Navigating UX in the World of AI

What we're looking at today is really a, you know, relevant topic in the UX and research space in general, but navigating the UX or navigating UX in the world of AI. Um, I have the pleasure of introducing Dev. Uh, so I'm Brett Kjuski. I'm the VP of research here at Accelerant. Um, I have the pleasure of introducing Dev. Dev and I worked together for around three years or so. Uh, formerly at Walmart. Uh, he's now at First American Insurance. um and he's a director of design and research there with a plethora of uh titles to go along with what he's doing and and all of his credentials academically and he is working on a PhD as well. So, I'm sure he'll he'll add that in the near future. But, uh Dev, I'll throw it over to you. I know you have a hard stop at four. So, >> yeah. Yeah. Thanks for having me. Um, I'm gonna go ahead and share. I put together like a little deck and um, Brad, if you could, I guess, manage the um, the questions and then just interrupt me if they're coming through. >> Sounds good. >> All right. Can you see my screen? >> Perfect. >> You can. Okay, cool. So, as Brett mentioned, um, we're here to talk about navigating UX in the AI world. I know this is like a hot topic right now, and to be very honest, I'm just gonna put it out there. Um, this is similar to like like when anytime there's like an industrial revolution, um, we're kind of living through it right now digitally. It's happening so fast that the things that I'm going to show you and we're going to talk about are I'm not going to act like I'm an expert in this. I'm going to show you what my perspective is as a professional that's been doing this for 16 years and what I'm seeing. As as um, as Brett mentioned, I'm finishing my doctorate at the University of Florida, which is R1 institution. There's a lot of research going on at at the academic level regarding AI. So I am privy to what's going on there, but I don't know everything, right? In fact, I don't I think anybody that that claims that they know what's going on or where AI is going is, you know, probably illinformed, right? It's really silly. With that, um, I do want to do a quick intro as, as Brett mentioned, um, my name is Devon Singh. I go by Dev. Formerly worked at Capital One, Facebook Honeywell Jewel Verizon Walmart. But I I have with me my friend Ranju who works in the right now. Reju, do you want to do a quick intro? >> Yeah, sure. Um, hi everyone. Uh, my name is Ranju. Uh, yeah, I um joined First American recently um after interning for them as a product design and now I'm a product researcher. Um, I'm currently a grad student at NYU and I'm wrapping that up right now and I'm working on my thesis. So yeah, excited to be here. So, one of the reasons I pulled in Runju was that, you know, I'm I'm kind of like an old dog in UX, right? Like I've been formally trained. I've been doing this for 16 years. This is my 17th year. And what Runju represents is a fresh a fresh perspective, a fresh lens. She's at NYU right now. She's finishing up her last quarter. She was she's actually been engaging in AI even before a lot of the corporations were doing it, right? just like so I think by having like the interplay between someone like myself who's a veteran in the field and someone like her, it's going to make the the discussions more just fruitful in general. So that said, I I will just say this. I always do this with all my talks. I give a lot of these talks um every year, but we are represent our um ourselves. We're not representing an employer organization. Although we do obviously are employed by a specific organization, anything that we say here is just our own view. With that said, um the way we're going to do this is we have um a several topics and with each topic you're going to see a blue screen like this and I'm going to try to make this as interactive as possible. I know it's a pretty big group, but um first to the audience, right? Like what was your initial reaction to AI? So when you first heard about AI and you know whether it was the doom or the gloom or even positive, what was your initial reaction to AI? So type it type in the chat. Um, and I can't see the chat, so maybe Brad, you can you can maybe just summarize what you're seeing or >> Yeah, I'll I'll let you know when people start interacting. >> Yeah. So once the audience kind of answers a question, what I'll do is um, Runju and I will kind of have a chat around this question. So, okay, I can see the chat now. So I see skeptical then amazed disappointed seem like a lot of hype same as Anna. Yeah, I mean I think I mean I'm I'm seeing the word skeptical hype, right? You you often see this, right? Anytime there's like a change in technology. Um, as I mentioned, I did work at Facebook before they were known as Meta back in 2015. I still remember very vividly back, I think it was 2016 when they were talking about like the VR worlds and this was from the presentation then. Um, I think it was 2016 and and Zuck had talked about like, you know, what if we can use something like Oculus Rift to engage with the world, right? And I remember I was also skeptical of this when it first happened. Um, cuz my first question as a as a UX folk was like, well, what is the job to be done that we're solving here? What does it solve? Or are we just like slapping VR into something? So, my initial reaction to AI was the same reaction I had back in two like 10 years ago to to VR was that, well, what is this solving really? Right. So, I was very skeptical at first. Obviously, since then, it's I don't think anybody could have predicted how fast it grew, right? And how much it's changed the landscape or even the market pressure. But I was very very skeptical at first. But I'll pass around you. What What was your initial reaction? >> Yeah. Um, I think for me definitely at first from like the more personal standpoint it was a little I mean it was very like I guess Fahrenheit 451 is what was always coming to mind for me of like that book that talks about like technology and its sort of perance in like the way society de develops around it. So that was me grappling with that for sure. But then when it started entering like the UX space, um, honestly I was really surprised at first because I always felt like UX is so human centric that it there's a little bit more of a barrier of entry for that. So I was like it was utilized but like I really only started like fully diving deep into like how we could use it um during like my work at in my masters. Um, I took like a course on experience design and AI and that really helped me grapple with like the ability to like use this thing but also maintain like the ethics and building on like the existing processes that we have to kind of keep like to keep design like an empathy focused thing. You know, I feel like that's it is possible to do that and balance it. It's just you need to be intentional about it. So that was my sort of reaction. Yeah. I noticed your image you chose was about like chat bots. You want to talk about that or like natural language processing? >> Yeah. Um I feel like it's such an interesting like I think the thing that's scary is how fast these things develop and how fast these things are like actually able to I guess keep up and create effective like results. Um, like I mean when I think of like a year ago, like if you're typing into cloud and you're asking like something versus like now it's like gotten to the point where you can pick like it'll give you options you can select from and like make its decision based on what you're selecting and things like that. And so there's such a rapid growth that I don't think I think even the internet like there was a little bit more of this like steady but like you know it was growing but this is like I don't think any of us have seen this before. So, I think that's a part where for me it was like at first there's like possibility to it, but I think that possibility like has that other side of like being intentional with that sort of >> Yeah, absolutely. >> So, kind of moving on, um I do want to address the elephant in the room. Um at this point, I think most people in our industry have seen this talk by um the design lead from Anthropic about like don't trust the design process. So I I had to I have to bring this up because it's like on every feed in LinkedIn that I that I see or even see social media. And if you haven't seen it, right, it is a pretty cool talk. You can you can just YouTube it. But um some of the key takeaways that that Jenny Wen says is that essentially like traditional design process feels outdated. UX should pair with engineers to make prototypes with real data, right? And interactions and to consider like a design some sort of fusion of a design product tech generalist. And really the end goal is to kind of shrink the road maps. So this was this is something that's been kind of like you know all over the internet right now. Um right now I think it was reported by like Lenny's Lenny's podcast has a thing on it that was really for from a PM lens right. Um and then NNG the Neil Norman group they finally they did respond like a week ago or so and they had a really cool article um and you can click these links. I'll share this deck so you can like click the links at the bottom left. But really what they said was design thinking process was never meant to be a rigid checklist, right? All it was is a visualization of the stages that people tend to move through, right? They don't always move through, but they tend to move through, right? And instead of thinking about it as like, you know, I to challenge UX folks to consider like process compression rather than abandonment, right? So one of the things if you think about if if you take if you take both these arguments like side by side right one saying like the design process feels outdated and the other one saying like maybe but instead of thinking like let's abandon that let's keep it but also let's just think about more of a compression. They do have some like the researcher in me is thinking like there are themes that you're seeing here right they both are talking about the need to move faster the argument to move faster about shrinking road maps compression right I want to stop for a second and talk to the audience real quick like is the design process dead so when we say design process I'm think like the double diamond right the traditional design process what are your thoughts based on this argument right here so feel free to like just type your answer in the chat that and then we'll kind of So Hannah said nope not all dead just faster. So yeah Ann Hannah's essentially essentially agreeing with NNG. Anybody else? I I think it's adapted and that's something we discuss internally as a consulting agency is a lot of the stuff that was taken on by more junior researchers is now being shifted to AI and a senior researcher can do that plus the junior role right with become a superhero with the AI capabilities. So not necessarily dead but adapted and and expedited. >> Yeah definitely. Johnson says, "Definitely helps to save time, compress the process." Yeah, I mean, my my take I I'll take a stab at it. My take is that I um I think there's truth in both, but I do align more with the NNG. I think that I've always viewed the double diamond and these frameworks as they are frameworks, right? They're not meant to be like, oh, I have to I'm like a robot and I have to move in this. Anybody that's been doing this for a long time knows that like there are times where you have to skip the first diamond. You might be looking at like quant data or previous data, right? And you might have to speed things up. There's times where you do generative research and you might just launch something and test while it's in production, right? There's there's the real world doesn't work so rigid like the double diamond, right? So I think like there's theory and then there's there's application, right? So I I agree with NG in the sense that it was never meant to be like a rigid checklist. Whether you use like the Stanford design school or you're using like the double diamond or like the IDO process. There's so many different frameworks, right? But in general, the idea is that like you have a framework, it helps ground yourself, right? And I do think that AI helps compress it rather than like abandon it. I do think that if we abandon a framework, then it just becomes the wild west, right? And doesn't really solve the actual issue. RJI, what's your what's your take? Yeah. Um I totally agree with that last part you said of like it becoming the wild west when you drop those frameworks. I think I mean I think we can see that when like one thing with all of this is that UX has become more accessible regardless of like like um the sort of training or background someone has. But I think the difference between when you can see like a seasoned UX or like approach a like solution versus someone who's kind of getting started but has AI on their side is those sort of little things where I mean design itself is a process and it in I think empathy is that key part in that process and I feel like that's the thing that's like it's it's rather just being packaged in a different way and it I think it's more important than ever now. So yeah. No, and and I think the article like the NG article is really it brings up this good point of like the these like double diamond frameworks were made to kind of like capture what a seasoned designer and like UXer like thinks through to like approach like a problem and solution. So yeah, like I I agree with the point that like these frameworks are important. It's just that the they're more fluid in the in the way it's shaping basically. >> Yeah, absolutely. All right, topic three. So in the new AI world, right? What remains the same in the design process. So let's assume that okay, we're not going to let's assume that we are going to stick with some sort of process, right? But audience like in the new AI world, what remains the same in design process? What are some of the things that are kind of like table stakes? as UX practitioners, we just can't afford to. It's too risky for us to abandon. Let's see some Justin says things should be tested amongst consumers. Absolutely. We don't want to like just like create it in AI and just like ship it, right? like we want to there needs to be some sort of some sort of like feedback from the actual end user methodologies could largely remain the same but use AI assist. Yeah, I think I would that's a really good point. Um that I was I was talking to one of my friends and that works in UX and he said that like think about instead of thinking about like AI as this like scary like robot like it's Terminator 2, right? That's out to get us. Think of it more of as like it's it's a tool, right? How can we use it to just be more efficient in what we do? if you're a designer or a researcher or a content strategist or even like a PM, right? Or engineer, like how can you use it to assist you? So, for for these next couple conversations, I'm going to keep the double diamond up only because it's, as we know, it's like the most popular framework. I'm not saying that's the only framework. It's just the one that it seems like a lot of companies use. But I'll take my stab at like what I think is going to um stay the same or should stay the same. So we think about start from the from the left to right. Um my view is that I think that in the generative phase what needs to maintain is that we need to have discovery right like because just because I have AI doesn't mean that AI might help me you know look at what like if you have a team that has like a bunch of research that's like set up in a repository I might use AI in in the very beginning phase where it says discover to search that repository almost like a library to find out what's been done already, right? Almost like a secondary research, not research I'm going to do myself, but what's what's my team currently done so I'm not repeating it. So that's something that teams do right now in some form, but it might be a little more manual. I think AI will speed that up and make it more easier. In the event that I have to do research, I think that we do need to maintain going out and doing some sort of discovery, right? Because AI is not always going to help us, especially with hallucinations. I was talking to my one of my good friends um that's an attorney right here in the New York area and he was telling me like in the legal field he was testing out just like anecdotally using various like like AI tools to to look up like legal text for New Jersey and and New York and there were so many hallucinations like I think he quoted like it was like 40% of the time it would be like a hallucination that it just quoted the law wrong or interpreted the law wrong. So we do need to go out and still do some form of discovery right um in that first time and I think that's super super important because again generative research make ensures that we're we're solving the right problem right or we're building the right thing I think that's going to be super important and then to to the point in the chat I think when we move into that second phase whatever we use in AI to create something really fast right um we probably need to maintain some sort of iterative quick testing, right? I think um I'm not going to jump into like what's going to change. That's the next question, but Ron, I'm going to pass to you. What do you think is going to remain the same in the design process? >> Yeah. Um I think some similar to what you were mentioning for the generative phase, I feel like the process of like identifying the problem I think is always going to matter and always going to be important. um within this process like like for example like theming like on theming as an individual like on our own and like interviewing I think this is where I feel like it's important to bring in that human involvement and like keep that like sort of that final part the same because I think it's where like our ability to like listen and identify creative patterns I think it's I think when you give it like from what I've seen like when you give AI like that sort of task like let me say like right now like it's it like identifies solutions but it's not like but there's it's missing that sort of human creativity I feel um where it's like absolutely able to like you're able to interpret what you're seeing and like similar to what you were mentioning with the law I think it comes here where it's like properly interpreting what you're looking at and like creating some a solution that like is I guess creative is where it'll like that process will stay the same. Um yeah. >> Cool. So, topic four is the inverse of that. What do what do you think is going to change in the in this new world? So, think about the double diamond or whatever framework you're using, right? So, so chat or audience, what do you think is going to change for the better, for the worse? So it says I think nuance will be lost when it comes to data interpretation for worse even more time pressure to generate designs quickly. Yeah. Yeah. For the first point I think new nuance will be lost when it comes to data interpretation. That's interesting because like yeah, there's more than ever we're going to have like an information of overload and you know it's just like drinking from a faucet sometimes, right? Um regarding this the time to generate things quickly. I think I think yeah in UX we all are constantly under pressure like a time crunch to get things done. It's just going to be more pressure to just go faster, less time. Allocate the thinking. Creativity will be lost if we don't continue to keep the human perspective. Absolutely. Here's my take. Um, so I mentioned what I think is going to stay the same. I think again just to rehash is I think the general phase needs to largely we need to like make sure that we're doing discovery and some form of testing. I think the big changes are going to be that in the general phase because we can create concepts so fast like faster than ever there there might be a lot more method methodologies that you know in the older days we might or I say the older days meaning like you know pre-ai right we might traditionally do like some like a survey or a contextual inquiry right where we're just going out and like asking people questions and doing observations but I think they're going to be there's going to be more generative methods that deal with concepts like like you know um concept tests where it's like hey I'm going to put a design in front of you. I want to see what your reaction is. It's still going to be generative in nature but it might utilize a concept because of the fact that like I want to be very careful here. I'm not saying concept first design right what I'm saying is there are many different methods in the first diamond that you can utilize artifacts to extract generative insights. That's what I'm talking about. Right? I think that's going to become more prevalent in generative research. I think moving into the second phase, you're going to see, you know, the development of designs much faster, but not just regular designs. Like the thing about the double diamond right now is that the development phase is after the second diamond. I think those two are going to happen more parallel where it's like AI will allow us to build a prototype that will also be coded at the same time. So while you're doing that loop of of of evaluative research, you're also kind of programming it. It might not be like 100% programmed, but but it's there's going to be some code involved with like real data. And I think that's what's going to happen. So when the NNG says it's going to compress, that's what I mean. It's really going to be that second diamond that's really going to compress a lot. What are your thoughts, RJ? >> Yeah. um pretty similar where I think my interpretation is like simple but like polished con I feel like are just going to become more expected like regardless of like like the person behind it like I think when you have things like Figma make and like cloud code it's just it's much more easier now to like visualize the solution like you're proposing. So, I think that's where we're going to see a huge change of that of the sort of expectation that like when you're building something or like when you're proposing something, there should be like some sort of like um I guess visual or like work like working visual to like go off of. And I see this in my thesis class like a lot when I look back on like honestly when I started my masters to now. um in a lot of my like UX projects like in my first semester it was like it was fine if you kind of had like the sort of um basic wireframe like going on and that was your final solution and but now especially during thesis we're seeing like I see more and more where it's like even though I know with that person like they're more experienced on like I guess the research end their designs look just as polished you know and I think that's something that starts to change >> and I mean a point she um she said going back to the first statement about don't trust the process I think the lines between code and design and research will continue to blur and what are your thoughts to that? >> Yeah, like I think I I I I like what you said of like it just starts to like be like parallel. I think that's where I feel like I was mentioning this before. I think the shape of like it might not be like these two like the the shape that we're seeing right now. Maybe it's just we have to like kind of change like adapt the process and like visualize it in a different way where it's just kind of happening in tandem because it's so easy to like kind of get these results and quickly and rapidly like iterate. >> I think there's I think there's still value in specialization. I still think so right? >> But I do think that at a base level you'll see more and more practitioners that know a little bit about everything, right? I think that's going to happen. um similar to like med school where like every doctor has to go through four years of medical school with like clinicals and rotations and then residency and they're specialized in something but every MD knows the basics. I think in UX what happened for a long time was that we went down this path where we were like hyper specialized at least in the big companies right like in the in the metas and the apples and like the Microsofts that's what it was like Walmarts right you had like hardcore content strategists hardcore researchers designers right and um I think this world has kind of challenged that which is like you're going to see some of those blends where people are a little more and what in my experience Those unicorns usually would be like in smaller companies or maybe like startups, right? But I think you're going to start seeing it more and more at even bigger companies now. >> Yeah, definitely. >> Let me see. So, we kind of ran through this Brett. Um, but we do have last topic is closing thoughts about AI and UX. This one isn't like a question to the group, but more of just like Ranju and I both have some thoughts. I actually um I stole this image from um my boss who came from Amazon. She um she she had a talk with me about AI recently. So um want to give her credit for showing me this. But if you think about like the change curve, right, which represents like the different like stages like psychologically that people go through, right? When they have when there's any change, not just about AI but anything, right? So you go through like there's like shock and denial, frustration depression experimentation, which as you're experimenting, you might go back and fail and have more frustration. It keeps looping back like that and then at some point you push forward and have decision integration, right? Um I think this is this is a good way to think about like anytime you have like big change and this is what we're dealing with right now. I think myself where I'm at with with AI and UX is that I went through the shock and denial that was like when I showed you the images about the VR and meta. It was the same thing. I saw it. I was like, h it looks cool, right? I don't know if we're going to really be using it. I don't know how long it's going to really pop off, how long it's going to take to pop off. And then it started to happen overnight where it seemed like you went to sleep, you came, you woke up, and then AI was everywhere, right? It was like an AI overload. And that would be like the frustration cuz it like turns your world upside down. And then you go you have like a low where you're like ah I guess we're this is it. We have to deal with it. I think I'm in the experimentation phase and that's where a lot of people are now where there's so many tools out there. We're still trying to figure it out. Um even as we were walking through the double diamond and I was telling you my perspective that's just me experimenting and thinking about like what I think is going to happen. I'm sure there's people out there that might disagree with me. There might they might agree with me. Um, but I think that's where I am. I I am that orange part. Where where would you say you are, RJ? Like in this >> um I think I would say that I'm a little bit like on the way to like decision I guess because of how prominent it start. I mean, you know, like kudos to like academia, I guess, of like they really pivoted really quickly when like this like sort of explosion of AI happened. And I saw it overnight like changing kind of like how classes um worked and like wanting to prepare us for the professional world. A lot of people like or a lot of teachers really focused on like bringing that into the classrooms. Um, so I feel like I've been around it enough where I'm like I have like and I'm going to touch on it of like my thoughts on it of like I I feel a little more equipped and I have like this agency of like deciding what when it's integrated into my life and when it's not, you know. >> Cool. What are your closing thoughts about AI and UX? Ranju. Yeah. So, I think I think we're at this really interesting point in time where we're at the precipice of like we have the ability to kind of shape how this all of this kind of shakes out and where we like where our role is in in this and also where the AI is, you know, like even if it's exploding and it feels like like bigger than us, I think at the end of the day, we still we still have that agency and um ability to kind of to kind of shape how this goes. And I think that's where ethics and like frameworks come in. And um I learned about this like framework from class called um created by anthropic like the 4D fluency framework. And it touches and a part of that framework is like assigning a role to AI. And so in this case and we kind of talked about this um during this presentation, but like maybe AI in your situation is like a sort of research intern. you know, it's like equipped to like kind of help out in that role of like simple tasks where it's like organize my transcripts for me or something like that. Or maybe it's like your collaborator and that you feel you have more trust in it. And so that's where I think like the you know we still have this sort of choice and decision and like how the boundaries that we're creating and um and where we like kind of explore and see how our work can be better like as designers and as UXers, you know. So yeah. >> Yeah, absolutely. I think I I agree with that. I think we are definitely building the bridge as we go and there are times where it could seem scary or frust frustrating but just to remind like remind any practitioner spent a lot of time in the field that it doesn't necessarily have to be doom and gloom. It's easy to go on LinkedIn or read like a medium article or watch YouTube video and it seems like the sky is falling, right? But it doesn't have to be like that. It's all it's all perspective, right? Um, I've been surprised of what as a researcher like what AI can do. That's really cool. Like things like being able to like read through all of my notes and code and find themes faster than I would have to go do manually. I still think there's there is there's value in me doing that, right? Especially if I have to like really understand the data, right? But there are things like even like playing with certain design tools and how it makes design even if the design the the output design breaks like a ton of usability heristics right the fact that it can do that is kind of cool as a starting point right um I think there's going to be a human in the loop required in this entire new world but yeah I mean I think I think we definitely have a we have a role to play and we're going to be as as I said like building the bridge as we Go. So >> yeah. >> So that's pretty much what we have group. Um I know it's like a really quick chat, but we kind of put this together. I tried to keep it really current, right? Um but I'll take any questions anybody has. Feel free to bra. How do you want to do this? Do you want them to come off of >> They'll they'll put it in the chat and we can just ask it there. Um, I have some softball questions for you and and more so for you to expand and want to hear your thoughts, but um, you know, you talked a bit about the where we're at with with AI, right, and acceptance or denial and you know, everyone's at that I guess different stages of grief, right? Uh, you can equate it to. Um, you know, what are your thoughts? Um, is it the integration stage or where do you see it across the line where you said like it's the wild west right now, right? where you don't know what tools you're using, you're exploring, you're trying to figure that out, you don't know how to use it. Um, and we're all kind of out there and trying to figure that out. At what stage do you think we start having a processy, right? Because >> yeah, I think I think what's happening right now and again this is just my own subjective opinion. just from me talking to my friends at other companies within the Fortune50 is that like I think what's happened is that there's market pressure right now for every company to use it because there's always that fair like let's say that let's say Brett you me and Ranju are in the um the field of making I don't know I have a protein shake here like protein shakes right and let's say the manufacturing for this protein shake is right now it's all manual right now whatever and we start we start hearing about AI and we're like wait man if we don't get if you start in integrating AI in our process what if the other company that makes protein shakes down the street is using it and they beat us to the punch so I think there's like a little bit of fear and anxiety around the market which is causing market pressure for everybody to do it across every industry it's not just like tech it's like every industry is starting to use it law medicine right you're starting to see that and when you have that type of pressure, like it tends to seem like the wild west because there's a lack of framework or instructions, right? It's almost like we want you to do this, but we're not going to really tell you how. This is where I think frameworks come in really handy and like to push back against like, oh, process is dead. I would disagree. I think process was never meant to be like, oh, you have to do it this way. It's something to fall back on and say like, "Hey, at least when it seems like the sky is falling, I have like almost like a script on what I should be doing." Like like you're right. Doesn't mean I can't change it, right? It's not malleable, but I do think that um yeah, I'm not sure if I answered your question, right? I think I like I kind of went off on a tangent on other things but >> no valuable, right? Knowing you, process is key, right? Uh the other kind of point which is going on Joselyn's question or statement here um you know what we're seeing and and some of the stuff we're seeing with more analysts or junior researchers is the over reliance on AI right and and Joselyn kind of sums it up as my 16-year-old is replying back to his dad saying I don't care what AI says what do you actually say um I think there's a dangerous point where those that are entering UX or research in general uh are over reliant on AI and don't know the right answer. Um, so whatever AI tells them is the right answer, right? >> Yeah. Absolutely. >> Very dangerous. I mean, like I going back to my friend that's an attorney. This is a guy that has a jurist doctor, right? Like a law a law degree, right? And imagine if he didn't know law and he wasn't doing this for 20 years and he was just say, "Oh, because AI told me, this robot told me that this is the legal interpretation. I'm just going to go and go with that." Right? It's very scary. I've been talking to my friends that are in med school. Same thing. They'll like look at things and they'll know it's wrong. They'll say like, "Yeah, this is wrong." Like, it's just not that's not accurate. Right. So, there are hallucinations. There's um I think New York City, cuz I'm that's where I'm based obviously. New York City there is it's on the it's on one of the um the voting measures that's coming up. the idea of like it's just an idea. The idea of like of controlling like putting parameters in place that like certain things um cannot be like searched via AI meaning like things with laws, things with medicine, things that have like really big repercussions. Someone have to fact me check fact check me on that. It was something I I saw online like in passing, right? But I do think What are your thoughts on that, Brett? like the idea that like certain things like AI like hey AI is fast it's great but there's certain things that maybe it's not there's too too much risk right to go in there what are your thoughts on that I >> I think it's across the board yes right I I think we need guard rails and that's kind of where I was probing on that first question is when do you use AI and when you don't right like I don't need AI if I'm going to quickly respond to your text dad and say hey man can't wait to see you later awesome right it takes me longer to put that into an AI platform, generate that, copy it, then put it to you, then it is see you soon, right? Then there's the other component of the stuff that's dangerous that I don't even know about, right? I couldn't tell you how to do brain surgery and I shouldn't use that for AI. Um, so I think it's going on kind of the conversations you've had is there's going to be a fall back to actual skill sets, right? And I think that's what you know having you make this uh start this conversation and lead this talk. You are someone that has such a rich and thorough skill set within the research you know space right whenever I have questions whenever a lot of people that you and I have relationships with have questions you're one of the first calls that they go to of hey talk to me about this technical research methodology. Um, so I think it's a fall back on the the old way of doing things, uh, which is actual technical skill sets, but then the ability to expand that with scale and speed and efficiency with a platform after you learn it. And I think that's the part where people try to take that shortcut. Um, whether that be in medicine, law, or whatever, where they need to learn and get those skills before they can utilize this awesome toolkit that is AI. Right? If we go back to like the the the um that image I had about the VR worlds and stuff. I'm going to go back to it real quick. Um I remember when this came out the idea here was that >> D you have uh both screens up like um or not both. Yeah. Moving that. >> Sorry. Can you can you see it? >> Yeah. >> All right. Yeah. If you look at the image on the right, the interesting thing about the image on the right was the idea here was that you're wearing like a VR headset to look at a so you're looking your eyes are looking at a screen and then in the world you're looking at another screen. It's like so like meta. I think that's where they got the name meta right from. But it just is so weird to me at least that like I guess I'm trying to say like to your point like I think that there's going to be a there's going to be a normalization right of the excitement of any anything you see like something that's new right there's always excitement and people react to it markets react right and I think at some point once people realize and the industries realize what what are what what are things really good for like what is this tool good for and what is not good for. And there's more research out there that shows that we might not want to use it for this, right? You're going to see start to see a normalization of like the usages of it. You see with even things like Whimo and stuff, right? Like it came out, people were scared about it. They were It was um I saw videos of like crazy accidents of like these cars like flipping out and crashing into to pedestrians and stuff and then there was like laws passed and like you know it's starting to normalize like the use of it, right? Um it's going to take time for it to really for us to really understand like how to use it properly. Like even when we walked through the double diamond and we were talking about like when like how it's going to change and how it's like we're such in the early phases right now. It would be so like irresponsible for me to be like, "Oh, this is what's going to happen. This is what's going to happen." Nobody really knows. I truly think that we're going to see again in that that first diamond, you're going to see more uses of like concepts because like we can create a concept with like a couple clicks in a prompt, right? I just be like, "What do you think about this design?" Right? And you're going to see more of that generative research with a concept. And then the evive research again is going to be more with like actual like real data like a real prototype that's already been like pro like actually developed also. So but I think the double diamond or even just any type of framework is still pretty sound. >> Yeah. >> Let me look at the chat real quick. I know there's a couple comments. So Josyn said explaining about the AI. Any thoughts on how different generations are going to adapt? So, I have both of my kids are are young. I have a four-year-old and a 2-year-old, two girls. Um, I think that the the scary thing about AI is that I think there's pros and cons. I mean, it's cool that like these these this newer generation's growing up with like so much information. It's like literally we thought we were growing up in the information age. That's like they're in a completely different era now, right? It like redefineses it. I think that the hallucination part is what scares me, right? That's what scares me. Like RJ, what are your thoughts on hallucinations? Like you're in NYU finishing up right now. Like what's your what have you seen about like AI hallucinations? >> Yeah. Um, I think I've definitely seen it like I' I've seen some sort of hallucinations when I was like working on my thesis project or like when um as are like we're assigned into like peer like groups to also like help with each other's thesises and everything. And there are some things where it's like if someone didn't catch that, I feel like the project would have gone in a different like in the wrong direction. And I think that's where my take on this is that a person's ability a person and like a professional's ability to discern it becomes like mandatory you know it's no longer like a skill where it's like it's a positive and it's a plus I think it's like you have to have it you know and in order to sort of make things that are like accessible and safe to people like if we're talking about from a UX perspective and then as a user like being able to navigate something And like every time you're asking AI and you're like asking it to make something or create something like the other part should be like okay well what do you think like what's your like feedback and what like you know what's jumping out to you and I think that's the part that we really need to emphasize for this generation that's like it they're you know AI is already existing as they're coming into it. feel like that's where I would love to see more of a push in that like curriculum and stuff like that of really understanding discernment and you know navigating that part of it and also um this kind of topic it reminded me of this book I don't know if um anyone's read it here but it's called culpability and it talks about it tackles this exact thing of like growing up with AI and um having like situations of like hallucinations and how far this thing can go and how important ethics play a role in this. So, culpability, it's a good book. Definitely check it out. >> I haven't read that. Has anybody read that that book in the chat? >> I don't see anybody. There was another comment um that talked about like I I'm just going to read it. The frameworks and methodologies we know are important to hold because it helps us establish a standard and understand where AI can fit in the process. When we blow up everything, abandon frameworks and just say use AI to be fast, that's where it can become dangerous. I think so. I think like again I think there's like this misconception that frameworks are meant to be like scripture like you can't break from it. I've I haven't used the double diamond like to scripture since like my early years in my career because anybody that's been in this field for a long time knows and run you I mean keep me honest right as someone that's newer in the field coming straight out of university right like you came out with this mental model of like oh this is the double diamond this is like this is a framework what was your reaction jumping into the field in your first year working as a UX researcher like what are some of the things that you saw Yeah, I think that was an interesting sort of journey for sure of I mean I I think it's it's like school can prepare you to like adapt is how I'd put it where it's like I don't know it was just interesting learning how there's this process, yes, but it's not as cookie cutter as what we're learning in school. And I also think um those processes sometimes account for like you know we'd have to do it in like a semester or we'd have to do it like we'd have the set amount of time and it's like done and and there's like no stakes in that situation as well since it's like a school project so it's really up to you. I think in a professional standpoint, there's a lot more at stake and there's a lot more people involved and there's um more like time in a sense. Like yes, there's time pressure, but it's like also the goal here is to make something like that's as effective as possible, right? >> You still you have to evolve though, right? That's the that's the point, right? I I don't know. I do think that like I tend the more we talk about this, right? I I really truly I feel like I personally like align more with NNG about the idea of like compression rather than abandonment, right? That's just me, right? Um I do think we have to move faster. I think we will move faster, right, as an industry. But I think abandonment of like frameworks that are like peer-reviewed and and time- tested over like millions of of workers and use cases out there, right, for like decades is dangerous in my opinion. So, we have 15 more minutes. Any other questions or anything? If not, we can give everyone some time back. Um, Dev, always a pleasure. Um, always value your your expertise and um, well, never mind. There is another question. Have you ever been able to or have you been asked to create AI features? >> To create AI features? Yeah. I mean I think like every I think I think at this point a lot of myself or like my network is being asked to create AI things. I think it's like AI is such like a broad term that um within AI there's so many like subgroups of like features. There's obviously like like natural language processing, right? There's like chat bots which that's essentially what it is, right? I think chat bots were like the first one that I saw that really became popular. he started seeing them um and now they're super popular. I think that's why Ranju shared that graph and showed you like it's probably the most used that you see. Um chat bots are super super like popular. I think like then you had like um like like generative AI, right? And then aentic AI which is where we're at right now. But um yeah, I think those are like the three big use cases I'm seeing where like you're starting to see these features and they're um I'm reading more about things like neuronet networks and like basian logic and things like that which are much more advanced but like these are things that actually have been around since like the 1940s. They they they're not like new really. They're things that scientists have been working on for a long time. and they just recently like everything connected in like a massive like nuclear bomb and they're like whoa it's actually working now but if you like actually this original talk I actually was going to spend the first like like 20 minutes or so talking about the history of AI before we jumped into the topics and I actually removed it right but during that during that like I was doing some quick research on just to kind of prime myself on the history history of AI and stuff and what I was finding what was interesting was that like the first like thing that I found historical context of AI was actually that book Guliver's travels from like the 1700s. It was the first time that someone had mentioned like a machine. He called it like the engine that would help humans like write books. But since the 1940s, we've been actually creating like features like AI features. It's just now up until now it's just started to recently like really come together. So >> yeah, Dev there I think it was one of my last projects at Walmart. I think you were gone by then, but um we were creating the chatbot uh like a smart assistant, smart shopping assistant. We've done that here at Accelerant for a couple different Fortune 500 companies since uh but creating that framework, helping understand what are the limitations, what can it do, what can it do. Super interesting work as well. >> Yeah, the history I'm seeing like people love the history and context. I I'll send an article to you, Brett, that you can share with the group that it's actually an IBM article that lists out the entire like chronological like history of AI. It's really interesting. There's like there was like in the 1950s there was some of the most AI research actually that is applicable to UX was actually done in 1940s to60s and that was surprised to me because I'm not like an AI expert, right? My expertise in my doctorate is about learnability and cognition, right? Not necessarily AI. Um, but I I was fun just it was fun for me just learning and reading about it. So, I'll send the article so you can look at it. It's really cool. >> Yeah, we'll share we'll share that out to everyone. >> Hopefully, this was hopefully this was um interesting and useful. You know, I get I get asked to do these talks every so often and I always I never want to like just lecture to people, right? I think I like to make it as interactive as possible. Um, but hopefully it was interesting and at least it sparked conversation. Hopefully it was applicable to some of the things that's happening right now. So if anybody has any other questions, feel free to add um Renju and I on LinkedIn and just reach out to us if you have any other questions, you just want to chat about like AI or anything else about about UX. So >> awesome. Thanks Dev. Uh thanks Mu for the time. Um appreciate you guys coming in and uh hosting the conversation. So uh like I said um all these will be found on our website. So this talk among the other three or four that we had today and then we have our next one coming up in April. But >> what were the other talks about today? >> Uh let me look real quick. Um we had about four today. So we started off with AI. Uh Bill, our founder and CEO, spoke about um the quality of participants within the AI spectrum or the AI landscape, right? Maintaining quality within AI or the age of AI. Uh the next was consumer insights. Uh so someone from Roku, Alex Dress, uh spoke about consumer insights, uh where the lines kind of blur and then uh innovating for inclusivity uh and impact from uh Viking cruises. >> Nice. I dropped the IBM link, by the way, in the chat. I found it. That's the That's the link. It's It's one of the better like um articles on the history of AI without being like super academic. So, check it out. Cool. Um appreciate your time, Dev. >> Yeah. Thanks for having awesome. Thanks everyone. Take care. >> Thank you.