In Season 2, Episode 6 of UX Leadership By Design, host Mark Baldino chats with Bridget McMullan, Director of Technology Innovation at Elevance Health, about the intersection of artificial intelligence (AI), user experience design, and product management. They dive into the critical role of human creativity and critical thinking in effectively integrating AI into product development and design processes. Bridget shares insights from her diverse background in product development and her teaching experiences, highlighting practical examples of AI applications. The conversation also touches on ethical considerations, potential biases, and the importance of human intervention in AI’s development and implementation. Whether you’re a designer, product manager, or simply an AI enthusiast this episode explores the transformative potential of AI in creating innovative, user-centered design solutions.
Key Takeaways
- AI and Human Creativity in UX Design: The integration of artificial intelligence in user experience design highlights a unique overlap where AI enhances human creativity, leading to the development of innovative solutions that blend the best of technology and human insight.
- Human-Centered Design in the Age of AI: Despite the transformative potential of AI, the core principles of human-centered design remain paramount. This approach ensures that design solutions prioritize user-friendliness, intuition, and ethical responsibility, keeping human needs and values at the forefront.
- Ethical AI Use and Bias Prevention: The ethical integration of AI in design practices emphasizes the importance of bias prevention and aligns technological advancements with human values, ensuring that AI serves the genuine needs of users.
- Evolving Role of AI in Product Development: Artificial intelligence is increasingly shaping the future of product development, introducing new methodologies that streamline the design process and encourage innovation, thereby transforming how products are conceived and brought to market.
- Critical Thinking and Human Insight in AI Implementation: In the implementation of AI-driven solutions, critical thinking and human intervention play crucial roles, ensuring that technology is applied in a manner that is both practical and deeply attuned to user needs.
- Navigating the AI Landscape in Design and Technology: As AI continues to evolve, understanding its impact on design and technology becomes essential. This encompasses both the benefits and challenges of AI, highlighting how it can enhance user experiences while also presenting new considerations for designers.
About Our Guest
Bridget is a product professional who loves creating products that truly make a difference, no matter the industry. With a knack for guiding teams to success in product development, technology, and strategy, she’s currently leading a technology team at Elevance Health (BCBS) while also mentoring students at Northwestern University. Bridget specializes in helping teams understand what customers really want, making sure their products hit the mark.
Resources & Links
Chapters
00:29 Bridget’s Background Split Life
04:33 The Role of AI in Design and Product
07:23 Using AI Effectively
08:15 The Benefits and Challenges of AI
12:55 The Role of Humans in AI
23:00 The Risks and Limitations of AI in Healthcare
26:08 The Importance of Human-Centered Design
30:16 Balancing AI and Traditional Design Thinking
33:37 Continuously Improving AI Products
Transcript
Mark Baldino (00:01.678) Hello and welcome to UX Leadership by Design. I'm Mark Baldino, your host. I'm also the co -founder of Fuzzy Math. Fuzzy Math is the user experience design consultancy that brings consumer grade UX to business applications for B2B software and enterprise tools. Today, I chat with Bridget McMullan, who is the Director of Technology Innovation at Elevance Health. We start by getting meta and nerding out a little bit as Bridget explains how she ran a jobs to be done session with her team about what to talk about in the podcast today, which is just really, really funny. But then we get to the heart of the conversation, which is the insanely high demand for AI, AI all around us. How people in our roles can learn to use it, apply it in their own processes, and then ultimately, how we can best integrate AI into the products we design and build. And we center around the fact that for AI to be successful or successfully utilized in products, there needs to be critical thinking, analysis, and creative skills by human beings. Adding AI alone can't make a product more valuable to the people that use us. It's on us. We have a role in ensuring that AI solves real problems, delivers what people need, and ultimately creates something that people love. I think you're going to enjoy the conversation, and thank you as always for listening, folks. Mark Baldino (00:02.326) Bridget, welcome to the podcast. Bridget McMullan (00:04.48) Hello, thanks for having me. Mark Baldino (00:07.09) It is my absolute pleasure, excited for our conversation today. As I asked most of the guests, I guess all the guests, would love for you to give folks a sense of your background, how you kind of got into your current role. I know you have many irons in the fire, lots going on, and I think the listeners would love to hear. Bridget McMullan (00:27.121) Yes. Yes, I think most of your listeners are probably design, but my entire background is product. And currently, I do have kind of a split life between corporate working for a company called Elevance Health, that is one of the largest licensee holders for Blue Cross Blue Shield. So health insurance, and I deal with leading a team that does kind of an internal technology incubation. So just really fun, all sorts of projects, just it's the best and brilliant team. And then my other split life as I teach at Northwestern University for multiple classes. But the one that I'm doing now is one that actually you have come guest lecture at. It is an intro class to digital product development. So it's for juniors and seniors that Either they have background, maybe they don't, but they want to learn the whole process and we partner with professional clients and take them through that and it's really fun. If you went down my LinkedIn, I have a ton of product jobs. I've done consulting, startups, big companies. I've done physical product development, digital product development, you name it, I've done it. So I have a lot of experience on building essentially any product you could think of. Okay. Mark Baldino (01:49.09) That's fantastic. And I think I might have first met you when you were in the physical maybe and shifting over to digital, and then you were nice enough to invite me to your class. Was it last spring? It must have been last spring. Which is always like a great reality check for somebody who spends a lot of time with corporate clients. And I have designers who are, you know, junior who are learning, but that level of like... Bridget McMullan (01:57.631) Yes. Bridget McMullan (02:01.896) Yes, yes! Bridget McMullan (02:10.444) Yes. Mark Baldino (02:16.786) like mentoring people who are in school aren't doing it on the job. Like it's a whole other ballgame. So it was really wonderful to see and I appreciate you inviting me. And I appreciate you for joining me on the podcast. Bridget McMullan (02:29.32) Oh my God. And this was so serendipitous that when you reached out to me and said, do you want to be on the podcast? I couldn't have timed this better. A person on my team, I'm going to call out Kristen because she listens to this podcast and she's amazing, sent a team's kind of message to the whole group that said, hey guys, I've been listening to this amazing podcast. Here's some inspiration from it. Like we should be following it. Great nuggets in there. And I said to her, I go, is this Mark Baldino's podcast, the UX leadership? And she said, yes. I go, I'm going to be on that. And she freaked out. And it was just amazing because it just goes to show you. I know you don't know every individual listening, but a lot of people are listening and you're making a big difference. So kudos to you, especially my team. I was like, this is amazing. And then we did these really fun exercises where I was like, well, if I talk, what do you want to hear? What do you want to talk, what do you want me to talk about? And then we went even more nerdy because I am a nerd on jobs to be done. And so I want to share the three that she shared with me on what your podcast jobs to be done is before we get into AI. So she said, the jobs are teach me something new. That's why I listen. Validate, I'm doing the right things and specifically soft skills. A lot of like UX leadership soft skills. Am I doing the right things there? and then inspire me when I'm stuck, probably more in like creativity stuck in how to think. Oh, isn't that beautiful? Mark Baldino (04:03.622) Yeah, I mean, fantastic. Well, first of all, Kristen, thank you for listening. Thanks to everyone for listening and amazing that you, Kristen, that you brought it up to Bridget as we were talking about Bridget. Joining me, I also think, Kristen, to be quite honest, it actually maybe prompted Bridget to schedule it and get going on it. So I appreciate any push there. And I love the meta product digital design nerdiness of doing jobs to be done about. Bridget McMullan (04:07.732) Yeah. Yep. Bridget McMullan (04:16.44) I can't believe it. Bridget McMullan (04:24.5) No! Mark Baldino (04:33.41) podcast. It's fantastic. And it's the I mean, it's kind of it's funny. It's the process I probably as a I consider myself a practitioner of user centered design probably should be talking to my audience and asking them what they want to hear more often as opposed to assuming. So this is super helpful all the way around. Bridget McMullan (04:33.419) Yeah. Bridget McMullan (04:50.188) Well, podcasts are hard. I mean, when have you listened to a podcast and filled out feedback? It's not the same as like other channels, so there you go. Mark Baldino (04:58.033) Yeah. Mark Baldino (05:02.438) But hey, this is absolutely fantastic and I love learning about it. And I think I told you this when we were chatting pre-call, but I tell most people, like I've been doing the podcast for a little over a year now and it's been the highlight of, you know, those 14 months, like the conversations I'm having are just fantastic. And then, you know, we do kind of promote it on LinkedIn and just some of the conversations that happen that I'm not part of where people link back to it is it feels very rewarding. Bridget McMullan (05:05.048) Yeah. Mark Baldino (05:30.134) But what I feel is my job as a podcaster is really to give you and other guests a forum to share their knowledge. And I really do that first job to be done to teach me something new, give me some skills that I can take away is really something I want people to leave with. Because I don't operate my company or my life in like always my head in strategy. I'm really practical and pragmatic. And I think I want people to have that. when they listen, which is like, oh, that's an interesting idea. Let me jot it down. Maybe I can use it on my next project or... Bridget McMullan (06:04.568) Because it's really just as if you and I were having lunch and someone gets to listen in that doesn't know us. And if you get that at scale, I do think it's comforting to know like what are different leaders talking about? What should I be thinking about? And it's a lot faster than reading tons and tons and tons of articles. Mark Baldino (06:24.65) Yeah, that's for sure. And plus, if you can handle listening to me at one and a quarter or one and a half speed, you can also speed it up and they go a little bit faster, remove all of my numerous sort of ums from the equation. But I mean, you mentioned part of what we wanted to chat about in your role, which is, I don't know if it's an overhyped or underhyped topic in terms of AI, but I personally am trying to get Bridget McMullan (06:34.252) That's true. Bridget McMullan (06:43.989) Yes. Bridget McMullan (06:50.767) Ugh. Mark Baldino (06:53.926) more product folks on the podcast because I feel like the dynamic between design and product is it's still underutilized in some way shape or form. I don't know how it's possible working together as teams. We've been working together for years, but I just feel like the maturity of both of those organizations needs to grow at the same time for them to both be effective. If not, one is sort of, you know, has too much on their shoulders. So I appreciate you coming with kind of a product perspective because I talk to a lot of my designers about AI. Bridget McMullan (07:04.064) No. Bridget McMullan (07:23.445) Yeah. Mark Baldino (07:23.602) and how we should be using it internally, and how we should be helping our clients use it, and how, as with digital product designers, how do we actually wanna integrate it into the products of our clients? I think you're doing everyone a disservice in your sphere if you're not thinking about AI right now, and I don't mean hyping it or whatever it is on the curve of exposure. It's really like, how do we use this as a tool as digital product designers, as product management, how can I get the most out of it and how can I learn about it? So I'm kind of excited to sort of to talk about it. And so like when you're talking to your team or maybe personally, what are your, I like this, what is the recommendation for like, how do I make AI work for me? Bridget McMullan (08:10.721) Hehehe Bridget McMullan (08:15.604) Yes. So, and I think it's a split life, right? You have the, if you work at a company, startup, corporate, even as a freelancer, how do you use it with what you build for whoever your audience is? But then you also have the other life of how do you use it for yourself, for like productivity, right? And I would say that the biggest thing that I've learned with it is There are people, I think, that automatically hate hype. Automatically, they're like, it's not as great, it's stupid, and they sit on it. And they don't, like, and I would argue, I have had this personality in the past. Now I think it's something that is, it's so kind of ubiquitous in everything you do that you really should just jump on it. Like, just open up ChatGBT. Just try to use it. Would you actually use it? For our students, there's some really fascinating arguments on whether you should allow students to use AI in the classroom. And we're teaching a digital product development class. So we say, yes, you should use it. And here's the policy and here's the pros and cons. But I think it's one of those things that if you commit to it, make it work for you. Don't just do that thing that you tried it, you hate it, and you just throw it away. for designers, like the time you spent finding the perfect stock image. I mean, use AI to help that you get it faster, right? I even read on your website, you guys wrote about this. I feel like you wrote about a lot of like product recommendations. I'm a nerd and I do my homework, okay? Mark Baldino (09:56.454) We did write about it. You're the best advertisement for team Fuzzy Math right now. I love it. Listen, you're doing the research before the call. I appreciate it. We did. Because the team and a few team members in particular, not that they want to get caught behind, but they're like, there's some real power here and what can we do? And that's specifically in idea formation and generative text. And then you have the whole image. Image Dynamic, Image Generation, or as you said, like Image Curation, where when you're, I mean, the reason a generative AI, which is one sort of classification, what AI can do for us, but the reason generative AI is so popular is because it's doing stuff that takes human beings a really, really long time to do. And that's insanely powerful. And then it turns us into information sorters, which we're a little bit better at, right? Like, Our job, maybe we're not the most efficient at coming up with 50 variations of a landing page or 50, finding 50 stock images, but we are pretty good at sorting through those and using quality criteria to help filter AI. And that's kind of what I talk to my team. Bridget McMullan (11:09.652) Yeah. Well, and it's so, it's so fascinating too, because if you were alone in a room designing a webpage and you came back out and you showed what you thought it was, and then you also use an AI tool to do it, are they the same? Because could you have elevated your design if you also had access to what the spit out answer would be on this audience, this tone, this edginess, etc.? But I think you touched on a good point of like, we're the information sorters. And that's what I'm very passionate about, which is like the age, if we talk about soft skills, the age of critical analysis and spending your creativity on the hard stuff. Not the stupid stuff, you know? Not the stuff that is like. Collecting at the end of it. You can't you don't show a lot for it You know you formatted a Miro board to be really good to present some research. No, I mean, what's the good meat? That's the stuff that I think if you use these tools in a way that it helps get you there faster That's the stuff That you should really be focusing on and a lot of times like the biggest thing That is fascinating to me is like you're your own quality control with some of these tools You are the one, and now you've had X years of experience, your version of what good design is, could be something different as someone else with one year experience. Who's right? It comes back to how we live our day to day now. And so it's just very, very interesting to be critical of the tools in a way that it elevates, I think, what your outputs are. Mark Baldino (12:55.126) Yeah, I mean, I'd like to get to some of the... There's a cautionary tale in there as well. But first is that critical thinking component, right? That's why we have the brains in our heads. And like I said, can you get things, or as you said, can you get things faster? Can I get more alternatives out? And how do we help people on our team, not just use it, but think critically about it coming back and make decisions about it and applying like a human lens to the output. And the truth is, I think people are scared because there's a lot of unknown in AI, but we're doing, you do this with tools all the time, right? There's lots of data in, a little bit of data into Google for, you know, what, 25 years, right? You get a lot out and it's up to human beings to sort that information. I mean, there's obviously an algorithm behind it, but you know, to figure out which of these is the most, is the most helpful for me. Bridget McMullan (13:38.146) Yes. Mark Baldino (13:53.846) So it's not a pattern, I think, that people aren't used to. It's just some of the output, specifically the fact that you could say, give me 10 variations of analytic dashboards for executives who are trying to sort through financial information. You know, you're gonna get a bunch of, if you do that in Google images, you're gonna get a bunch of barely applicable images. But if you do that in an AI generation tool, you need some really interesting ideas, but that's not the hard part. The hard part is figuring out which of those apply in this situation and how do I evaluate those. Bridget McMullan (14:22.272) Yeah, well like an Excel formula, math is math, one plus one is two. But if you were like, what's the best design for my logo? Subjective. And I'm sure you have thought about this a lot too, but the bias, right? Like, you know, so many companies got slammed. I think there was like an Amazon story where they used an AI tool for hiring and in the algorithm there was a bias to hire more men. Mark Baldino (14:32.791) Yeah, yeah. Bridget McMullan (14:51.644) and they had to scrap it. But it's the same of like Google a doctor. Is it a white dude? Isn't it right? I don't know. And so there's all these things that like, you are the ones that have to assess whether it's up to your level of standard and that only is going to happen if you continue to test yourself, upscale yourself. Mark Baldino (14:59.945) Yeah. Mark Baldino (15:16.636) Yeah. Bridget McMullan (15:16.872) The other thing too I think is like IP how fascinating is an IP discussion with this because so many times on Anything that I make especially power product PowerPoints all day. You want to write copy for me? Thank you You want to find a stock image? Thank you You want to tell me what X leader would love to see because you pulled it from a ton of quarterly statements great But if I asked you for the source and you can't tell me what it is. That's a problem. That's a big problem Mark Baldino (15:33.252) Right? Bridget McMullan (15:46.74) Because you're not in the meeting with me when someone challenges me on a number, on a phrase, on a definition. I can't be like, oh, chat GBT told me it was this. I have to answer it. And so I think that's a really fascinating thing that I think, I'm sure it's changing, but I know Google Bard is really trying to be on top of that and chat GBT isn't as much. But it's even the companies that integrate it. Are they informing their employees where they're getting it? Mark Baldino (16:15.446) Yeah. Bridget McMullan (16:15.956) That's the powers, they get it everywhere, but then it's like, well, we can't actually tell you where we get it. Right. Mark Baldino (16:21.506) we get it. Yeah, no. All that stuff is like to be continued or to be sorted out in the future because people haven't, they're just starting to run into those. That's a great future edge case, which is what's the source of this? Who owns this? How credible is this? And if you're using it as a design tool, as we're a consultancy, right, and we're coming up with ideas, I mean, I guess we Bridget McMullan (16:29.173) Yeah. Mark Baldino (16:51.122) But the truth is, I think it's to the detriment of our clients not to be looking for accelerators in the way we think and operate. And we need to be able to come in credibly and say, hey, we used a third-party tool. Now we're going to kind of do a level of refinement. But we got to that initial result, that initial set of ideas, concepts, comparables faster because we used AI. Bridget McMullan (16:56.429) Yes. Mark Baldino (17:20.502) push comes to shove, if you have to defend that idea, it's gonna be hard to source. But I do think for designers in general, sometimes that level of subjectivity that they have in their designs is sometimes hard to justify. Which as you know, is why we try to focus on a foundation of user research and of actual kind of building a case and an evidence. And the way I tend to use AI is I very rarely ask it for a question. Bridget McMullan (17:30.604) Yes. Mark Baldino (17:49.046) I generally precede it with lots of information and then ask it to sort of be additive towards it as opposed to solely relying on it as a source of truth because as you know, if you've used it, sometimes it comes back with answers that just literally do not make sense or aren't clear. Then I find out the more kind of I give it and I have my own build to the model and my own addition to the model, I can then get a little bit better answers. Bridget McMullan (18:05.835) Yes. Bridget McMullan (18:15.316) Yeah, and I think too, that's why it's like you gotta find your people. Who are your, what are your credible sites? Who are your people that you can check in on and say, you know, how do you go about doing this? What's the best way to do it? Even your answer right there. If I told students and I said, you can use any tool you want, but you need to give me a source. And if... And if you do one iteration of it, you will most likely be wrong. So if you hand that in, you're not going to get as good of like a grade as you will of leading it, what you're saying. It's a tool, it's an assistant. It's not the only answer. Um, so very, very fascinating, but I know AI is there's a zillion types of AI and we're never going to cover all of them, um, in this podcast, but I did want to share like a personal use case example. And. Um, the takeaway and I'll, uh, I'll take out all the proprietary stuff and keep it vague, but I would say the takeaway was I learned so much more about it by just building like a project, by just getting my hands dirty. And if you do get a chance wherever you work to work with AI or generate AI or large language models, like just try it. Try it, you'll learn so much more than reading anything, than taking any course, and that's coming from me as an educator, unless you like get your hands dirty. And so the use case, so part of my team, we look about, I would say, six months to two years out. You know, we're not like an innovation group that does like, here's your five, 10 year mind blowing idea. And so we're a little bit more practical with what we do. And we were looking at, you know, kind of enterprise large language models. Cause I'm sure you have, maybe you wanted some specific information, let's think of ChatGBT and it says, say you wanted information from LinkedIn. I wanna know Mark Baldino's latest posts on LinkedIn. ChatGBT will say, uh, that's behind LinkedIn's paywall and you have to log in and I can't find that information. Well, every single company is dealing with that. So they're all trying to figure out like, what is their version of ChatGBT and how do they you know, utilize this technology to be better. And so the use case was when, since we're a health insurance company, when our customers call in to ask questions to our call agents, could we give our call agents like a chat GBT? What would that be like? Instead of like our existing solutions today. And one of the most fascinating things is, so we did this whole built it from scratch, research from scratch, the whole product development process. The pros, like the actual factual pros, significantly faster. So instead of a call agent searching a site, going through search results, summarizing it, shortening it, then figuring out how to speak to a person for it, that saved massive amounts of times. The cons, and we touched on it earlier, quality control is a nightmare. Quality control. when you don't have X question equals Y answer like math is so subjective on what you should include or not include. You know, I think about like, if you go to the DMV and you're like, what do I need? They either could like hand you a piece of paper and that's like the not helpful person or the super helpful person's like, stand here in this line, do this, do this, do this. So what's the line? You know, what should your benchmark be? That's impossible to figure out when you're essentially like starting from scratch, essentially. And that was really, really interesting as well as I would argue internally, we're a health insurance company. We can't get wrong if surgery is covered. But if you use chat GBT and it tells you the wrong recipe for your like chicken piccata, that's fine. Mark Baldino (22:10.286) Thank you. Bridget McMullan (22:31.872) The stakes are lower, right? And so I also think it completely depends on, you know, like how your industry is, right? And where it makes sense. And so I have so much more detail that I have gotten out of that, you know, let's say six months experience, then I would have taken 12 classes or read, you know, 800 articles. It's really doing the work. Mark Baldino (23:00.798) What without giving away anything proprietary, I mean, like the risk, as you mentioned in healthcare of it. I again, let's generalize, because I don't know what exactly this tool is doing. But if it's looking up deductibles or coverage, it's probably not maybe not giving health advice, but that sort of level of care, as you said, it's one thing to use ChetUT for personal reasons. It's another like a financial, you know, how much what's my insurance coverage or what's my return on investment if I put $10,000 into this group of stocks? Yeah, there's dollars and cents involved, but it may not be life or death. There are things in healthcare that are life or death or people choosing to have a procedure done or not have a procedure done or, oh, by the way, I'm $50,000 out of pocket versus zero. I've designed enough health insurance plan, screens, like those coming back and trying to break down all of the costs. I've probably done 15 over the course of 15 years at Fuzzy Math, and it's really, really hard to do it in like a graphic and a visual. Like, what did you all learn around kind of those risks? And did you have to set a tolerance level? As you mentioned, benchmarking, like what exactly was that process? Bridget McMullan (23:59.608) BOOM Bridget McMullan (24:20.564) Oh my god, it's such a good question. And this is my other big giant push is like, it was people. People ended up helping. Everything we do is a test because I live in more of the innovative state. And they were the ones that had to really help us. Their brains, their people brains had to really come back and say, this is good, this is bad. And then essentially you're looking at maybe your best call agent to your like, okay, call agent and looking at a pattern of what they're saying about similar topics and determining what the benchmark is for that. There's no AI in that. That's manual, right? That is people. And so I always say that you need to have this like subject matter expert promoting the fact of what this could do for you. And they saw it. They were like, we love this. We want this. But we also know you have to like really go through a lot of information to get it to the point that we want. And so I always say that it's the people. It's the people that are the ones that are ultimately going to make it the best based on what your quality standard is. And for us, it's very high. The highest. We have laws on how high it should be. Mark Baldino (25:39.266) Yep, 100%. And I think it's a good message, it's a good lesson that there... Like it's our first example of just like sorting information that's coming back and filtering it out. Now you're sort of training your model, but humans have to be a part of that process. And that's another critical thinking, assessment, review skill for accuracy of information that's going to be part of this. Bridget McMullan (25:54.646) Yes. Bridget McMullan (26:02.252) Yes. Mark Baldino (26:08.342) This isn't a pitch for AI, but let's let it accelerate things. It can accelerate and not be too scared of it. But let's also acknowledge that there's going to be a role for humans to assess and train and teach and validate. And then it's not just like we're replacing 100%. But we're going to have a vital part in making Bridget McMullan (26:17.921) Yes. Bridget McMullan (26:24.981) Oh yeah. Mark Baldino (26:38.218) that are AI enabled, us as designers and product management folks, we're going to have a key role in this. At least that would be my take. I think you'd share that. Is that fair to say you feel like we have a big role here in making better AI products? Bridget McMullan (26:53.761) Yes. Yeah, because even, you know, I can't think of names of it, but like software that can take your hand sketched, you know, low fidelity design, you plop it in and it makes a beautiful Figma design for you. That doesn't mean you're being replaced. That means I want you focused on is that design good? Do people want it? Do they want it or are they going to use it? And it comes back to the traditional design thinking. methods and processes withstand any new technology. That is a lifelong trend. What do people want? Why do they want it? And how are you going to give it to them that's better than anyone else? It's all what you do, right? Your company offers this. We certainly do it internal. We have custom research tools where we get to talk to real users. That's where you should be spending more of your bread and butter because that's the really hard part. Predicting the future is the hard part. Doing amazing technology and I hate this, but it happens a lot of, we have a hammer looking for a nail and I'm like, no, that's not what you do. You start with the problem and you start with who needs this problem fixed and then you go find the hammer. It's not the other way around. And it's just, it's... Mark Baldino (28:17.838) I mean, as you said, everyone's walking around with the AI hammer and trying to knock things down. Bridget McMullan (28:21.316) Oh my God. And I can't try to knock things down in it. And it's really just a matter of saying, okay, well, what is it? What is it really offering? Take out the buzzwords and the jargon, right? And, and why would the audience want this in the case of what I just told you, like these college, it's would love to get information summarized faster. Like they would love that they're on a phone call and they want it to be quick because nobody wants to be on a phone call long when they're talking to customer service. And I think it's just like, once you learn exactly what that is, then you have to cater it to that. And that's, you know, that's another giant people piece to this. Mark Baldino (28:59.83) Yeah, for sure. I mean, we have the tools in our toolkit, you know, design thinking toolkit, human-centered design toolkit, which are around hypothesizing what the problems are, doing some discovery around them, figuring out what the opportunities are. And then it's the same thing of what can the product do to solve for those opportunities, right? To solve for those pain points. And now we have another thing in the arsenal that's, you know, at the forefront of technology right now around AI, but it's really like, how do you mold AI or build a use case for AI that's actually, you know, that's centered around a foundation of a user need and a user goal? There's always going to be marketing and sales pushing for it to be part of the product because, you know, they really want to just put AI on the website or, you know, another bullet point of like, this is an AI enabled. But I think it is the responsibility of the humans of the designers as a product management, of those folks that are human-centered in their work, it is our responsibility, as you said, to make that product be useful for the humans who are going to end up using it and not do it for technology, you know, don't do technology for technology's sake. Bridget McMullan (30:16.34) Yeah, and I understand some of these technologies is like, you gotta start somewhere. And people are smart enough, I believe, to see the potential of it. But if you're starting somewhere, then you need to promise that. And you can't like promise the world, right? You have to make the match. I'm sure with your clients, they want the world and you're like, I gotta start with step one. Mark Baldino (30:36.45) I mean, if we can solve that problem on the podcast. Mark Baldino (30:41.879) We are in great shape. But I think you probably advocate for the same processes, which is let's center it on a problem, let's generate some ideas, and hey, let's go see if those ideas actually work in the real world or in a prototype or in a test or something. Bridget McMullan (30:57.472) Yeah, because even the case of like, you know, oh AI can predict that and predict this, unless some human is consuming that and changing their behavior, it doesn't matter. It doesn't matter what you add to it. Adoption, engagement, usage, paying for things, it all, that's the true validation, right? But I think embracing the AI piece of whatever that means to you, whether it's your productivity side or whether you want to put it in your products, it will only help as long as you couple it with some of these traditional methods, these traditional design thinking methods. But it's not going to hurt you. I think people were originally afraid of, and maybe in 50 years we'll be on CNN saying this and getting called out. But I don't know, I have more hope in the people designing it that we can make it work for everybody and be ethical and unbiased like we do with any product development product. Mark Baldino (32:10.55) Yeah, as you said, keep the focus on the humans, on the people, both those who are part of the process of designing and building these tools, but the people who are going to consume them. Go back to the basics of build products that people want, need, and are going to pay for and love. Like there's nothing wrong with that process, I think, but I think it's a good reminder for the audience that a lot of times when you're working in an organization, it doesn't always feel like everyone's pulling in that direction. Bridget McMullan (32:22.038) Yeah. Bridget McMullan (32:25.438) What? Mark Baldino (32:40.554) But that's our design lead, product lead, designers, individual contributors. It's our responsibility to advocate for that and kind of push for it. Bridget McMullan (32:49.856) Oh my God. And we all know, I mean, I know at huge corporations, especially designers are like, it could be this. And the engineers are like, no, it can't. And then product is like, but can we try to be in the middle? And, you know, and then, you know, you have marketing or, you know, money people saying, well, we only have this and all of those things too. I, the biggest thing that I always like to think about is even if the delivery of your first round of it isn't great, right. And all digital product stuff. Mark Baldino (32:59.082) You Bridget McMullan (33:19.704) Um, you should be focusing on that fifth and the 10th round. So at least you don't just like do it, throw it in the trash. You're doing it to get to something and you're just slowly making it better and better because as fast as we want to move, like it takes time. Mark Baldino (33:37.902) I like it. I think I found a good place to wrap up the conversation, although I could do this for another hour or two, but probably only Kristen would be listening at that point. Yeah, yeah, yeah. So I know you're on LinkedIn because I'm connected with you and follow you. Is that the place for folks to find you if they want to reach out? Bridget McMullan (33:43.936) Same, same. Bridget McMullan (33:48.318) Everybody has left except for Kristen. Thank you, Kristen. Bridget McMullan (33:59.436) Find me on LinkedIn. I'm happy to talk to everyone and anyone. And if you're a Northwestern student, take my class. And then you get a front row seat. We'll talk all about this. Mark Baldino (34:09.718) Yeah, in a perfect world, you could also meet me at some point. I might just show up randomly to your class at some point. So, well, thank you so much for time and energy today. It's been a delight to catch up with you and connect about AI. So much appreciated. Bridget McMullan (34:22.7) Thank you, I had so much fun. Mark Baldino (34:24.61) Fantastic. All right, take care.