UX for AI

EP. 82 - How AI is Reshaping Creativity and Productivity in UX w/ Aswin Ranganathan

• Bonanza Studios

Share episode

Send us a text

In this thought-provoking episode of UX For AI, Behrad and Aswin Ranganathan dive into the exciting and rapidly evolving intersection of AI and design. The conversation explores how artificial intelligence is reshaping the creative process, redefining the role of designers, and unlocking new possibilities for productivity and innovation.

Aswin, a designer-turned-educator, shares his inspiring journey from working in traditional design roles to becoming a full-stack developer and eventually founding his own academy. He discusses how his hands-on experience with development and AI tools has given him a unique perspective on bridging the gap between design and engineering. Aswin’s academy focuses on teaching designers and non-designers how to build AI-powered applications, Figma plugins, and productivity tools, empowering them to work smarter and faster. He emphasizes the importance of embracing AI to stay relevant in a competitive industry and shares practical insights into how his students are leveraging these tools to create real-world impact.

Behrad, the host, brings a visionary perspective to the conversation, diving into the future of UX design and the rise of AI agents. He explains how AI agents are transforming the way designers approach problem-solving, enabling them to create more personalized, adaptive, and responsive user experiences. Behrad highlights the growing importance of prompt engineering and how designers can use AI to generate dynamic interfaces, streamline research, and automate repetitive tasks. He also explores the potential for AI to revolutionize collaboration between designers, developers, and stakeholders, making the design process more efficient and inclusive.

Throughout the episode, Aswin and Behrad discuss the challenges and opportunities that AI presents for the design industry. They touch on the ethical considerations of using AI in design, the need for designers to become more technical, and the importance of staying curious and adaptable in a fast-changing landscape. Aswin shares examples from his academy, where students are learning to build AI-powered solutions without extensive coding knowledge, while Behrad provides a forward-looking view of how AI agents will redefine the role of designers in the years to come.

The conversation also highlights the importance of community and continuous learning. Aswin talks about the value of collaboration and feedback in his academy, where students work together to solve real-world problems using AI tools. Behrad encourages designers to experiment with new technologies, take risks, and actively participate in shaping the future of design. Both agree that the key to thriving in an AI-driven world is to embrace change, stay informed, and continuously refine your skills.

This episode is packed with actionable insights, inspiring stories, and thought-provoking ideas for designers, developers, and anyone interested in the future of creativity and technology. Whether you’re looking to enhance your skills, explore AI-powered tools, or simply stay ahead of the curve, Aswin and Behrad’s conversation offers valuable takeaways and a fresh perspective on how AI is transforming the design industry. Tune in to discover how you can harness the power of AI to unlock your full creative potential and stay ahead in the ever-evolving world of design.

You can find Aswin Ranganathan here:
https://www.linkedin.com/in/aswinckr/
https://www.moderndesigner.ai/


Interested in joining the podcast? DM Behrad on LinkedIn:
https://www.linkedin.com/in/behradmirafshar/

This podcast is made by Bonanza Studios, Germany’s Premier Digital Design Studio:
https://www.bonanza-studios.com/

Welcome to UX For AI. In this episode, Behrad and Aswin explore how AI is transforming the design world. Aswin shares his journey from designer to educator, teaching others to build AI powered tools and Figma plug ins. We'll be also diving into the future of UX design, discussing AI agents and how they're reshaping creativity and productivity. Tune in for an inspiring conversation on embracing AI to stay ahead in the ever evolving design landscape. So this is sort of our own curiosity. We might a bit a bit more nerdy than the rest of the folks that are listening to this, and there's a lot of folks that I've seen that since my last post. Subscribe to the podcast. So I want to bring it back to sort of like where the collective brains or awareness is when it comes to AI spring, and maybe it would be really helpful for the audience to get a bit more about your background on what you're doing and how you end up with AI. That could be really nice. Necessary preface to our conversation. Great. Yeah, that's that's a that's a perfect way to start because people would be wondering, why should I listen to this guy? Like, you know, that's that's that's a good, good point. At that time before I, like, start setting my opinions. So. Yeah. In the design field, I have, like, close to a decade of experience. I work with different sizes and shapes of organizations and different complexities. Worked in, like, start ups in India to like companies like Grab in Singapore, which is like a much bigger organization. I worked on products that, like, have reached like hundreds of thousands of users at least, or have used in some way or the other in something that I've built. Not only have I like design something, I've also seen how it worked. Took some learnings, iterated on it, which is also a critical part of my, design. Shaping process. In terms of how my opinions have evolved. And during this time, there was also a phase in my, career where I decided, okay, you know, why am I stopping myself with design? I could probably be also building things. And, so I kind of took a couple of years out of my design career. When I fully transitioned into, development. So I became a I would say I would, I could call myself a full stack developer in some way, but it was all self learned. The reason I, I mean, I a calling yourself like, I'm not a skilled full stack developer, but I acquired most of the skills required to be a full stack developer. So, I wasn't like, I didn't have any professional training, I didn't go to college or anything. So so I tried to learn all of them on YouTube and stuff. So I kind of learned like bits and pieces of everything, and I try to make something with it. So I learned some backend. I learned like some databases, I learned some SQL, I learned some, I built like, an application with Django. I learned, sockets. I learned some frontend frameworks so I can. I was all over the place. So I realized, like, you know, after, like, trying really hard to understand, like, what was this things and try to build something out of it. I went through that, like, really, it was a very difficult journey to learn this and grasp it and try to build something, especially with having no prior knowledge. And, at the end to make something super simple was very complicated for me because it was just like, well, I learned so much. But to make some a small product, it would it would take such a long time. And it was, it was really like then I started to question it. Is it is it really worth, you know, going down this path? There should be an easier way. And that sort of like where I left off and then I went back into design. Then I was like, okay, you know what? I'm going to like I have all this development knowledge. I like designing, but like, you know, making the whole thing is like super time consuming. So it gave me a unique advantage in my design because I had that development exposure. So, yeah, I was able to uniquely have conversation with engineering. I was able to, come up with solutions that were not just like user driven, but also had like some technical considerations to it, how we could launch fast, build fast. The test fast was something I had a lot more to add to that conversations. So yeah. So these these were some advantages that I picked over time. And then I thought, I like this whole, like building with culture building with, Claude, things kind of like, changed like my, like, I was mind blown looking at how much time I had wasted. I've not wasted how much time I had spent. It was just like nothing, you know, it was just like I was able to do it in, like, a week and. And that was like the moment for me. And it was like, okay, things are going to change right now. Like things are getting real. So I took it upon myself, like, what if I can teach this to someone is, who is not a who's not a, a developer. And if they're able to do it, then my theory holds true. You know, like, this is like, if it can be done by a non, developer, then maybe VR, like everything that I'm thinking is going to happen is probably happen. That's sort of like how I started and I started teaching and I built like Modern Designer. I started the YouTube channel, and pretty soon I started seeing results. Like I started like, people are building plugins. They start, they build plugins like Figma plugins, and they tag me and they're like, oh, I found this job because of this plugin. And like, people are like at my company, people are building, like different internal plugins from the courses I built. So indirectly, all this stuff that I was teaching was creating like real impact. And previously people who had no development background were now building things. So I was like, okay, this is really working like, this is this is something really big happening here. And from that point of view, I kind of have like so sort of like opinions on how this is going to change, like where the whole industry is going and how the whole productivity system is going and how like our designers roles might evolve with that. So that's kind of like a quick intro into like when I'm speaking from. So there is a lot of different direction I can take this conversation. But maybe, perhaps we can double click on your read. Current UX design, product design trends, maybe go a few years back and then come all the way to now you have a design leader role. So not only for your own sake but for your organization. See, you need to basically look at the trend and see where it's headed, which will affect all the decisions you'll be making on a day to day. I think that's a really good place to start with. Yeah. Interesting question. I don't know if I can, plotted across a timeline and keep that the same across all organizations because I think it differs based on where you work and how your organization is structured. But I can speak from my, that my experience, when I started working in startups in India, things were like, very, messy, I would say, like, things went like. So there was there wasn't, like, really a design process for me, you know, to be very honest about it. There was just like, we need to build this. So it was just like, okay, let's make something cool. And then they would just be like people, oh, what if we did that? That looks much cooler. You know, that would be more interesting. And then and then people would be just figuring out how to build it. So I kind of started off at that space and and I'm talking about a reasonably successful company, which was like working in this model, like the company, I mean, it's called Housing.com. And like, yeah, they were at the time pretty successful for building, quite tech savvy stuff because they were, they were innovating a lot in that short, and honestly, like, it did push a lot us a lot in terms of how we thought about our interactions in general, like visual stuff and everything. But it was much less informed on like the UX research or something like that. In fact, I was I was one of the first ones to bring UX research into the into the company, into like as a, as an. Yeah, like as an intern. So, this was not an intern was my first job. So, so it was like I started from there and then I moved to, like, slightly mature companies where there was like some information that was available. And then when I moved to grab it was a lot more mature, like we had a knowledge base that we could refer to. We had like a research team. We could get that help on doing conducting research. If we didn't have any information, we could like go and ask, we could set up like user research sessions with agencies or like, organize these with like based on the requirements. You would go with discussion guides. And we had like more thorough research process. So typically what it would look like this. Right. So you have a product requirements from a product manager. Then we would like come up with what the research requirements that we need to do. We need like we need to do a direct interviews. We need like some secondary research. What do we need. Whatever we could do we do it ourselves. We whatever. We need a research team. So we get that help, and then we get all those findings and we, we estimate how much time we need for the entire deliverables. And then we plan accordingly with the product managers. Then we go into our like whole ideation session. And then there's like back and forth feedback and try to refine it until we get it right. And then there's another testing, session after we get the final designs. And testing is also done with the help of the UX team, the UX research team. And finally. Yeah. So this is this this was like our mature design process. I would say that we were following at my previous company. And even here, like, we follow very, the current company that I'm working in, we also follow very similar mature design process, for specific projects in some specific projects, we it's a project based. It's like, you know, for more smaller, features that we need to launch in much shorter timelines. We probably won't go through an extensive, research, study, but we would use whatever insights we have available. And best case scenario, can make the most informed decision. Yeah. So I think this has been a the traditional traditional process of research, research and design in general. But looking at where it's headed, I think, yeah, that's where things, things are a lot more, confusing and unclear for a lot of people. And I take as well and I think the, the I started to see a little more clarity when I started to question this from a different point of view than just looking at it from a design process point of view. Let me explain. Like, so like if you, I think still the pieces that we are discussing where like, you have to talk to your users, you have to, do your research, you have to somehow, you know, like, come up with different explanations. You have to still define it. All that process still is going to stay the same. There's nothing changing in that. It's just that when I get mixed into this, a lot of, productivity gain is going to, get mixed into this. So, like, things are going to get faster, things are going to get more efficient, things are going to get, like where a designer, a researcher would spend ten hours, would not only need an hour or two, you know, that sort of thing. So now the question is, well, where's that eight hours going to be spent? So like that's where that's where the real question comes. Right. So then like you would be like, okay, either as an organization we move faster, we build more stuff, we ship more stuff. Then the developers also have to ship faster, which is now possible because you have tools like cursor, you have tools like, you know, GitHub Copilot that just basically codes everything with it. So now we can ship things so much faster. And now like will design be able to catch up, you know like now we'll design. We'll have to use these methods and try to catch up with that speed to be able to match up with that thing. Or you could like, you could generate many variations and then you can test them in different environments. So there's a bit of how we test might also change. So I think if you take out individual pieces in the whole process, I have, I is having influences on those pieces. And you know, like how they say like like a designer who does know how to use. I will have definitely a better advantage in terms of being being faster with, with their work or being more efficient with their work, or just like creating better documentation or, or coming up with a workflow that the entire team can benefit from. And yes, I mean, as an organization, we move to this like faster delivery, direction. This is going to matter. Like having the AI skills is going to matter. So yeah, I mean, there's no one process that defines, that is defined like how we could define it from back to now. But this process is being invented right now. And like, it's interesting that we are the ones creating that, that shift. And we are identifying where like, you know, these process efficiencies can be improved. And doing that with I. Know if I want to add to what you said, I think there is a lot of merit to what you have shared is that AI is not going to replace, for example, double diamond process. It's not the intention, in my opinion. No one wants to replace the creative process because without the creative process, how can you go to understand the problem and then come up with solution, test and idea prototype, whatever the case, AI is going to augment? To your point, every step of that process to a level that was not imaginable before. To your point, if you used to spend eight hours for one unit of research, now you can do it in two hours. That's like six hours saved from that particular part of your process. Not addressing, not integrating AI. It's a massive risk because just if your researcher and the other researcher compare your work to the other research on the other, and using AI and integrating into every step of the research process, the output that they're providing is going to be larger at scale, with more depth. Most likely if they get it right. So it's not even a comparison here. So I think that's what really important UX is, not that creativity is not there, but you are blessed with a very powerful tool that could help you at every step of the time. Double down the process or any other process that you're following and give you the results that are not even comparable to what you're doing right now. Yeah, it's it's not it's not the what. It's like the how that's that's changing. Right? I mean, it's not like the way we come up with things, is fundamentally when, when change I think like it's still we are we are building something for some which are this is a specific problem. And the whole problem discovery process, which is the whole double dipping process, is the same. Like you, there's a discover and there's a design like it's I think it's pretty straightforward. Yeah. I mean, yeah, you're right. I can completely agree. Let's but maybe like look into the different parts of the design process, like go from research to prioritization, ideation, prototyping, testing, any step that you think, for example, wow, this this part I could do a lot. Maybe if you have some examples up your sleeves, I'm using this tool. That would be fantastic. The reason I'm I would like you to go in because I know that because of your academy, you're sort of like in the trenches playing around with these tools. So maybe our audience will pick up on a couple of these tools or like tactics that you're using. There might be some like, you know, moments for the audience listening to this. Yeah. So from my, observation on biggest when, like, what has already worked, our biggest win has been, a few areas. One is, Figma plugins. This this has, like, you know, like, we there was this, a person in my, in my team who built, plugin for, linting, like icons. And if it's using the exact design system, I can, and that plugin is already dead, and we integrate it. And, last I heard, this person is like, shipping plugins, like, you know, once a day or once a week, which was you can never imagine something like that. And someone in your team working the design systems team is just able to to build this, already show us what, like, you know, you it's, internal productivity audit. It's just making design systems more effective, right? Like, it's just ensuring the usage of design systems is the way it's supposed to happen. The Figma files are documented and used properly and that sort of stuff. Gen AI is, of course, another aspect that like, I'm pretty, close to, the project that, I mean, one is we are using, generic for like, icon generation, content. Our content team is using generic for like some content descriptions and stuff like that within, products. You know, food descriptions. There are quite a few, like, areas that being identified. Region is and I'm sure like customer support is also, used. So this is already like taking, starting to shape things, like, like just creating a knowledge base is also a big, this is also happening in, like in our company, and it's also a big win that, UX team can have, which is like, just put all the, product requirements into a knowledge base. Right. For all the competitor analysis, not the, not the product competitors, but more like the design details later competitor analysis into a knowledge base, but all the UX, UX research into a knowledge base. This is huge because now when when your like coming up with like your designs you it's just a chat bot away from finding out information related to how you come up with your explorations and well, you would not have like, you know, not have considered all of these if not for the the easy availability of the tool. Like, I mean, unless you did put a lot of time to in your process to go dig up this information that will inform your designs, which I'm sure not everyone is going to do. Now, at least this makes it more accessible. So I think I think this whole, serializing all your documents, putting into a vector database and just making it accessible is a huge win for for a lot of it. Hey, and, like, you know, when this I was earlier having a conversation with my team was that one of the biggest challenges especially senior designers have is on alignments and stuff, like, you know, when you're when you're in, when you're in conversations with other stakeholders and you have to like, consider different, the constraints and stuff. Well, I mean, these parts are still like really rely pretty reliant on like yourself, scale your communication and everything. So, yeah, there's still not a lot of AI intervention there. I think, but we're still exploring. So but one big, another big way I would use AI is like if, like if I were to generate something with AI, let's say I want to generate a screen for checkout, right. If I put it on a tool like bolt or V0, it will be a generic one. It won't have. Like any context of the challenges, the team configurations, the internal dependencies, the all the roadblocks, every decision that went through before all of that stuff. It does not have any context. So it's going to come up with a generic design which is not usable. So if it's not usable, no one's going to try it out. But if I'm able to create a knowledge bank of all that information, which I can send it to bolt, or which I can say send it to V0 to consider before coming up with the design, where things get different now. So now the question is who's taking the initiative to create this knowledge bank? Right. So we all are trying to come up with things without having all the foundation we need to make. I work for us, so we still need to create that foundational work. We need to still create that that assets we need for us to be able to generate. It's like crafting the right prompt. So the right prompt is like like we we need from the design, say someone to be like, okay, you know what? This is the right prompt. This is the, you know, the agent that refers to these, vector databases for this knowledge and these tools for like, you know, making these calculations. And now based on all of this now it creates this, like, design. Now the designs will be far more usable, far more effective. Right? I'm sure that Figma is working on it. But like we, we either wait for Figma to do it, which is a whole another conversation. Because like, well, Figma is gonna try and retain into its Figma ecosystem where it's going to try and understand your design system at most, let's say it does a great thing, but it will still not have so much context on your product requirements. Unless they build a feature where they can understand your product requirements as well, which kind of goes out of its scope. So I don't know is that and we're going to happen. But as a design team, you have that ability to bring these AI pieces together and connect them together and make it work for you and make it create something for you, which I honestly like. Yeah, I feel like more AI enthusiasts like us are the ones to be taking that initiative and exploring it from our front and be like, you know what? We believe in this. We've seen the value in it. Let's make it work for us. So we are the ones to like, try it out and, you know, show the world that this happens. But there would be a lot of people who would be skeptical about it. But, you know, that's not helping anyone. Like, but we've seen that this is transformed, like crazily in the development aspect. You know, you just like literally you just said and it's built. Right. I mean, what's like, how different is it from like just having a little more context on all these like, internal challenges and just coming up with ideas? I mean, it's it's not far away. I'm almost getting some goosebumps here because I think, you know, VR at the you know, UX has gone through a lot of gold rushes. Right. You remember that Drupal were filled with different like it was a fetish for dashboard design, right? IOS app design. Right. That was a gold rush of you. Works. Right? For very good reasons. Right? We needed that, like, sense of esthetic and enchantment that, you know, okay, this is an app. This is all its pixels, but it feels is special. I think a lot of UX designers feel insecure because. It's I think there is no surprise anymore that we are going to face we are entering a era in the digital, product evolution that interfaces are going to become less and more fluid and more responsive. So, okay, do all you do all you want with a menu. But the menu is a menu has input fields dropdown that, that that that's your menu okay. You want to see the Figma file and like tweak it and tweak it and push every pixel. That's you right. That's your decision. But that's not going to be the real any innovation anymore. We've gone past that phase. Us as UX designers and product designers and innovators, we need to accept this fact and embrace it more. So, like you said, something really important. I think I truly believe that's going to be the future of UX design, the future of UX design, not. And also the UX designers are product designers that there are research savvy actually is going to be the golden rush for those folks because for the past six years, as as long as I remember, UX research is complaining that no one look into our research anymore. We do a lot of research for weeks and weeks. No one take a look at it. But in the future that most of the apps there will be AI agent integrated into them, or there will be some responsive and. Real time interfaces that change based on user feedback. You're. I think the work of UX researchers, UX designers will be about. Creating the service instruction or AI agent instruction and maintaining it and developing it. So unlike your boss or executive team that they don't look at your research, they learned to be able to. Basically, set up that agent. They are going to look into your research with all its nuances. And because they have to make but we need to make sure the AI agent is going to provide responses that our company could put its name behind it. Right. So all your research that you're doing and if was not being seen now, it will be seen, it will be incorporated into the output of any AI agent or sort of hybrid interface agent applications. And that's why I'm really excited about the new phase of AI. To your point, is going to be really different. And it's us as basically folks that are at the forefront of the UX design to basically chart the path for the rest of the designers. But it's going to be a very exciting era because now you are our job is transition planning from designing interfaces to designing the behavior, the interactions between the AI and the users at the same time? Absolutely. Yeah. The designing agents. Yeah, 100% is is what's truth? The truth and what you said. Yeah. Designing agents would be a very important skill to learn as a designer. I truly believe in that because, I mean, think about this, right? I this is how I, I'm working on a product, and this is the exact steps I followed and also inspired from some popular creators in this space. Right? I had a a basic idea of what I wanted to build. I put it inside ChatGPT and said, like, here's a template of a product requirements document. Create a product requirements document it made like and the own model gave me a detailed product document. I made some tweaks to it and stuff like that. Again, I took took that product requirements document, put it inside, own model. Again, I said turn this into a UI spec and it took the full build. It gave me the complete UI specs for all the screens that I needed with the entire flows. What create, which button should interact, where, like all in a text document. Right. So I had this product question document. I have this UI spec document. Now I put both of them into Curser and I said no design my product. The product was there. And and literally I just had to go make little tweaks and the product was ready in like an hour. Now, if this world is is dead, this world is already dead. We are here. And this was possible like the creation of product requirements to the the creation of the design spec can like a product manager can do that. Like where is the designer going to going to come here. Right. So like the generation of ideas itself or coming up with the UI for the product design is is done like this is it took care of it for myself right. So then the real question becomes like, like what a what a VR designer's doing in terms of these agents building. So plugging into the context that you brought like now this agent sitting and analyzing my US spec, my product requirements. Also, if I design this agent to look into my research, also design this agent to go, look into much more organizational, the, you know, details to consider some other like, ownership related stuff like, you know, only touch on these screens but not on these screens or whatever. Like if my agent was smart enough to, like, figure out all of this, what for coming up with UI like that becomes a designer's job because, like, then it becomes a lot more effective for the designer to, you know, to come up with good stuff. But, I mean, it's kind of debatable because, I think eventually organizations will move to this model of working because like, I saw this, like recent post from like, I think PolyGram on on X, where it was like there was some company that moved, to, replicate directly, from Figma. Well, yeah. So I think that kind of also indicates a shift in the same direction. Right. Because if you if you have Figma and you, you design something on Figma and then you build it versus you start something with a spec and then you already have something built and then you make modifications to it, things out much faster than the second way, right? And much faster because, you, you have something. And making changes is also not very hard. You, just as a designer, have to just go say no, make this little thing if you have a little bit more skills where you can work with tools like also you can like easily make changes to your talking about web, but you can also do it on iOS apps and stuff. But yeah, like and then leave to the developer to handle all the, the testing, the, the scaling aspect, the connecting to the backend aspect, you know, to, to connecting to the APIs to make sure the API deliveries is accurate, you know, the loading experiences. So there is plenty of things for a developer to do. While a designer can still do a lot of that while sitting on the technical side of things, you know? So there's that shift happening. So we would be like getting a lot more technical. I think, as designers, when this shift happens, like, you know, when this, when this will move towards AI development space happens. And we will be building agents, like you said, like we would be building agents that, not only uses the context of that prompt, but also uses internal organizational context. 100%, I think the behavior of the agent, I think agent by agent, company by company culture. But culture is different. Like, I mean, you cannot expect that. For example, Zalando agent will act the same as, Nike agents. Right? Because the, the brand vibe and culture and mood, them feel and mission and purpose are different. Right. I think us if we claim to be we are designing experiences, we are designing characters. That's where we can influence a lot. Right? There is a technical aspect of it. I think. I mean, what you're doing with your academy, I think it's really important to design and that's it. I think there is no other way around that we need to get, into trenches with technical stuff, not like to level that to your point that you can create the, you know, you can fine tune the backend logic and make sure your database modeling is intact. And that this is a scalable is I don't that the the technical core is coming in, but enough to be able to, create the front end at least while this like, you know, and then, you know, get it to a point that you can pass it on to a developer, right? I think that these are the areas that we need to really like, say, okay, this is the reality. We need to get in the trenches. But beyond that, I think I believe like once you start like building one agent, then you see like, oh, actually it's not really difficult. Let me see what other agents I can create to facilitate my work individually and my work with my team and my work with the organization. Right. So the I think I think it that's why I'm really excited. That's why like I just changed the name of this product actually works for a because, oh, it's so untapped. One question regarding Vercel and the other one is Replit. Replit. Replit. Yeah. So, your process, you created the so you had the product requirement, you created the interface, specification, and then you move it to the verso to basically create the front end for you. Right. I, I created the product requirements, and, you edit spec and put it on course or directly, and, I have a, I start with a template, like a basic project template, and then ask user to use these two files. And I would say, now go design the dashboard. And so now it will understand from the UI spec what's needed in the dashboard. It'll understand from the product spec what is needed for the entire project. And then I will go create the dashboard. And does it have a Figma integration that you if you see that, for example, part of the interface, I could just do it myself instead of going back and forth. I had, figma reference. In fact, I didn't even. Wow, I couldn't even I didn't even imagine what the product would look like. Like, I just that's what I tried. I pushed myself to do it right, because, I mean, this is this is the course I'm making within my, the course on my page. So, like, I wanted to challenge myself, like, what if. Can I just do all of this? Like, I don't even know. It would have taken me some time to go do these explorations on on Figma. Try to think about what this thing would have been like, which is nice, which is creative exercise. I love doing that. But I was like, like I wanted to really challenge myself. Is it really necessary? Like, what if I what if I could just generated first and then modify it? Right. So but because I have my take to it, do I like it or not, I can follow my own opinions, but I don't have to like come up and design things when this is so much faster. Because just just after I finished, I looked at the interface and I was like, wow, this is really cool. Like, I didn't even think like I didn't. I had nothing in mind. I didn't know what to expect. But this is completely usable interface like and this is this will work for what I'm trying to build. Well, yeah. So that you could you could kind of say that's very similar to like, using Figma as I told you, just like, say a prompt and give it. So it's like same as bold or visible, but like giving this whole context of like, your project requirements and US spec makes it a lot more, intelligent. That's fascinating. That's sort of the process that we, you know, touch upon design, process and how we entangles with the project development of project management, side of things. So like if you if you zoom out in and think about applications, it's of that the user experience of the application. How do you see the typical application we are using, ranging from Spotify to a B2B application like HubSpot or Calendly or the likes, like, you know, it's a wide range. So I don't expect you to be saying, here's what I think about the future of Spotify. Here's what I think about the future of Canada. But what kind of like general trends you see, when it comes to the future of applications? Interesting. Yeah, I didn't think about it. And I wanted to, I wanted to, like, have, like, a general framework of how things are going. And the best reasoning I could make was that I see I am trying to, like, differentiate applications into two buckets. Like one bucket is like applications that are heavily, reliant on user generated content, and applications that are not so much user generated content, but more like service related stuff like, it helps you find something, it helps you, find a service or like, you know, or a marketplace or like an e-commerce. So that sort of stuff is one bucket. The other bucket is user generated content. Like, you know, you're making a LinkedIn post generator or like a social media, or a TweetDeck or something like that, you know, so on the on the user generated content bucket, I think AI powered apps has created the biggest disruption now. Like that's that's where the all the AI powered app focus is right now. So if you're working, if you're if you are in the AI space, like if you want to be a designer in the space, I think that's where you're probably going to look for a job because that's where you're like, okay, actively thinking about how to, like the UX, the AI, UX patterns and stuff like that. These are like stuff that we are going to be focused on. And yeah, that's that that is the changing, big time. While the other aspect isn't as much yet because the number of places where it's applicable is quite limited, like search, for example, is quite an amazing, space. Right? But yeah, the impact of using such AI powered search is, it's still like the question of like, not, not, not all organizations will want to invest in doing an AI powered search, because if the search results in so much revenue and that sort of conversation, it comes down to, like, I think there's a customer support which also exists in, in both, but in the UI powered, sorry, AI powered, UX wise, in this first bucket of like, you know, marketplace, apps and that sort of stuff. There's not not a lot of like changes that I see yet. But I might be wrong. Like, you know, when, when this whole generative UI becomes a norm, things could be much different. Like, I mean, the whole imaginative way to look at it is like, yeah, you know, you just ask and then the UI is created for you and the UI just adapts based on your part of the journey, based on your personalization and stuff like that. However, in this first bucket, I think the biggest impact is happening on the analysis side. Like, you know, the recommendation logic or the machine learning algorithm that, you know, runs on the backend, you know, so it's more on the data science team that is more actively engaging on the products on this bucket and much less on the user experience designers who are being involved in it, at least the I mean, if I would imagine if I was like, UX designer at Netflix, I would like because it's not a user generated content kind of platform, but the recommendation algorithm is should be an AI generated one. It should have, like, you know, all these machine learning algorithms that would create this thing. Should a designer be involved or not like that? That's a question, right? I think a designer should be, I don't know, like if they are typically, a designer should be curious on how these algorithms are written, like how how these data science teams set things up and at least share their opinion, share their like inputs into the whole thing. Also bring some user insights into that thing so that they are creating these algorithms. But yeah, so I like there are varying degrees between the two I think. So that's where but yeah, for in the user generated content space, it's completely disrupted. Like I think you're, you're like, you know, for every single form, you just start with a basic idea and you just use like, you know, generate and then it expands on your idea. So you don't actually type anything anymore. You're just like, just type out. What do you have? What do you have in mind and let I finish the rest? Or like, like, you know, generate, generate, case studies generate, things like generate everything, generate presentations, generate so many things like all of that is like completely different right now. Yeah. That that's going to like the how they call it as a ChatGPT wrapper. Which it is just sort of like the big goldmine right now. I take like for, anyone can build a chatbot wrapper and it's like, and it has a huge productivity gain, I think, and there are so many use cases that people can, like, benefit from, from these, like, little ChatGPT wrappers and wrappers are just like, like agents, right? Like you're just like building an agent that, that just like, takes takes something unstructured, something like a simple idea. And then figures out what information it needs to expand on it, just ask the right questions at the right time and compiles and creates a structured output at the at the end. Now it's just about how do you prompt that agent to go from A to B, right. So that's yeah. So I think yeah, this is also a critical skill that like designers can learn because like like you could, you could technically with just the knowledge of prompting, build this whole experience. Like how to go from an unstructured idea to a structured output is a is a great skill. Like you could apply that same skill to pretty much your productivity, agent or an agent for something else within your team. Or like, or your own app for all you want to do, you know? So, it's it's. Yeah. I don't know if I answered your question correctly, but. Yeah. So I see, like, these two, like two drastic, ways to observe changing, but depending on that bucket. That's a. Like that's a really fascinating way of putting it and categorizing it. And yeah, there might not be so much use case for UX designers in the first bucket, which is mostly about algorithm and like recommendation and prediction, there are going to be some use cases for sort of like showing the prediction or recommendation or insight to the users, but limited because most of the work is done by data scientists and mathematicians, so to speak, rather than even like I don't think programmers would be able to do anything over there, because it's just basically about can we offer a better recommendation when needed by users? Right. And that's purely mathematical and, data science, topic. But on Jen, I said, I see your point clearly. I think there is going to be for every use case, especially specifically the most boring ones. Right? Right. Now it's sort of like now Jim and I, now I can access it through my email in my, my Gmail. So they just release it to our workspace. Just summarize this email from someone writes this email threat to use cases like summarize this email or summarize this email threat. Well, one use case with one variation. If massive productivity gain, especially this massive threat that all of us are in, we don't know why. So if you are in and then summarize the whole thing, okay, this is this this. Okay, I have enough nothing. But before I needed to go through every email in this theory and read, okay, Alex, that the sushi and sushi set the cells and set the there was a oh my God, was going to now summarize that. Okay, do I need to pay attention based on the summarization? No. Move on. Thank you very much. That good for you. And why wouldn't you do it? Why wouldn't you use it? I think so it's time intensive. Yeah. And then I think the other Jen I said I think is going to be basically every, every app that used to generate something before it was user generated. So users need to come in and do something that the app enhances it. Now every app that used to do that is up for grabs. So if you come in and to your point, actually you could you know, when you think about the structure of a typical application, every application has more or less certain algorithm or secret sauce does something really specific, right. And there is certain programing backend logics in place to make sense of the algorithm. And this computes you need to compute it. It's a different it's a different case that you, if you want to generate one image versus how is an image you need to have compute. You need to have a system of a scale. Right now the first two parts now could go into a prompt algorithm. Like what what. My secret sauce is what my knowledge base is. Maybe your knowledge base is all the blog articles you've written in the past 20 years. That's your own specific knowledge. You want to turn it into recipe or secret sauce. That's right there. What does the AI out there is to do with that recipe? The logic could at least 6,070% of it could go into the prompt later. So, you don't need the developer to come in and do 60, 70% of what an app required to do the programmer needed to do to launch application. You could influence it right away from the prompt from the service instructions, which I think is going to be massive for any gen AI application. And that's what I think. If designers positioned themselves as folks, that they can really create prompts that actually does that thing really well. You position yourself in the organization or in a very good in a in a very good position. You put yourself in a good position because, well, the, the your lead team knows that, okay, you are doing actually the job of the programmer as well. Absolutely. I think the the challenge that designers would face when doing that is like, I mean, what is the context limit? There's there's only so much that you can say in one prompt, and at some point you need to that's why it's like prompt engineering, right? It's like you need to you need to think about how to break the break it down and they create a system. So it's like you're creating, an, an agent to yet break. You're saying that, look, if you want this information, go fetch it from there, decide what information you need first before going and looking for that information. So you're this is like pretty closely related to how you design flows. But there is no visual interface for this. It's just, you know, the, the agent interface. But yeah. Yeah. To your point, I'd. Right. I mean, this is, and the other challenge would be the, the cost, the cost for, I mean, yeah, this is something designers probably don't need to get into because, you know, for some, for some generations, you wouldn't need that, like, super expensive models. You you don't need, like, reasoning models. You just need a type model. So and for some of them, you need a super amazing one. And for some use, cases like Lama might work better for some use cases like deep sink might work better. So like to make that trade off is also someone deciding of how each model like works and what it does best. And like what models that we, they can use for that specific use case is also like a useful knowledge for designers to, to understand. Now time flies. If you're like 60 minutes into this conversation, I know that you're working on your own academy helping designers, non designers to really make sense of the, how to create agents, Figma plugins. Maybe as a closing remarks, maybe you want to touch upon that and like sort of like, tell folks how to find you what that academy looks like, what it does. Thank you about that. So, yeah, you can find me on, Modern designer or I, so what I'm trying to do there is, like, I started with teaching how to build, like, interactions, like some fun stuff with, with so with just AI. And then I taught, like some, I have some courses on how to build Figma plugins. Now, I'm teaching how to build full stack applications, without code, or just like very low code, which is just using cursor and on the backend completely using automation tools like anything. And so these are like build how to build AI powered apps. So this is also in my course. And soon I will like go into teaching just fully how to build anonymous agents for like pretty much most of you use cases like how can you, improve your productivity or like how can you create your knowledge base with AI? So you how can you be a ten x designer or a 20 x designer? This kind of like the objective of how can you how can you make AI to make yourself much more efficient than a non I design? That's kind of like the goal of how I'm trying to approach. Yeah. The the course details are there. And I also have a YouTube channel which is also linked in the website where I share most of this content for free. But the the paid course gives you a lot more, access. And there's a community to, to address specific problems. And I work closely with the people in the community to, to create this content. So, it's more relevant to the people inside. And. Yeah, I mean, everything is, this constant feedback loop. So I make sure I don't create something that's not valuable. So. Yeah, I think the importance of being part of community is so there is stated, in this, in the DNA, because if you know that if you are part of a community, there is 100 other designers doing and trying to learn this thing. You motivates you a lot to a great extent. I mean, thanks a lot for being on the part and sort of like be on this intellectual wrestling with me and, sort of going back for any last remark. But thank you for having me that I mean, it was, it was really it was really, nice to have that intellectual lesson. As you said, it is nice to chat geek out about. Yeah, topics like this. But yeah, I mean, like I said, right? I mean, this is, this space is whole. The whole space is new. People are all of us are just figuring out, as we go. But, yeah, I mean, I think this is it's exciting. As designers, we as design leaders should be the ones to be driving the change. And I'm pretty excited for, for, for the next, times, you know, next year or two until AGI arrives. Yeah, until the guy comes in, and then we don't have to do anything else. He's still got folks. It's going to be one episode amongst many new episodes that we're going to basically bring in folks that are at the forefront of AI and AI revolution, and any, on basically every front. I'm actually trying to bring a data scientist to basically help us, educate us about what this data behind this AI works. And until the next episode, keep it cool, chill. Thank you for listening to us for AI. Join us next week for more insightful conversations about the impact of artificial intelligence in development, design, and user experience.