feel like they have this kind of imagination opening experience where work effectiveness or productivity isn't hanging in the balance.

That story is amazing for so many reasons. One, that your 95-year-old grandmother is texting. I find that fascinating. She does. And second, this notion of making it personal first to get people connected. And this idea of bringing people together to share at Stanford, where Jeremy and I both teach, they do this thing called Appy Hours. So people come together to share the different apps that they've built so that you can then learn to leverage it. And just the name Appy Hour, and they do serve drinks. It's a fun experience to share, building more on that creativity.

I want to pick back up on that notion of allowing for time for this. How did you at the Trailblazers actually give permission to people to take the time? Because I'm sure people are saying, I already have a full-time job. I don't have the time to do this work. David, did you do anything in particular to give people permission? I think the...

Couple along with what you all are just saying, I think it does start individually and looking at personal use of cases to with an end goal of like empowerment. Can this tool empower me? Because if you reach that level, then you stop thinking about replacement. You stop thinking about things that are negative connotations with general AI, and then you can

be a proud displayer of what you're able to come up with. This is what I did. Let me share it with my staff, whether that's personal or something work-related. They see it, and then instead of saying, "Hey, are you using AI?" So they're immediately put on the defensive.

You say, "Here's what I did. Have you tried this? Have you tried using AI for this?" or something like that. Just trying to rephrase it around empowerment and trying to get an end result has been super helpful. And just making the time is difficult, but in the end, once you start it, you realize that there are benefits of using it that cuts down on some of the manual labor that you have to do and allows you to focus on more important things. I refer to it as a utility. It's going to be a next utility for us. Implementing electricity, okay?

companies got rid of gas lamps. One of the most obvious things to do right away. But the people who really take it to the next level is like, how can they use electricity to improve their production lines, improve revenue, improve all those different things. So I think you have to do it because it is a utility. It's not a new flush in the pan.

Yeah. And we would also like, we had a Slack channel where we would just encourage people to share like, hey, I just took a picture of my lunch and asked it how much protein is in it. And guess what? It knew exactly how much protein in what I was eating and blah, blah, blah. And so we just were like constantly encouraging people to share how they're using it so that it is demystified and encouraged. And so the result of that is that David has someone that works on his team who just like

went off without even asking, built her own software that replaces an existing software that we spend a lot of money on today. And I think it's because we, like she knew top down that Dwayne, our president and others supported, it's part of our business planning process, et cetera. But also it's just so encouraged across the organization from that initial AI committee, from David's lunch and launches, like it's encouraged and celebrated. And when you celebrate something, people tend to want to do it.

The literature on motivation is very clear that if you put people on the defensive, they're not going to be motivated to do something. And it sounds like you've worked very hard to reduce that defensiveness and give people an opportunity and to celebrate, as you said, to help them. And that's really an important step.

Can we talk a little bit about the change management to actually get people across the organization to use these tools? It's one thing to build them. Is it simply that people see the benefits, so therefore they use them? I can imagine some people are really comfortable in what they're currently doing and the current way they're doing it. Have you done anything to help with the change management to keep the momentum going?

I think individual groups adopt faster than others. And I think the best thing is just to come back to the luncheon launches. Within those groups, there's people who have not used AI yet. There's people who have, but it's a shared space where we could talk about a similar topic and talk about solving problems.

And it's specialized to what their main focus is. So I would say that those are incredibly powerful. One, we are there to build something impactful. But for the shared space of communication and things like that, that's a major motivator for change management. And to go back to Jeremy's sandbox point, the David Detractor Kelly kindness example that I shared earlier. So when David first built it and launched it, it's not like people just started using it right away. People...

a lot of people didn't use it at all. So then we had another meeting and a conversation. We're like, hey, what is keeping you from using this? What would make it easier for you to use? And so you iterated that product for a couple months before, and now it's just like everyone's using it and it's everywhere. But it took months of iteration and learning and feedback and communication to get to that point. Were there any...

key revisions or key iterations that you feel like unlocked people's ability to use it? Like, what was keeping someone from using it? I'm just dying to know. It was me. I think one part of it was that I was not solving for the user's problem. So I needed to stop and overproduce something that I envisioned would be helpful

And that's, this was before a lot of these lunch and launches and big component of what we now implement, but getting with them and saying, does this actually solve your problem? And what ways, why, and what does it allow you to do? What impact do you think this will have? So that's the main thing. If you're solving for it by yourself, you're going to have revenge.

you're going to overdevelop and you're going to probably have less adoption. Matt, just thinking about your question of what drives adoption is kind of what you're getting at. To me, as I'm listening, I go back to the very beginning, which is everything's rooted in employee pain points.

Of course, people want to use something that's actually making their life better. But I think importantly, critically, David and Krista did the hard work of figuring out how their lives needed to be improved. So they didn't start with a broad mandate of let's just use Gen-AI in general, right? I think the usage metrics are largely irrelevant there. What they did is they said...

What are the problems? What really stinks in your job? Let's build a solution there. And then it just creates suction. It creates pull. This makes it easier. This makes it better. No brainer, right? Absolutely. And like, I love the, it sucks that prompt. That's a really good framing. It turns out people really enjoy talking about what they hate about their jobs.

And so we set up our sessions and then we would always have to cut people off at the 50 minute mark and be like, okay, now we're going to switch to solutions. You used the word earlier, it's cathartic for people to be like, oh, and if only, and for us to get that insight across the company of like the systems, like the key systems across our company and how they do and don't work effectively. And then the problem solver, those also gives us a lot of insight, which my hypothesis is that ultimately it will help us be able to figure out

where and how AI is going to significantly transform our business. Like right now, this is all incremental innovation across a lot of different work streams. But by getting this insight across the whole company, it is already opening our eyes into ways we can like really potentially transform the business more broadly.

Anybody who knows anything about me knows I love alliteration. So catharsis catalyzes change. I like that. For those in the audience looking to expand AI in their organizations, whether they're technical or not, I'd love to hear from each of you what's one concrete action they could take next week to start making progress with AI.

Why don't we start, Krista, with you. We'll just go down the line this way, if that's okay. What's one thing they could do? I mean, it's self-serving because this is what we did. But I think just recognizing that you don't have all the answers and don't need to have all the answers when it comes to AI and how to implement it. Like, you have to step and trust that the path will follow. And also admit that you don't know everything because literally nobody in the world knows everything about AI. So the idea that you would be expected to

It's a little crazy. So I think just admitting vulnerability, starting with curiosity and understanding like what's already happening at your organization so you can tap into the latent motivation and create momentum from there. Excellent. So start with vulnerability and follow with curiosity. Very good. David, what's one thing people could do starting next week? I think at the leadership level, critically about who within your company or under your team who exhibits like behavior of

taking ideas and turning them into action. I think that's a person that you should have a one-on-one with and present AI as an opportunity for them because that's a huge unlock if you have someone who can move things forward that way because this is a powerful tool in the right hands of someone like that. As an individual, like I mentioned earlier, do a complete audit of some of your systems. Whiteboard out, here's one for example, like every month I have to balance my credit card and so I need to know all the codes to send all these different charges to across marketing, across our corporate partners, all that sort of stuff.

I know it takes four different sheets. I now have to reference control F all these different sheets to find these different codes. It takes me two hours. I'm a cathartic like this is one that I absolutely do not like. And so within that, I was able to mark just a little carrots. Maybe I could use something that understands all of our codes. Maybe I can do something that can be accessible within Slack and answer it right away. So that's something that's taking it to the next step. But to answer your question, it's just like audit something and see what's possible and then go from there.

So the audit point is well taken, but you bring up something that we haven't really talked about, although you mentioned it, is buy-in from more senior leaders is really helpful. And taking the time to make sure they're on board can help, and in your case, to really drive the event. Jeremy, what's one thing these folks could do next week to make a difference? We talked about it earlier, but I think have...

AI interview you about either your life or your work to identify opportunities. Tell AI it's an AI expert, which by the way, a role is a critical part. If you've been playing with AI at all, you know this, you got to give it a role. If you're talking about a parenting challenge,

Hey, you're a child life psychologist with a specialty in childhood development in teenage girls. I have four daughters, so I've used that prompt a lot, right? But the point is, you're an AI expert. You're here to give me a consultation of how I can use AI better in my blank. Would you ask me five questions, one at a time, because I'm a human and struggle to answer more than one question at a time, please?

Something that simple. The other thing I would say about leaders, by the way, leadership buy-in is not some nebulous abstract thing. If you're a leader and you want to give buy-in, do something yourself and tell the team what you've done. Because to say, hey, y'all have permission to go do it is totally insufficient and it's far too passive. And the best leaders I have observed, they are actively showcasing what they're doing on Zoom calls. Let me share my screen for five minutes. I

That goes so much farther than, no, really, you're free to try it on your own time. No problem. What's even more important is to, as a leader, to share your struggles and challenges and failures because that gives permission for others to do that. Because it's one thing to say, go do it. And here's what I'm doing. People feel there has to be perfect, just like yours was. That can be challenging.

Before we wrap up every episode of my podcast, I ask some typical questions. Due to time, I'm just going to ask one question. We'll do it very quickly. The final question I always ask is, what are the first three ingredients that go into an effective communication recipe? And since there's three of you and I'm asking for three ingredients, just very quickly name an ingredient and then we'll wrap up. Krista, what's one important ingredient for successful communication? Oh, since it's his birthday, I'm going to say something that Jeremy does really well, which is turn complicated things

objects into very clear messages and communicate them very well. Make them accessible.

Constructive problem solving. If you have a problem and you want to present it to your leader or your staff, present it, but now you have this tool, AI potentially, where you can come with lots of solutions and you can flood a problem. So I think anytime you can come to a leader and say, hey, I have this problem, but here's some things I want you to consider about how I want to go about solving it. That's a completely different type of conversation. Lead with solutions. Very good. Conviction. If you don't believe it, don't say it.

Three very valuable bits of advice and important ingredients and lots of interesting steps and recipes today to help all of you be successful in deploying AI. Thank you very much for your time. I hope you're taking something of value away.

Thank you for joining us for this special South by Southwest live version of Think Fast, Talk Smart, the podcast. To learn more about AI and communication, please listen to episode 77, where I interview ChatGPT, and episode 134 with Jeremy Utley and Kian Gohar.

This episode was produced by Ryan Campos and me, Matt Ibrahams. Our music is from Floyd Wonder, with special thanks to Podium Podcast Company. Please find us on YouTube and wherever you get your podcasts. Be sure to subscribe and rate us. Also, follow us on LinkedIn and Instagram. And check out FasterSmarter.io for deep dive videos, English language learning content, and our newsletter.

Please consider our premium offering for extended Deep Thinks episodes, Ask Matt Anythings, and much more at faster smarter dot IO slash premium.

I wanted to share with you that over the past few months, I've had the amazing opportunity of talking to listeners across the globe about the impact the podcast has had on them. I love learning how people are applying the principles and concepts that we cover on the podcast and the impact that it has had on their lives. It is truly inspiring.

Thank you, and here's to another 200 episodes.