Research the needs, not the roadmap with Preeti Talwai

Listen

Watch on YouTube · Apple Podcasts · Spotify

Summary

Preeti Talwai, UX research leader and product strategy advisor, on researching products that do not yet exist — how an architecture background trained her to think in systems, thresholds, and liminal spaces, translating that into technology research that zooms out to the whole ecosystem rather than one feature; why futures research anchors on enduring human needs and experiential simulation rather than the technology itself (the home-robot study that separated experience reactions from technology reactions without ever building a prototype); the sushi conveyor belt approach to delivering insights (bite-sized, sequential deliverables that shift the perception of speed without sacrificing depth — low-fidelity video reels to high-confidence topline to full insights deck); measuring research impact through four organizational shifts (definition, prioritization, direction, inspiration) rather than product launches; redirecting tech-first engineer enthusiasm toward human needs by showing the gap between technical difficulty and experiential value; defending qualitative research from speed-at-all-costs AI tools (fewer participants by design, the human process is the point, fix the delivery cadence not the method); AI as a material not a tool (unique strengths and limitations like concrete or cotton — build with what it is uniquely good at, do not fight its nature); the learning-by-doing distinction (automate autopilot tasks, preserve the wrote tasks where processing teaches you something — note-taking as the canonical example); closing principle: new technologies are always emerging but human needs are enduring, and bridging new capabilities to fundamental unaddressed needs is the researcher's superpower.

Guest

Preeti Talwai — UX Research Leader and Product Strategy Advisor · Marketing / startup

From this episode — top claims

• AI should be thought of as a material, not a tool — like concrete or cotton, each material has unique strengths and limitations. You would not build a building out of a porous material and be sad when rain leaks through. AI's unique superpower is in identifying patterns and predicting things humans cannot see because of access to a huge corpus of data; its underutilized potential is in helping people with unexpected change and pattern recognition at a scale humans cannot match, not in generating more or making things faster. — Ep. 75, Preeti Talwai
• New technologies are always emerging, but human needs are enduring. The most fundamental needs are basic yet complex, tied up in complex systems, and not fully addressed even now. The researcher's superpower is deeply understanding both the technology and the human and building the bridge between new capabilities and fundamental needs in actionable ways — there is no need to chase the next flashy thing when really basic, 'boring' things in people's lives remain broken. — Ep. 75, Preeti Talwai
• Even when a working prototype exists, researchers should separate studying how people react to the resulting experience from how they react to the technology enabling it. The experience reaction tells you whether you are solving a real need; the technology reaction tells you how people feel about the enabling mechanism. You could love the idea of your clothes put away and be creeped out by the idea of a robot in your house, or vice versa — conflating these two reactions loses signal about where to focus. — Ep. 75, Preeti Talwai

Full transcript

0:00 The sushi conveyor belt was a way to visualize balancing like you said the speed of insights and then the durability of insights. In early stage research that's really hard because you want to move fast, but you also want to make sure that if the team happens to pivot or change direction that your insights are still relevant and they're not suddenly going to go stale. Hello fellow insight seekers. I'm your host Molly and welcome to the Curiosity Current. We're so glad to have you here. And I'm your host Stephanie. We're here to dive into the fastm moving waters of market research where curiosity isn't just encouraged, it's essential. Each episode, we'll explore what's shaping the world of consumer behavior.

0:00 From fresh trends in new tech to the stories behind the data. From bold innovations to the human quirks that move markets, we'll explore how curiosity fuels smarter research and sharper insights. So whether you're deep into the data or just here for the fun of discovery, grab your life vest and join us as we ride the Curiosity Current. Today on the Curiosity Current, we are joined by Priwy, a product researcher and strategy leader whose career has focused on helping organizations understand emerging technologies long before they become everyday products. Pret's path into research is anything but conventional. She studied architecture before moving into technology research and has spent more than a decade helping teams explore AI and other emerging capabilities long before they became fully formed products.

0:00 Throughout that work, PT has developed a unique philosophy of research. The further into the future you go, the less your work is actually about the technology and the more it's about understanding timeless human needs.

0:00 So today, we're exploring what it really means to research products that don't yet exist. how researchers can influence decisions even when those products never launch and why human- centered thinking may be the most important capability in the AI era. Pretty, welcome to the show.

0:00 Thank you. It's so great to be here.

0:00 We're very happy to have you. As Molly mentioned, your path into research started in architecture, not marketing or human computer interaction or or even psychology. Looking back, can you connect the study of physical spaces to the way that you understand people today?

0:00 So, you're right. I studied architecture and my background in the built environment and in architectural theory and in research has really deeply shaped how I think not just about people but also about technology. And I think it's done that in two main ways. I think first is that architecture is really about systems in larger context. Um, and one of the things that I've brought to UX research is understanding people's needs and their use of technology holistically. Really zooming out to think about the whole ecosystem, not just one feature or one task or one product. In architecture, we talk about things like thresholds and movement and liinal spaces. And it's a very like alive and active sort of research and translating that into tech research.

0:00 Um, I think I've always tried to make sure the teams I'm on are thinking about people's broader goals and how a technology fits into the complexity and the messiness of someone's real life, the seams and the transitions between products and features, which I think is not often done in industry just because of the way that organizational structures are often siloed. And so, yeah, so one way is really zooming out and understanding like messy holistic real lives. And then I think the second way that it's shaped me is by training me to think in terms of language and frameworks and taxonomies which make up a lot of what the contributions of scholarship in the humanities are. Um like if you read theory in any of the humanities um and architectural theory in particular is very multi-disiplinary.

0:00 So it takes from anthropology and sociology and philosophy um and making sense of again a lot of that complexity of life. And so um a lot of my own contributions um over the nine years that I was at Google were really going into these deeply nent and ambiguous spaces and helping them feel tangible through a framework or through a vocabulary and that has certainly been shaped by the the academic training that I had.

0:00 No, that makes a ton of sense and especially I think when you think that of course architecture is, you know, a human endeavor, right? And I love that line about seams and transitions. That really got my brain turning. It's really fascinating.

0:00 Well, let's switch gears a little bit to talk about your anticipatory research, which is always so fascinating to me. Like, what's the research crystal ball we're all looking into? And so, you've described that your work is researching products that don't really exist or don't exist yet. What makes this type of research so fundamentally different from traditional UX or product research?

0:00 Yeah, this is a good question. So, I think there are two categories of differences I think about. The first is difference is in research approaches like how you actually do the research. Um when you do work on futures and bets um and there's also differences in how you navigate the organization in doing that sort of reach sort of at a high level in traditional product research you're typically working on something that exists that has users you can probably ask directly about um and that probably has some sort of like roadmap or plan that you can align your research to. With futures research, I think a big difference is that your users may not have any mental model for the technology or the product that you are thinking of um designing.

0:00 They may have folk theories that might be informed by sci-fi or increasingly informed by social media, right? But they're not they don't really have a good grasp on it and none of us have a really good grasp on it, right? It's it's new. Um so change aversion is a really big thing as well as anchoring in the present. it's really hard for all of us to sort of think beyond our present frame of reference. Um so that takes sort of creativity to navigate around. And then of course when there isn't a product or even sometimes a fully functioning technology yet, you can't get direct reactions to it. Um you end up focusing a lot more on the needs of people um and the experience and your methods need to be resilient to a lot of organizational change because these early stage products move really fast and they pivot fast.

0:00 And so you have to be not just aligning to the road map, but you have to be part of charting what that road map even is and driving with your insights and be really comfortable like being in the driver's seat. And then I think in terms of navigating an organization, it's also a lot less straightforward. It's often hard to get buyin for something that may not show immediate impact. Right? You may hear I have heard many times in my career the notorious faster horses quote. Is that are you guys familiar with that?

0:00 Oh yeah. So, faster horses is a quote that I think is actually misattributed to Henry Ford, but it's the idea that not him.

0:00 Okay.

0:00 I have heard that it's there's no actual proof that it's him. So, I'm not I'm not sure, but but it's often attributed to him. And the idea is that when you're um you're building a car and you ask people what they want, they would have told you they want faster horses. I have there's many pet peeves with that quote because underlying that if a researcher even did that ever did that which they wouldn't because we don't ask people what they want but if they were to do that the underlying sentiment there is that people want speed right um I think that that's that's a good takeaway and that's what cars provide um but it's often used I have often had it thrown at me as a way to say hey we early stage research isn't really viable or is not really trustworthy because people don't know what they want so you know that is something that a lot of education with stakeholders ers um and creativity to to navigate that is something that you need and I also think you know it's often hard to get a seat at the table when people feel like it's not time for research yet which is also that skepticism around like until we have something we don't really know if there even is something to research what would you even do um right so I think it's really to me about building trust with your stakeholders as a research leader I feel like doing bets research is a very kind of human and collaborative endeavor you really need to have deep trust with your stakeholders, your leadership, so that you can do what you want to do and you can like go out and be a little bit of a research cowboy.

0:00 Um, because that's what this sort of research is. And I I think that you also need to be more conversant in the technical side of things with emerging technology than traditional UX research. Um, so you can bring people along the journey and they you can both speak the same language. Um, you need to be a storyteller to help people understand the human impacts of these technologies. you need to be an advocate for responsible technology. Um, so I say all of this, I think the theme here is that I think of emerging technology or futures research um, as being a lot broader than you might think of traditional UX research being. You end up wearing lots of different hats like a strategist, an hertologist, a researcher, a critic, and all of that sort of wrapped into one.

0:00 Priy, you started talking about this a little bit already, but I'm curious if you could dive in a little more. when there isn't a product or an interface or even a clear use case yet. What do you anchor the research around? I mean you mentioned needs I think is is one area but what other anchors do you use in research in these situations where you know what are we evaluating? Probably nothing.

0:00 Yeah. Yeah. So I think there's a couple things like you said first human needs. Um, and I think going a little bit deeper into that, I think needs tend to be really durable and unchanging over time. And you want to look at also I think the gaps where existing technology is failing these core needs. Um, so as one example, a core need that I found in my research like eight years ago is the need for people to be able to recover from curve balls or unexpected changes in their life, right? And yet, if you think about even with all the advances we have with AI, technology is still pretty bad at helping people be resilient in the face of change or exceptions to the routines. Like routines are easy, but exceptions are hard, right?

0:00 And so, you know, we live in an age where I can like vibe code an app in an hour, but is still super manual to let's say you miss a flight to have to find a new flight and then like reschedule your entire day, right? like that is still a really stressful process even though that task is sort of more contained than like by coding an app right and so you know same thing if you're like trying to recover from a financial emergency or like modify my diet when I'm traveling those are still not really addressed needs and so there's often no matter what domain you're looking at I think probably these deep pockets of unadressed needs and um helping point a team towards the sorts of problems they should be solving I I think is is one thing to definitely anchor on.

0:00 The other thing that I often anchor on is the experience. So like what will the user's life be like as a result of using this technology and can you simulate that? Um so a good example that I like is if you're trying to do research on like consumer home robots, right? And you want to understand what might people want to use robots in their house for. you can do a lot of that research without ever having a working robot prototype by just simulating the end product of what a robot would do. Um like for example, let's say they're loading the dishwasher or the robots floating folding your clothes or whatever the task is, right? So what does it feel like to come home to the breakfast dishes that are totally done, but maybe a cup is chipped or broken, right?

0:00 Like maybe that's what it would be like if a robot did it. What does it feel like to come home to your laundry folded and put away in your closet, right? Even if in a study it's just a human who's doing that who's simulating the robot, you can learn a lot of things about these use cases. You can learn about people's expectations of performance and quality. You can learn about their need for things like control and feedback, right? You might learn, hey, actually it's super important for people to give instructions about where clothes go in the closet or load the dishwasher a really specific way. And you'll learn about, you know, what delights them and what concerns them. And that can inform how you do build that technology even if zero robot technology exists at that point.

0:00 Right? And I would even go further to say that even when you do have a working prototype, I think it's really wise to separate out studying how people react to the experience and how people react to the technology. So you don't actually want to conflate those two things. So the way people react to the resulting experience tells you whether you're solving a real need. like do people even want their laundry folded and put away? And the way people react to the technology tells you how they feel about like that enabling mechanism. Like how do they feel about a robot doing that thing, right? And those two aren't the same. You could love the idea of your clothes put away and be really creeped out by the idea of a robot in your house or vice versa, right?

0:00 And so I think actually not immediately having the technology be part of a study or part of the research is actually a really crucial part of doing emerging technology research. No, that makes a ton of sense. And then it I can just imagine that, you know, in these scenarios where I love the experience, but the enabling mechanism is kind of creeping me out that it helps you focus on what you have to overcome to make this a viable solution.

0:00 Exactly. And I think a lot of times for all of us when we encounter something new, it's not always obvious to us why we're reacting a certain way, right? It's not obvious to us if I were to see a robot putting away my dishes or my clothes, I might say, "Oh, I don't like that." But I may not even know why I don't like that like what part of it, right? I can't even tease that out myself. It sort of has to be teased apart for me like you know in the structure of a study or the way that you do that research. Um but yeah, exactly. It helps you know where you need to focus and what you need to improve without mushing it all together in one thing.

0:00 Yeah, makes a lot of sense. Yeah, it's fascinating the context of the solution as well because especially as like a mom, a busy working mom of a young kid, I would love my laundry put away but a giant robot in my house. I Yeah, I don't know. I fall into that camp for sure.

0:00 Yeah, exactly. It's so interesting, right? Like you definitely don't need a we don't even need to have anything working or a prototype to engage with. But it's interesting like if this were the feedback from a study that you did, you might be like, "Oh, well people, you know, maybe it's not the the use case that we need to focus as much on. We really need to think about the form factor of this robot, right? Like it evokes something that's big, right? So what if Ashley was small? You know, like you need a little cute robot, Molly. Yeah.

0:00 Yeah. Apparently I also anthropomorphize my robots. I have a little robot vacuum and he's got a name and he's got a little personality and spite my son loves watching him go by and we call it a him. I don't know, maybe that's me.

0:00 That's fascinating. Yeah, it's like so much so much to learn without needing the technology in front of you. Well, let's talk about teams that want to move fast on products. Um, especially in product evolution and wanting to capitalize on trends, teams often feel like they have to choose between getting the stuff done and then also having research that lasts and that is actually going to be meaningful long term. And you've argued that that's actually a false trade-off. So I wanted to talk a little bit about something that you've mentioned before which is interesting uh about your sushi conveyor belt approach to delivering insights.

0:00 Yeah, sure. So the sushi conveyor belt was a way to visualize balancing like you said the speed of insights and then the durability of insights in early stage research. That's really hard because you want to move fast but you also want to make sure that if a team happens to pivot or change direction that your insights are still relevant and they're not suddenly going to go stale. Um, so if you think about your research insights as a meal that you're serving, typically when we have like evergreen durable insights, we think of those as coming from large foundational pieces of work that are sort of naturally slower to produce. And for those sorts of pieces of work, researchers, I think, tend to deliver those insights like an all you can eat buffet, right?

0:00 There's like these lengthy dense reports. They come out all at once like kind of go underwater for four months and you show back up. I think some of that is changing with with the use of AI now. But but I think traditionally that's how people do that research and I think also a little bit why stakeholders are skeptical or a little worried about those kinds of research because they're like it's it's not going to be on our timeline, right? But instead I like to think about delivering insights on a sushi conveyor belt. I use that analogy because each piece of sushi on a conveyor belt is composed. It's complete and it's satisfying, right? So, it's not like a halfbaked idea, but it is also bite-sized and it's digestible and it's produced at kind of a reasonable and consistent and reliable pace.

0:00 And so, I think for research and practice that looks like a menu of different deliverables that are shared sequentially, not all at once. And you might start with something really low fidelity and low confidence like a video reel or a couple quotes. Um, or maybe using an AI tool to generate some highlights, right? And again, low confidence, low fidelity. And then you might move to a topline report, which is high confidence, but still low fidelity. It just has a couple of the main bullets, lacks the details. And then you kind of move to the traditional insights deck, which is that high confidence and high fidelity. And finally, you can have things like workshops or playbooks or things to like really activate the insights and make them stick.

0:00 Um, and that can come even later. So I think you know even if the project itself actually takes a while has many phases is more ambitious I think the perception of speed is actually the delivery of insights it's not really like how long is your research project I think it's more how frequently do you deliver something that's useful and insightful and so I think that sort of sushi conveyor belt or like multi-course meal however you think of it can help that the team move faster um even when you are doing um a more durable foundational sort research study.

0:00 That's a fascinating way to think about instead of this big sort of shift in what the work is, it's more of the workflow and the delivery of the products at a certain cadence, but it's not changing fundamentally the work.

0:00 Exactly. Exactly. And I think obviously there's times where the research itself needs to be more scrappy or like we're not doing these big projects, but again, I think in emerging tech world, it really is about getting to those core needs. It really is about finding something that's quite deep and driving the product at a really foundational level, not a surface or a tactical level. And so I think again, you know, I I know these days in sort of our AI era, it's all about like moving fast. And I think often we think that means well, let's change our methods or like let's use AI to make everything go faster. But actually, I think it's like shifting the perception of speed. And you know, there's there's a lot of things we can do to shift the perception of speed without actually being faster.

0:00 in the in the research itself.

0:00 That's such a good point. Switching gears again, Piti, some of the products that you've informed have taken years to emerge and I'm sure that there are probably some that have never launched. In that kind of context, how do you measure the impact of research when success isn't necessarily tied to like a shipped product? Yeah. So, I think about research impact as a tangible and ideally a quantifiable shift in an organization that a research directly brought about. So as in you can say you know before this research prior to this research the state of things was ABC and then after we did the research the state of things was XYZ. I think the good news is that you don't have to have launched a product to have orchestrated a shift.

0:00 And so to get more sort of tactical about it I think there's four types of shifts I like to think about when I do especially with emerging tech research. The first sort of shift is definition. Um so that an example of that is saying before this research there was no defined user or use case and after the research we had defined X user groups and Y use cases uh and you know we're really defining what the team is going after. So I think that's like one sort of shift or impact. Another one I think is prioritization. So it's common that when you do early stage research there's like 40 different concepts or something like that, right? And there's not really defined work streams. It's very messy. And so if you're able to say, hey, we came in and did this research that actually resulted in a framework that helped us pick like the top three ideas to pursue that resulted in these three different work streams with 10 people each working on them, right?

0:00 And that's a very quantifiable shift that you can talk about um about how you went from many ideas to few ideas and and and putting numbers to that. Um the third one is direction. So I think that is sort of saying before this research the team was going in one direction and then you do this research and now the team is moving in this other user focus strategic direction. So helping a team pivot um or helping a team accelerate in the direction that they're already going in. And then finally inspiration. that someone being able to say as a result of doing this research we identified let's say a particular need um and that might inspire a certain feature that didn't exist before an idea that didn't exist before and got put on a road map and now we have a team of 20 cross functional people working on it or whatever that is and so as you can see like in each of those cases definition priorization direction and inspiration um even if the projects you worked on like never saw the light of day you still had tangible and quantifiable impact with your research and I think that The key for me has been to focus on the organization more so than the product as a recipient of that impact.

0:00 So as long as you can identify before and after there um you you can show impact regardless of what happens to the product and you don't have to be so precious about that. I think I think that's an important framework to attribute because there's ongoing conversations all the time about researchers having to justify their own work because it when research is done on a product or a business decision, it can typically be that nothing actually happens, right? Like the product goes well, the launch happens, everything's great, but there's no ROI on the research, but it's just because it prevented a costly mistake. So how you actually attribute uh an in return on investment for research is always a constant challenge.

0:00 Yeah. And and I would add to that too that made me think of the flip side of this which is to say that just because something launches doesn't always mean you had impact right like if you did research that contributed to a launch but it didn't fundamentally change there was no like before and after that you can point to where you're like the research actually made the shift happen. I would argue that that that's not super impactful research, right? You sort of got swept in the current of the launch and you did some stuff but no one can really point to that versus like you said Molly even if you prevent if you prevented something that actually is a shift in direction that can be attributed to research and that product may or may not be launched but it's really about those shifts and less about less about the launch.

0:00 That's super fascinating. You mentioned this a little bit at the beginning and we we have to talk about AI. Is there an example of maybe a conversation where someone reacted strongly to the idea of AI but then you went in and you shifted the discussion towards the human need which is a lot of what your research is about. Is there a moment where then the insights or the approach changed completely?

0:00 Yes. I would say there's probably a particular category of conversation that comes to mind here which is around the value proposition of using AI in a product. what is a thing the AI is good at or is good for I mean and should do or should not do in a particular product. And so one conversation that comes to mind is there was a group of um engineers that I was working with who really wanted to use AI to do like the technically hardest thing that was possible to like achieve the technical feat right organization. They want to publish a research paper. They want to do something that's really really technically hard to achieve. And I was able to show with some of the research that I did, the human- centered research, what are the real priorities for people, right?

0:00 And where are the gaps and needs that we could address. And it's not to say um we shouldn't pursue this technology, but it's rather to say let's pursue an application of the technology um that is something that aligns with a real gap for for humans and and also that this thing that they were interested in doing is actually not something that users were really caring about or excited by. Um because although it was technically difficult, the end experience was not something that was that compelling compared to other use cases that we could have been pursuing. And so I think it's really about just like making sure that we aren't just enamored with AI and sprinkling AI into everything because we can, but um you know going back to the basics and like really focusing on the reality of people's lives and making sure that um our value propositions for what AI is used for are aligned with that and having a really strong sense of what that is.

0:00 Yeah. Having some intentionality behind the application of a thing.

0:00 Yeah. Exactly. And if you look at how we have LLMs today, it's like it's just out there in the world and we're figuring out what is it good for and what is it not good for and where can it be actually really dangerous, right? Um and I think that uh it's quite common to to encounter a very tech first approach. Um, and so a lot of I think the coming in with human insights when people have a strong opinion about AI is um coming with a human first approach as opposed to a tech first approach.

0:00 Totally. Well, Pety, we have segment on every episode called the current 101 where we ask all of our guests the same question. So the first is in your experience, what is one trend or practice in market research um or UX research that you would like to see stop and what is one thing that you would like to see more of? Yeah. This is something I would like to see stop is the trend of AI powered tools that market themselves as ways to do qualitative research faster and at scale. Um, and I think and the reason for that is I think that there's a misconception here that qualitative research is unnecessarily slow and that it's unnecessarily constrained by the limits of a human researcher. There's this assumption underlying a lot of it's marketing I guess that it's better to be faster and it's better to always better to talk to more participants but that's not true like you know things we know qual research has fewer participants by design because you don't need 500 participants or 5,000 participants to find patterns and in-depth interviews and immersives and deep conversations people I think are see some of the few deeply human touch points left in the tech industry.

0:00 Um, and so I think if you know my belief is that if you're on a team where the qualitative research feels too slow to your stakeholders, I think the issue is often that you're not designing a research program well enough or sharing useful insights fast enough. Kind of going back to the sushi conveyor belt, right? I think it's like the perception of speed. There's an issue there and just using an AI tool to make the methods go faster I feel is like a band-aid solution, right? Um, and I think there's a place for things like qualit scale. There's a place for synthetic personas just like any other method. But I think what I really want to see stop is the suggestion that AI should be used to shortcut the qualitative process.

0:00 I think the human process is the point and if it's like too slow, find a better human process um for that work. So that's one thing I would like see. And then one thing I want to see more of is immersives and creative field research. Um, I think that was really big pre- pandemic being you remember Google for a number of years had this research van that would drive out to different places and it sort of like had the the lab contained within the van which was really awesome. I think post pandemic in the industry at large. I think field research never quite recovered but I am seeing now a resurgence a lot of it is in response to these AI tools a resurgence of really like getting back out there um and doing it old school old school that is a space I actually get really excited about AI use in because I'm like what can AI tools now give us?

0:00 How can they augment field research? how can they augment, you know, being being really immersive with people and, you know, whether that's wearables or ARVR or translation or, you know, whatever that is. Um, what can the tech do to field research? So, I'm excited to see that resurgence and where that goes.

0:00 Yeah, I love those. Those are great.

0:00 Yeah. And I think it's having that discipline to know what can be sped up and what still needs to remain with the researcher. That judgment is always I feel going to be there no matter the technology.

0:00 Yeah. Exactly. And I think that there's like you I like to think of those two categories of things you mentioned as a set of research tasks which are where you're on autopilot where like I know exactly what I need to do and I just like got to go do the motions and it's like great

0:00 repetitive same thing.

0:00 Yeah. Yeah. Where you're not really like the process of doing it isn't teaching you something. But then there are many other things and some of them are very wrote like note-taking for example is an example of something I think is like very wrote and it is you are doing so much processing while that's happening that you're learning a lot through that process and so it's very tempting to have like AI noteaker but like you process things differently when you don't take your own notes or don't have a person who's you're collaborating with taking those notes right and so I like to think of like learning by doing tasks and autopilot tasks and like yes automate the autopilot stuff but um where you're learning by doing it, even if it is wrote, even if it does take longer, like find ways to speed other stuff up and keep that stuff really human.

0:00 Well, and I love that example in particular because to me that's an area where the answer is both, right? Like I totally agree with you and I I've seen all the studies on like the importance of like taking notes yourself, right? But boy, it's sure handy to have that like recording or or that transcript to return to as well, right? So, it's sort of like, like you said, it's augmenting something that still should sit firmly in the in the world of of the human researcher.

0:00 And I think it comes down to what I have found in my practice with AI tools is that it requires breaking down your tasks into more granular subtasks than maybe I did before because it' be like this specific part of this task like the transcript, right? Like it's fantastic to have really accurate AI transcription. I remember like 10 years ago really struggling with like maybe the transcription the automatic transcription would capture like every third word correctly or something and today it's amazing right but it's it's sort of not letting go of still some of those practices like taking notes as well that um keep you engaged in the work that you're doing and there's studies too about handwriting notes the the mind and connection yeah I'm the same way I have like my writable tablet that I still always used for everything.

0:00 Yeah. And I think I think ultimately it's like the value proposition for AI. I feel like even outside of use and research just like in the world is often speed. Like I feel like that's the number one thing that is the AI's market is like making you faster. And I think yeah, I think there's um a lot of work to be done about really examining like what is the benefit of slowness? You know, what is the benefit of speed and where is that beneficial? Well, so if you don't mind, like in a world where speed is not the most compelling value proposition, what what is it just in I mean, and I know it's going to vary by use case, but when we think broadly about consumer use of AI, is it like the mental load, the reduction of the mental load, like what what do you find resonates more than speed?

0:00 Yeah, it's a good question. I think well, it's funny you say mental load because I think that AI often increases mental load, right, but I also think it's hard to say definitively this is what AI should be used for. is a broad bucket, right? Because it also depends on the maturity of the technology and what it should be used for. I think the way that I think about it is, you know, supercharging yourself to do things that you may not have like that you would never have even done before or you would have wanted to do but couldn't do because you were sort of, you know, limited in a certain way. And I and sometimes that might be speed. I'm not saying speed is useless. Um, sometimes that might be speed, but I think that, you know, when it comes to research in particular, I think a lot about times where many years ago I've sacrificed things like being able to show illustrations or concepts to users because a designer didn't have bandwidth to do that and that was not priority thing, right?

0:00 And so it's like, oh, it would have been really nice to do this, but I actually have to move forward in this kind of suboptimal way. Um, and so I think like that's a really great space that now researchers can actually use AI to, you know, generate something. It doesn't mean take the designer out of the equation, but it's kind of like doing some of what they may have considered that busy work away, right? Same thing with presenting your insights in much more compelling ways by creating a website for them or creating something interactive, right? So, it's maybe not the most eloquent answer, but I think that that there are ways where we can supercharge the things that you wouldn't have had access to otherwise.

0:00 that that's really compelling. But I think when it comes to any use case of AI, you have to think about what's lost and what's gained. And I think you have to sort of do that calculus a bit. I think in the most ideal sense like the I I think that the optimistic view for me about what I think AI could do for people generally regardless of product is really help people with identifying patterns and identifying and like help working with unexpected change. Like I think that that's so needed and I feel like AI is actually really good at identifying patterns that humans can't identify and predicting things that humans may not be able to predict and seeing things before we can because of this access to this huge corpus of data.

0:00 And I feel like that's really like underutilized in products right now. Um because this thing of like you can just do more and like generate so much more is being pushed. Um, I think we're aren't focusing as much and we should be focusing more on where does AI have superpowers that people don't have and some of the stuff that is happening in the AI world of like things like drug discovery and things like that quite interesting to me because I feel like we've moved beyond like literally humans can't do it. It's not like generating a LinkedIn post for you. Something that humans can't do and something that AI is like uniquely good at. Yeah. And and I think about AI as like I think about AI as a material and not a tool.

0:00 Like just like you would build with concrete or you would make a shirt out of cotton, right? Like these materials behave in different ways and they have unique strengths and limitations, right? You're not going to build building out of I don't know some porous material and then be sad when the rain leaks through it, right? You're just like don't build with that. Don't do that.

0:00 This is the architect talking to us.

0:00 Yeah. Exactly. I think there's parallels there. I and so you think about AI as a material like you end up thinking about what can this thing uniquely do? What is it really good at? And like what is it not meant to do and like let's not fight that thing. Let's just go do the thing that it's uniquely good at. Yeah. So so I'm all for for those applications of of AI. Well, to wrap this up into I think one piece of advice, maybe that might be a little difficult, but for somebody who's trying to build a career in researching technologies that don't exist in a world that's changing, a world that prioritizes AI use, a world that prioritizes speed, and a lot of tension points there. What's one principle that you hope these aspiring researchers never lose sight of?

0:00 Yeah. So I think maybe echoing something I said previously um when we were talking but I really think it's that you know new technologies are always emerging but human needs are enduring um and the most fundamental needs of many people even though the needs are basic they're very complex and tied up in really complex systems um and they are not fully addressed and so I think that helping build the bridge between new capabilities if you're like on a product team as a research helping build the bridge between those new capabilities and fundamental needs in really actionable ways. So like deeply understanding technology and deeply understanding the human is really a superpower if you want to be in this career.

0:00 Like I think it's important to not lose sight of the fact that some really basic things for us in our lives are not solved problems. You don't need to be going after the next flashy thing. There are lots of really quote unquote boring but fundamental things that are very broken um that we need to think about how new capabilities might intersect with.

0:00 That's an amazing and honestly refreshing mindset. Instead of just chasing the shiny things, worry about what makes humans fundamentally humans. I love that. Priy, this was such an engaging conversation and there's a lot that I'm going to take away, but one of the one things that's that's sticking with me is to remember that the further that we look into the future, it becomes even more important to consciously recognize that at the end of the day, we're still doing research to understand people.

0:00 Absolutely. I also love the reminder that great foresight research isn't about guessing what's coming next. It's about identifying the human needs that will matter no matter how or in what ways technology evolves.

0:00 Yeah. And your point about measuring impact of research is also very important because sometimes the most valuable research isn't just for the product that launches. It's the way in which the research changes how teams may think about even before they figure out exactly what they're going to build. Tracy, thank you so much for sharing such a thoughtful perspective on anticipatory research, human- centered AI, and what it takes to study a future that's still being created. Thanks. Yeah, this was really great.

0:00 And to everyone listening, thank you so much for being part of the Curiosity Current. We'll see you next time. The Curiosity Current is brought to you by AYTM. To find out how AYTM helps brands connect with consumers and bring insights to life, visit am.com. And to make sure you never miss an episode, subscribe to the Curiosity Current on Apple, Spotify, YouTube, or wherever you get your podcasts. Thanks for joining us, and we'll see you next time.

Produced by aytm

Curiosity Current is made by aytm, the consumer-insights company. Guests speak for themselves; the synthesis is ours. About aytm