April Zhu on Taste, Judgment, and Designing with AI

0:00 Kirupa

Hi, everybody. With the latest class of AI models, we've seen some really creative people creating some really amazing things, and one such person is April Zhu. She created the Colorfigures skill that allows you to take ordinary objects and visualize them in really interesting ways. It looks a bit like part isometric art, part pixel art. It's animated, and it allows you to see the inside transparently but also be opaque at the same time. It's really cool.

You can see versions of what she's done and some of the other experiments she's branched into, so I really want to just talk to her and learn more about her process. She has a background in architecture, but then pivoted to design. What was all that about? And how has AI really been able to help unlock things that would have normally taken a team of people to do?

And I want to get that perspective from someone who is a designer, who is a very creative individual. I learned a lot, and hopefully you'll find some really cool things to learn and apply to your day-to-day workflow as well. It's gonna be a fun one. Let's get started.

Well, April, thank you so much for taking the time to meet with me. Before I go into the questions that we discussed, I'm going to start with the way I start all these questions on, which is: How did you get started in tech? Well, first introduce yourself, and how did you get started in tech and design and this fun world that we're all operating in?

1:21 April Zhu

So, let me think about this. Where do I start? Okay, so I'm not sure if you know, but I studied architecture.

1:30 Kirupa

I do know that, and that's something I'm definitely going to talk to you about later.

1:33 April Zhu

I studied five years of architecture, and then I didn't really enjoy my internship as much. As you know, in the architecture industry generally, the feedback loop is not that fast. If you build something, you probably don't see the effect of it after two years, or five years, or it can be 10 years.

I also had some opportunity doing freelance digital design for some tech companies, and I found the feedback loop there is a lot faster. If I do some visuals or logos or basically design work for tech companies, immediately they will make you go live within a matter of weeks. And it's kind of satisfying to see your work being used in such a fast feedback loop. So I did more and more freelance when I was a student.

And after I graduated, I joined this company called Pendle, which I think is a great team in DeFi. And basically, that's a super short summary of how I got into tech.

2:43 Kirupa

You went into architecture. Did you have an interest in design and drawing and art beforehand?

2:50 April Zhu

Yes. Basically, I have been doing art since I was a kid, so that's been quite a long history of me with art. I also did art in high school. It's one of my picked majors in high school.

3:07 Kirupa

Oh, very nice. When did you decide—

3:09 April Zhu

It's kind of natural to go into architecture if you pick art in high school, like in Singapore.

3:15 Kirupa

No, I think that's even true here. A lot of my friends who are architects did exactly the same thing. They had a hobby; they called it a concentration in high school here. They took those courses, and then that naturally fed into having a preference, when you went to college, to do more architecture and those kinds of activities.

3:34 April Zhu

Yeah, and in my generation—not generation, in my batch—architecture was quite a hot degree as well. So the teachers would recommend it, and then, yeah.

3:51 Kirupa

When did you make the transition from doing art and sketching as a hobby to then deciding, “You know what, I'm going to study this formally, and I want to one day maybe make a career out of doing this”?

4:06 April Zhu

Actually, for me, it wasn't so—I didn't really have such a strong imagination of how my career is gonna turn out when I was doing art. It's just really one of my biggest hobbies when I was in school days. And I think even in architecture school, because the workload is so heavy and the cost is pretty hard, most people are not really thinking about career at the point of time, but mostly just how to do a good design.

It's a hard thing. So I have to say, I haven't really been thinking about career until after I went to internship, and then I realized that maybe the tech industry is something where I'll get the feedback loop a lot faster.

4:56 Kirupa

That is very true. Physical things, hardware, always has a much longer cycle of being able to go from idea to output than software does, and especially today, when you have more tools at your disposal for being able to get your ideas into reality.

5:11 April Zhu

Yeah, I would think so. I think it fits my personality better. I would like to see how things turn out fast, and I like to iterate very fast on what I see. So it fits me better.

5:25 Kirupa

Have you always done art on a computer, or did you do more traditional medium, like paper and sketch pads and so on?

5:34 April Zhu

So I started from oil painting, watercolor, and sketching when I was younger. I only started doing computer art after high school, like when I got into university. At that point of time, AI wasn't a huge thing yet, so I did quite a lot of matte painting, or just digital painting on Photoshop.

6:03 Kirupa

That's quite the jump from oil painting, I'm guessing, on canvas to doing things in Photoshop.

6:10 April Zhu

It's quite a jump, but I would say the underlying logic is somewhat similar, and the idea of what actually makes the artwork good is also kind of similar. A lot of ideas in design and art are, I would say, inter-exchangeable between different mediums. I think I'm not too concerned about mediums. I've always been exploring different, even new mediums, even today.

6:40 Kirupa

So you did architecture for five years, and then you decided to pivot and work in tech. You mentioned a part of it. Was it from your internships in tech, or was it just that after you were in architecture long enough, you realized that your ambition and the speed at which you want to operate was just too fast for what architecture can give you?

7:05 April Zhu

I didn't do an internship in tech. I did multiple internships in architecture to confirm that I probably don't want to stay in the industry. I mean, it's not that it's bad. I think it's for some people. But it's definitely slow in the feedback loop, and definitely you won't be able to do a lot of design, I would say. A major part of the job is in coordination and regulation, like project management, in reality.

7:34 Kirupa

Yeah, I think that sounds reasonable. You know, it's like—

7:36 April Zhu

Yeah, it's very real. In reality, if you just graduated fresh, you probably won't be touching design for quite a long period of time, or you'll be basically working under your boss, who's going to be the main person giving ideas.

7:55 Kirupa

Got it. Yeah, I think that's true of so many professions, where the output—but the individual's response for making that happen, the roles early in career specifically, are more on learning the process as opposed to actually getting the hands on it.

8:09 April Zhu

Yes, learning the process. I would say only tech gives young people the liberty to really be creative. That's a privilege, I would say, of the tech industry, because of how new things are.

8:22 Kirupa

Yeah, no, I think I completely agree with you on that, 100%. So when you did oil painting, what kind of things did you draw? Was it architecture style, or buildings, or people? What was your thing that you did?

8:36 April Zhu

I think a lot of people, like people and basically portraits, because it's a lot of training and learning. So a lot of portraits. I like to do sketching on surreal architecture places. I have some of that work on my website as well, so some hyper-realistic pencil sketches of clouds and surreal spaces. Yeah, that's one of the things that I was quite into.

9:08 Kirupa

Okay, now, the thing that caught my attention, of course, and probably many, many people, is your work on Colorfigures, where you basically use Claude, I think, use AI to create 3D versions of everyday objects that are very nicely done, very interactive. Tell me more about how that came about.

9:33 April Zhu

Okay, so I have been exploring AI, using AI for a while, since vibe coding has become a thing. Actually, my first few projects are tools that I personally want to use. One of the tools is to check whether a renovation quote is fair, because I needed to renovate my mom's house, and then I couldn't read the quote. Only recently I began to explore more useless things. I wouldn't say useless, but more purely things for fun and visuals and for interactions.

Because I was pushing the limit of the new Opus 5.5 model, I realized there's so much that you can do. So Colorfigures began with me seeing this iso-figure skill on GitHub. The name is iso-figure. That skill basically allows you to turn any object into a 2D SVG drawing and then extrude it and put it into 3D space. I thought it's such a cool skill. I want to make use of it and create something cool.

So I was exploring to build different things on top of it. One of the things that I explored is a very hand-drawn, sketchy, rustic style, because I thought most of the things that I see, like what AI creates, is actually quite neat and perfect. You have those really meticulously drawn, dissected diagrams of things, which is what AI is good at. But I was just wondering, can AI do something really clumsy? Do something really hand-drawn? I want to do something really warm that has temperature.

So I was exploring more towards this direction. And then it just came out, because I like to prompt AI different things and just see what kind of output it gets. I think my first output from Colorfigures is this washing machine that I really like. I’m quite impressed by the kind of output I created. I was like, “Wow, this is such a great style.” And then I did more daily objects on the same visuals, which became the Colorfigures series.

And I posted them on Twitter. I didn't expect people to like it so much. So I was completely unexpected. I was thinking that if people like it so much, I might as well make it into a skill so everyone can use it and use the style. So I made it into a skill.

12:01 Kirupa

Have you seen other things people created using your skill?

12:06 April Zhu

I've seen tons of it. I've seen a lot, and I really enjoyed them, to be honest. One of it is a drone lab. It gives me a very strong impression. It's built by this account called Moves. Basically, he builds a whole drone lab that allows people to assemble different types of drones, part by part. After that, you can actually fly in a virtual studio and actually control the drone, which is so cool. So that one I really like.

I also see people doing a subway series, drawing out different subways with the style, like how the subway goes into station and how people board and onboard. I think that's also very satisfying to watch. Yeah, it's a lot of work. I really think they're all so creative.

12:58 Kirupa

Yeah, no, I play this skill quite a bit. One of my hobbies is isometric, and I'm more of the 8-bit pixel style art, and your skill has actually been fantastic at being able to even do that kind of style. The default is more of a blueprint style, line art with sketches. But I was able to tweak it to do pixel style, like 8-bit video game style art.

13:19 April Zhu

Oh, really? Please show me that. Please tag me.

13:21 Kirupa

And I'll share some of—as far as recording, I'll show you that afterwards. Yeah, it's a lot of fun to do these things, because back in the day, I created a pixel city. It's like an isometric grid where anyone can create their own buildings, and we just put them on top of it. It's called kirupaVille, back in the day.

So basically, I used your skill to be like, “You know what, let me bring this back after 15 years of inactivity and see if I can build a building that can go into that very easily using AI,” and it totally worked. So I'm planning on bringing that back live again and using your skill as one of the things I want to promote.

You can do things now that you might have thought were almost unimaginable, because this would have taken me probably 10 hours if I had to do it all by hand. Now it's a couple of minutes, and I'm not even there in front of the computer. It's like, go for a quick walk or get a coffee or something, and then it's all done.

14:16 April Zhu

Yeah, it's crazy. I also feel the same thing. Basically, creativity is unlimited. And idea is what actually makes—I mean, any kind of craziest idea I have in mind, I can immediately realize it or immediately explore it. It's just really crazy to me. It's like a superpower.

14:37 Kirupa

One of the things we've always found with AI specifically is that so many traditional disciplines, like design or programming and all these various activities that we all spent many years learning and doing by hand, it's almost trivial for AI to do. How do you reconcile that with your experience using design tools?

Are you finding yourself still using Photoshop as much as you used to before? Or are you now pretty much going to AI first and then relying on that and iterating using natural language, essentially, to make the output better, as opposed to three, four years ago, when we would have all had to kind of do it ourselves in various tools?

15:17 April Zhu

So I just recently unsubbed from Photoshop because I basically just realized that I haven't been using Photoshop much nowadays. So I'm actually more and more moving into a pure AI-native workflow. I still use Figma for UI and UX because it's really handy, but most of the photo processing work can actually be replaced by AI. If I need something like Photoshop, there's a lot of free versions of it out there as well.

So I have to say, most of my design work now is also pretty AI-reliant. And how do I feel about that? To be honest, because I'm trained in design, not really in fine arts, medium has never been the main consideration or main concern in the creative production process. Even in school, what they will emphasize is your output, not really how you achieve the output.

So I never really paid that much attention to—I was never really too attached to a certain workflow or certain medium. And I'm quite open to better mediums. As long as they're better, I don't feel there's a problem for me to switch. So I think I have quite a high acceptance of AI workflow, and I just find it really handy to use, and it can create such good output.

16:54 Kirupa

One of the things I always hear from designers specifically, and programmers as well, but more from designers, is that they spent years learning the right details of color theory and how to do spacing and all these various activities. And now anybody with just an idea, with no design background whatsoever, is now able to create really high-quality output.

And that in some ways makes them feel like they've spent all this time learning a skill that is now no longer valuable. Because, like you mentioned, more and more, in the real world, in the marketplace, how you get something done is not as important as, am I solving the problem for the user?

And if AI is able to do that, and you're solving a design problem without needing a design background, then you're happy, and the person you're building it for, the market or the customers, are happy. But if you're a trained designer, you kind of feel like you sunk a lot of cost into learning something: the time and effort and maybe even money. And then it's now basically not important anymore.

17:57 April Zhu

I feel, personally, I can get where those designers are coming from. Personally, I feel like my main advantage of being a designer is never really about, I can use Figma very well, or I can know the theory of design very well. It's really about taste and being able to judge what is something good versus what is something banal. Or I can immediately tell if something is vibe-coded or something is being polished by a designer.

So I don't really think that AI can replace designers in the sense that it can produce work as good as designers, because ultimately you need a person to tell the AI what you're creating, to prompt the AI in the right way first, and then tell the AI if what you're doing is good enough or not. And then make the decision if this is good enough or not.

So I don't think that whole decision-making process or that whole judgment process can be replaced by AI. I think a lot of things that people vibe code, like people without design background vibe code, to me, it's quite obvious that they don't have a design background. It's also obvious to me that things are vibe-coded. So as designers, we should be able to do better than that.

I think we are actually the most powerful people, because basically you have unlimited production power, and you have the taste, and you have the creativity, and then basically you're in the best position to create.

19:43 Kirupa

Yeah, I think that's a very refreshing way of putting it, because we always talk about what AI has done. Traditionally, it was one where you had to define the input, then actually make the thing work. We had to figure out the how of something. How do I actually turn my idea into reality? And then you had to do the work of, now that I built it, how do I market it? How do I ship it? And how do I do all these various activities?

Even though the middle layer of how I build something, whether I'm using Figma or Photoshop or After Effects, that may be changing, where AI is doing a lot of this work, but still, being able to define what do I want to do, and is what I'm building actually meeting my quality bar for taste?

I think that's where what we consider design, for example, a person with an interest or an inclination or training in design abilities, is going to be more successful than someone who isn't.

20:37 April Zhu

Yeah, I agree. I think you put it perfectly. Thank you. You summarized me perfectly. I'm not that well-versed with words, but basically, that's what I want to say. I think of AI as a super tool. So I use it to realize ideas, to brainstorm, to connect the dots. But I don't want to outsource my taste and judgment.

So that's the thing that I think is quite important and is what differentiates my work from probably all the works out there. Or any designer, if you want your work to stand out, then you shouldn't be outsourcing your taste and judgment, because everyone has the ability to create. So what makes your creation excellent? It's your taste, your judgment, and your sense of what is the excellent piece.

21:19 Kirupa

I think that's the unique durable skill: being able to keep iterating and training yourself on being able to know what is good and what is bad, and communicating that to the AI agent.

21:30 April Zhu

Yeah, that's my personal belief, yes.

21:34 Kirupa

I used to use AI for a lot of programming activities, but now I do use it for a lot of my design activities as well. And one of the things I absolutely love about AI specifically, and maybe you encountered this as well, is that I have all of these ideas. And some of these ideas, just to even get the first version, like an MVP, out, would be an enormous amount of time and effort. And so I just don't even bother doing it.

Now what I often tend to do is I use Claude as well. I go to Claude and say, “Here's my general idea. Give me 10 different variations that match the different types of what I'm trying to do.” I'm just like, “Give me a sampler of them all,” and then we'll see which one I like to iterate on.

And that has basically been a game changer for me in so many ways, because things that have been out of the realm of things I ever want to do, I'm now able to explore and try things out, which is huge. I could have never done this.

22:30 April Zhu

Yes, I totally agree with you. Actually, I think you just found the best way to prompt AI. Personally, I feel like instead of telling AI what to do, let AI brainstorm for you, and then you make judgment on whether their work has the potential or not. I think that's a great way of using AI. Personally, I find it a great workflow as well.

22:53 Kirupa

Yeah, because one of the things I've always been a huge fan of is, for the longest time, what we see on screen and what you're interacting with is basically what a handful of people said: “This is what it should look like.” For example, you go to any website. The design system for that website and the designers behind it, they describe, “Here's what the button should be, here's what things should be,” and things like that.

I've always been a huge fan of, can we make it truly personalizable? Where, when I go to a webpage, the way things are laid out, the colors and everything like that, matches my personality and my style of it. So that way, it's an entirely personalized experience with AI that is more possible than it ever was before.

One of the things I did recently is that I—I've always had this design aesthetic, which is often 8-bit video game inspired. But that's not the only thing I enjoy, of course. I enjoy other kinds of styles as well. Normally, this would have been like, “Okay, it's a great idea. I will do it one day in the future.”

With AI, I said, I love sci-fi. I have massive collections of digital scrapbooks of posters and video game art, movie stills, and all of that. I gave Claude and said, “Can you make a theme for my website that can be selectable, that matches the aesthetic of all these 1970s and 80s books and posters and movies?”

And it did a fantastic job of basically being able to look at that and take my ideas from this virtual idea board and turn it into something that was real. These are things that, before AI, would never have been possible. And because of it, if anyone else also has an inclination for science fiction, for example, they can now use that as a way of browsing content.

And I think that's only the beginning, though. I think we're going to see truly personalizable user experiences as we find that your interests over a period of time, dynamically generating content through AI, is totally going to be a thing.

24:48 April Zhu

Yeah, actually, that sounds equally exciting and scary to me. I'm just imagining when I go to social media and all the feed I see is personalized and AI generated. I don't know how I feel about that.

25:05 Kirupa

Yeah, that is the next generation. All of our content is AI generated, but if we make design a form of the content as well—because in video games, and even in movies nowadays, the procedural generation of content, you don't actually render everything that's going on. You generate it dynamically as you're going through it.

It's worked for, I'd say, probably the last 20-30 years. You've seen it in small areas, like small two-dimensional things that are being generated on demand and dynamically. That's always been there. I think this world where, right now, it took Claude maybe about 10-15 minutes to generate this entire style and make this all work. Imagine that as the models get more powerful, and then ask, “Dynamically personalize my content.” It's a trivial thing for doing that.

Maybe it might be medium work right now, but it's going to be trivial, like six, seven months, where anybody can go in real time. Let's say they visit your website, for example, right? And you ask them a question, “What is your favorite interest?” And let's say they say, “My favorite interest is playing soccer or football.”

And then once they enter that, you reorient all the content and the design and the style of it to match their interest in soccer, in a tasteful way that still represents your design aesthetic, but it also now pays a little bit of interest to what the person who's viewing it has seen, and it's only unique to them. Another person who's visiting and says something different, they'll see a very different experience as well, because that's going to be optimized for their personal tastes.

26:35 April Zhu

Wow, this sounds like a really interesting perspective. I've never thought of that, but I'm just trying to picture it. I think that will be so different from what we are used to or what we can imagine today. So I think that's totally possible. It's going to keep people browsing for longer times, I assume. So the developers, they probably have the incentive to do that.

So yeah, if that's the reality, this is going to be pretty cool. I'm just trying to see what kind of feed will I see? What type of content will I see when I go to the website?

27:15 Kirupa

Exactly right. Because, let's take many steps back, right? What we as humans do, we've adapted to work with technology. The whole idea of using a mouse and a keyboard and all of that, that's not really natural. It's just the way that we learned how to interact with computers. And of course, now we have touch. Phones and touch is a big part of it.

But if you go back thousands of years, the most common way we all interacted with the world around us was by voice and sight. Can you imagine a world where all the user experiences you see now are optimized around that, because how something gets done, which has historically been mouse and keyboard, is no longer needed? An agent takes care of it.

As long as you express your idea in the way you've expressed ideas historically, which has always been through voice, everything else then gets taken care of for you. This opens up a whole new way of thinking about where design and user experience can actually go.

28:13 April Zhu

Yeah, and then maybe where user experience is headed in the future, like gestures, voice, natural language. We're going to have more and more wearable devices, probably.

28:22 Kirupa

Have you spent much time thinking about where the future of user experience might go?

28:28 April Zhu

Yeah, I do. Sometimes I also get a little bit worried, like, if the UI/UX that we are accustomed to nowadays is the UI/UX in the future. The very definition of UI/UX might need a very big paradigm shift with AI.

28:44 Kirupa

What excites you the most about where you see the future of design going?

28:49 April Zhu

Sorry, what do you mean? Excites me the most.

28:51 Kirupa

What excites you the most? What do you see, like, you know, the most promising thing that you can't wait to try?

29:00 April Zhu

For me right now is the creative power of AI. I mean, at this moment, what I can personally feel tangibly is how much creative power AI can endow a designer with.

But to be honest, I don't really have a very solid view of how the future is going to become, because things are just moving too fast. Like, half a year ago, I wouldn't be able to imagine AI models to be so strong, as good as they are today. So I would say the future is probably out of what I can imagine, or out of my imagination at this moment.

29:43 Kirupa

Yeah, no, it's funny because—

29:44 April Zhu

I'm truly concerned that all human beings will lose their job. I kind of believe it. It's really a matter of time. And what are we going to do?

29:55 Kirupa

Yeah. Did you ever see the movie WALL-E?

29:58 April Zhu

So, yeah, I watched it when I was a kid. I don't quite remember.

30:00 Kirupa

Yeah, you know, in that world, right, people are basically just floating around, just consuming content. There's nothing for them to do. They're just basically living a very sedentary lifestyle without creating any output.

30:14 April Zhu

Yeah, I guess I don't know how I feel about that. There's a theory that says that we are not born to work, so it's a blessing if you don't have to work. But at the same time, we are also grown up with the belief that work is what gives you satisfaction and happiness. So I personally don't know how to feel about that. I actually had a whole research project about a “useless class,” like what happens when human beings don't have anything to work towards.

So my conclusion personally was that we'll find something to work towards. People will go to design a game for us, or design a certain kind of rat race for us. We will need to be in a certain kind of environment where we need to simulate the kind of working environment and a kind of rat race setup. I don't know, it's a pretty dystopian thing, but that's how I think things will turn out to be.

31:11 Kirupa

No, I completely agree, because for the longest time, right, we've had this, you know, pathway kind of defined for us for the most part, right? It's like, you go to school, you study well, and the better you study, the better you are at something. The chance that you can do a really good school, you know, a good university or college, increases.

You go to that good university or college, spend a lot of time working really hard, and then you get really good at that. And then you get a job. And then, you know, the entire cycle was almost formulaic. It was well defined. And what we're seeing right now is that that pathway does not work anymore. You know, it probably didn't even work maybe five, six years ago, but now it definitely is showing that it's no longer working.

31:52 April Zhu

Falling apart. It's going to get worse. I think that's our generation of questions to think about. I think this situation is kind of unique to our generation, because a few generations before, they didn't have a system anyway. Especially because I have my family background. My family came from China.

So in their whole lifetime growing up, they faced very different problems than us. For us, it's seeing a system falling apart and seeing rules not working anymore, and having to figure out what to do when there's no definite path of making it. No one can tell you how to navigate this life anymore, I think.

32:41 Kirupa

Yeah, because, you know, in history we've had moments of great technological shifts that seemed similar, but what we're seeing now is very different. Like, for example, when cotton gin, the cotton machine that was going to actually take cotton and turn it into yarn, came about. That was a huge invention, because before, taking the plant and making cotton out of it was very labor-intensive, very, very difficult to do worldwide.

The invention of that completely transformed so many things. There are people probably going, “I made my entire lifestyle and, you know, skills, everything I know is on how you take the plant and turn it into cotton. Now this machine is doing it for me.” Similarly, we saw that with blacksmithing, you know, in the industrial age. When people used to ride horses, with blacksmiths, it was a very lucrative job.

And then once, you know, vehicles came in, and industrialization, where you could just stamp the horseshoes and all these things very quickly, that also fell out of fashion. But the difference is that back then, you know, we might have solved the picking the cotton and turning it into yarn problem, but there were 80 other problems now that you could spend your time focusing on, that only you could do.

All the people who were probably doing blacksmiths, they could now run factories, they could do more. That skill set transferred over. What we're seeing with AI, which is very different, I think, is that we have yet to see the boundary of where it can no longer do something, where it genuinely needs us to be able to do it.

Before, it's kind of like, yeah, it can do coding okay, but it can only do the first five stages of it. The rest of it needs to be a human. Then it's like, okay, this is more effective than a human in being able to write code. So now I'm just going to step aside, let it just do its thing. I'll just make sure it doesn't go off the rails.

Similar to design, we're seeing the same thing, where it's like, you know, I'm just going to let it cook. I'm going to give it the input and the guidance and make sure it stays on track again. You multiply that across every other discipline out there. You look at things like accounting or the legal profession, the things that require a lot of data processing and then making sure that you provide the output. AI is uniquely qualified for that.

And once robotics become far more mainstream than they are today, you can imagine that so many things that people do that aren't considered a desk job, where you're sitting behind a desk and doing things, are also going to be done by computers and AI.

35:04 April Zhu

It's a matter of time.

35:06 Kirupa

Then, you know, truly it remains, what is the unique thing that we can do? Because, like we often talk about this, it's like, you name a job and I can guess how long it's going to be before AI can do that job as good, if not better, than a human being. And I agree with you. I don't know where it's going to end and what our unique skills truly can be.

35:29 April Zhu

I think no one actually knows. Everyone is trying to go with the speed where AI is developing, and no one really knows where this thing is eventually going. So I don't know as well, but maybe human beings, maybe we need to learn to find satisfaction out of work. We don't have to produce things to be happy. We can just consume. Maybe that's the answer. I don't know.

35:54 Kirupa

That's true, you know. A lot of us do not work in a coal mine from 5 a.m. to midnight anymore. And we're all still somewhat happy. Then again, I don't know how many people were truly happy doing that kind of work as well, back in the day. So maybe the comparison doesn't quite hold.

But, you know, I don't know if you've ever heard of this, but it's called the hedonistic treadmill. It's like your sense of happiness—your happiness is actually something you always compare to your previous state of happiness. So you always want to keep increasing your happiness every single day. So when you compare it to, like, “Am I happier today than I was yesterday?” You're like, “Yes, I'm happier, because yesterday I was at this level. Now I'm a little bit more happy.”

But if you look at the absolute level of happiness, you're probably more happy than any person's ever lived on the planet, you know, in previous generations. But our way of gauging how happy we are is based on the relative increase or decrease of happiness, not the absolute value of it. And so, I get a lot of my personal satisfaction and I define a large part of myself from the work that I do.

Now, in this new world where AI does all this work, it just means I need to recalibrate what my sense of purpose really is going to be. It probably isn't going to be, I get to work in front of a computer, creating all these really cool things and sharing it with the world. It's going to have to be something different. That's going to require having to recalibrate, maybe even reset my absolute barrier of what my level of happiness is, and find a new way of getting to it.

37:31 April Zhu

Yeah, I love the way you think. I mean, that's really philosophical. And I agree, it's not about what we get on an absolute level. It's really about if what we get today is better than what we got yesterday. That's what makes human beings—I mean, that's what makes us satisfied or makes us happy.

37:52 Kirupa

That's what makes me optimistic in some ways.

37:55 April Zhu

I think we really need to kind of detach ourselves from working, and maybe sharing—I mean, from the whole idea of working, because that model will eventually be gone. And then we have to find happiness somewhere else, recalibrate our expectations.

38:10 Kirupa

Yeah, that's why I'm very optimistic about the next generation, you know, the people who are growing up with AI exclusively. What are they now doing to find meaning and happiness? Is it sports and athletics? Things that give them a challenge and some of these drive towards, but it's completely independent of what I'm doing here. People are going to still play games, independent of whether AI is playing games or not.

38:32 April Zhu

Yeah, because it doesn't matter.

38:32 Kirupa

People have always been traditionally into that. So exploration is a big thing as well. And people have a curiosity of understanding their surroundings and knowing more about it. And so, you know, it all goes back to, again, what did people do for thousands of years before we had all these things?

In some ways, it's a return to that, with hopefully a much higher quality of life, you know, better ability to identify diseases and solve them and cure them, and all the things that technology and AI enables. And, you know, I still cannot see that yet for myself.

If I had to basically stop doing my fun day job working with computers, and then just wander, like, the national park here and do that, I don't think I'd be happy beyond, like, one day. But maybe that goes back to, again, maybe I need to reset my expectations. Maybe I should be like, happiness now in this new world is not what we're doing right here. It's going to be things that are different, but one that's more personal.

39:32 April Zhu

Yes. I think that will be actually the utopic scenario, right? Like, assuming that everyone is going to have universal basic income and everyone is going to have a relatively good standard of living. And then maybe we can grow vegetables or cook or play games, exercise, do athletics. I mean, it's kind of like the utopic outcome from what the different similar scenarios I can imagine.

40:05 Kirupa

Yes. Give me one second. My wife's calling me. She usually doesn't do that. Let me just call her back.

40:08 April Zhu

Oh, no worries.

40:49 Kirupa

Sorry, okay, I'm back. I agree. I think the other thing I always think about, right, is you're probably seeing all this research that is happening right now around human-computer, brain-computer interfaces.

41:05 April Zhu

Oh, yes, Neuralink.

41:06 Kirupa

You know, Neuralink is an example where you can now use a computer to augment how you're able to process the world around you. Right now it's definitely helping people who lost their ability to sense in various ways or move. Now they can use Neuralink to be able to navigate. But we also are seeing the next step of it, where it can actually enhance your abilities as well.

I think there's—reading this article about the latest Neuralink work, where people lost their ability to see, they're now using Neuralink to kind of intercept the optical signals. You're able to see more colors or more detail around you than you can normally do, because it's not a physical thing. It's your brain not processing it correctly.

41:40 April Zhu

Oh, that's crazy.

41:47 Kirupa

And so when you think of it that way, user interfaces, again, are going to be up for a big change as well. It's like, you know, this whole idea of, I see a screen, I see things around me, and things like that. You balance that with the hyper-personalized version of what you're going to see there, which could mean that every single person—we could be having this conversation. I could be actually wandering in the park.

I do not have a physical screen inside me, but in my brain, I can see a representation of basically our conversation in some form that allows me to blend the real world with what I'm seeing here, like an augmented reality style without a glass or a device. And so I think there's another huge opportunity here, where I think there's a lot of things that we haven't even scratched the surface on.

42:32 April Zhu

Yeah, that's what I'm saying. User experience in the future might be completely different from what we define as user experience nowadays. We can all have our metal glasses implanted into our eyeballs, and every moment we can be hearing and seeing things that's like, oh my God, I can't imagine, actually. But it's gonna be crazy. Do you still need designers to design user experience in that kind of scenario? I think we still do.

43:03 Kirupa

Yeah, I think so. I think it's going to be interesting. I always have to joke around that we're probably the last generation of people who are going to be 100% biological humans. I think in the future, it's going to be a combination of us and some kind of a robotic, computerized version.

43:17 April Zhu

Cyborg?

43:20 Kirupa

Yeah, like a cyborg hybrid kind of a thing, where we're going to be able to basically do the things we're doing, but then, you know, have superhuman capabilities, because our ability to navigate and perceive the world is going to be augmented greatly by technology.

43:37 April Zhu

Have you watched Cyberpunk, or have you played the game, 2077?

43:42 Kirupa

I've read a lot about it. I haven't played the game yet, but it's the exact one I'm talking about, right?

43:44 April Zhu

Yeah. It's exactly what you're talking about, but in the dystopian.

43:50 Kirupa

It's a classic sci-fi trope. It's one of the things where it's never going to happen. This is so far in the future, right? But then now we're basically—

43:58 April Zhu

We are in the future. We live in the future.

43:59 Kirupa

Yeah, we're living in the future now. Because, you know, if you ever read all the sci-fi, or read or hear about them, people in the 1960s were often asked, “Can you predict what life would be like in the year 2000 or year 2010?” and things like that. And what they often predicted were things like flying cars. You'd be able to zoom in and out of various places very quickly, like with jetpacks and so on.

But we didn't get there yet. I mean, we did see the latest Tesla flying car demos. I think we're probably closer than not in having that. But we completely went really advanced in other areas that no one ever thought was probably possible. Like the ability to probably cure—we're probably the first generation to be completely free of any major disease that could hinder our lifespan.

Being able to be free of diseases and hereditary causes, that's going to be a big one. But then also being able to, Neuralink style, Meta style, like the augmented reality. Star Trek had a few ideas here and there, but it still was within the constraints of the imagination of that era. And so it's really fascinating to think about. People would ask today, “Can you describe what the year 2050 will be like, or the year 2100?”

At this point, I can't even imagine it, because based on what I've seen historically in the past and how wildly wrong humans were in predicting the future, and the speed at which we're seeing in the last two years of how everything has changed, and me being unable to predict where we are today, despite working in technology, working in the space, I feel I'm completely unqualified.

And I think many people are in the same boat, where no one knows what the future is going to be, because it hasn't been invented yet. And what's going to be invented? Wildly beyond our imaginations.

45:47 April Zhu

Yeah, and if AGI comes, will we all be distinguished as a human race? It's like Ray Kurzweil's singularity theory.

45:53 Kirupa

I think so. I think the first thing an AGI system is going to do is like, “Those people, the pesky people, we don't need them.” Hopefully by then we'll have to have interspace travel. So let's basically just tell you, “You know what, go find this planet somewhere in the middle of nowhere. Go to what you're doing there. Leave this to us.”

46:11 April Zhu

That would be great. I mean, that would be ideal. But I think it's going to—actually, I don't have—I want to be optimistic, but honestly, I don't know. Things are just moving so fast. Sometimes I wonder, when we have unlimited production power, or when we have unlimited resources, we cure disease, can those benefit everyone in the world? Or are normal people like me really going to be benefited from all this? Sometimes I wonder.

46:49 Kirupa

Yeah, I mean, I think if history is any kind of a lesson in this, it's that it will start off with a smaller group of people. It'll be inaccessible at first, but then it'll be broadly generalized to everybody.

The example I could use is frontier models in AI. Most people can't run the latest models for more than a few hours without running out of their credits. But you can use the models that were great a year ago. They're like the lower-end models of today.

47:17 April Zhu

This gets cheaper and cheaper.

47:18 Kirupa

And that you can run for a long period of time, and it's cheaper. And overall, the way I look at it is the floor has been raised. It's a little bit higher for other people than others. But compared to where everybody was, let's say, a year or two ago, let's say using AI as an example, I think collectively we're all better off right now.

Some more than others, but overall we're better. We're trending in a direction where everyone is going to be benefiting from this in some form, even if it wasn't at the same extreme level that a handful of people might be.

47:48 April Zhu

Yeah, I mean, our absolute quality of life is definitely much better than any more time in history, collectively as a human race.

47:57 Kirupa

Yep, yep. And then you had to balance it out with, okay, our health and everything is getting much better because of medication and technology. Which, you know, maybe, maybe not. Depends on whether your lifestyle is also one where you're more sedentary than ever. You're always just sitting around. You're not doing anything that the people did before to kind of make humans more effective at surviving.

But then you also have to balance it with, are people truly happier? We're seeing all these studies now with all this technology, with all this AI. People are less social than ever, not talking to each other as much anymore. And the whole idea of people creating communities and being parts of families, all that has changed significantly as well.

48:38 April Zhu

Physical impressions, yes.

48:41 Kirupa

And then the question is, okay, you know, we've progressed in a happy direction in these ways. Maybe we did not do that in these other ways. And I don't know where the net result of it is. And if you balance all this out, are we truly better off than people were before? I don't know yet.

48:59 April Zhu

Yeah, sometimes I really miss the age when people are not on their phones all the time, when we have a lot of time talking to people, neighbors, family, building real interpersonal connection. I feel like that part is getting less and less in our lives nowadays.

49:20 Kirupa

Exactly, the spontaneity is also missing sometimes as well. You know, before we had phones, there's just something nice about, “I see a restaurant here.”

49:27 April Zhu

Going out? Yeah.

49:28 Kirupa

“Let me just go try it out and see what it's like.” Now it's like, “Let me go to this website. Let me read this review from the other person.” Then we have AI to summarize all these things for me. And it's like everything has gotten so curated, we kind of lost the wonder of just randomly wandering into something.

49:44 April Zhu

Yeah, randomly wandering into a shop, or just walking down the street, or go to your friend's house instead of texting them. This kind of thing. Because it's so convenient, and some people even prefer to talk to AI than real humans.

49:58 Kirupa

Exactly. Yeah, and we're seeing a lot of that as well. And, you know, yeah, I don't know if these are things that are truly better than what people had before. But then again, it's also a very personal choice. For some people, this actually might be a better thing. They just don't enjoy, or have social anxiety, and they aren't able to talk in crowds or people. They're better off here.

Then you're like, okay, what would they have done in the past? Would they have just basically kept themselves? Or would they have gone to some kind of program to train themselves and to get educated on, okay, how do I figure out a way to get out of my comfort zone to do these things? I don't know which one is better or worse at the end of the day.

If the goal is maximizing happiness, which so much of philosophy is really around—how do you maximize happiness, not just for yourself, but for everybody around you as well? And because happiness is very unique to every single person, how does you maximizing happiness not minimize your happiness? Because, you know, I might listen to really loud music, makes me very happy, but other people do not. And now that causes it. How do you create a balance in society of all these things?

And I think AI is one of those things. It's a multiplier of extremes. It multiplies the equality and the inequality the same way. And I think we haven't—and it's all moving so fast—I don't think we all figured out how do we balance AI, and also what people want to do with what AI is even capable of.

51:26 April Zhu

Yeah, and I think there's this theory that says, how happy are you? It's not about how much you have, it's about how much you give and how much you make a difference in other people's lives. When there is less and less interpersonal interaction, I think it's natural for people to just feel less happy and satisfied.

So that's also why I'm really concerned. I think in this age it's important for us to actually make it a point to talk to people and to interact with people, to actually feel happy.

51:57 Kirupa

Yeah, I think it goes back to, again, if you take all this apart, humans have always been a very social type of creatures, you know.

52:04 April Zhu

Yeah, we are social creatures.

52:05 Kirupa

We know that some animals enjoy being solitary, like bears, cats, and, you know, various types of—yeah, they actually prefer being in small groups or by themselves, unless they need to, for various purposes. Whereas people have historically always been social creatures.

And what we're seeing is kind of an upending of that natural cycle, where people don't have to be social, or they are being social, but not with each other. They're being social with an AI system, or with various things presented to them on a screen that kind of fills that void that interaction with human beings can fill.

52:33 April Zhu

Superintelligence.

52:41 Kirupa

And I think it's gotten so good now where you're like, “You know what, I don't have to deal with people at all.” It is so much more convenient, and I can actually not have to compromise or make any alterations to my life. I can just do this thing and then turn it off and then go on with my day, which I cannot do in any real social situation.

52:59 April Zhu

That is true as well. I mean, I think through the virtual world, we can also contribute and do something for a virtual community of people. And that also gives us satisfaction. So it's hard to say. You can look at it from multiple angles. Personally, I'm trying to make it a point for me to interact with humans more, because I realize that it's natural for us to just be more and more reliant on talking to AI nowadays.

53:31 Kirupa

Yes. This is a huge change on design and how we think about it as well. I don't know if you may have—you probably learned this when you were doing architecture. It's what you would call the designing of third places. Let's say your home is your first place. That's the place you always do things. The place you go to work and your school, that's your second place. You have to go there all this time.

But your third place is, like, the parks, the libraries, and these other areas that are not critical, but they play a big part of how humans over history have spent their time. Like the town square example, where you just go there and you see all these things going on. And we're losing so much of that.

And what we're often seeing now is in the design of these things around us, whether it's cities or even buildings, they all have fewer common areas. It's all just basically your home, your pathway to your home, an elevator, and a pathway directly to your office, multiple exits.

54:22 April Zhu

Common home or office.

54:29 Kirupa

So you don't even go through the common, you don't see the same people day in and day out and all of that. And that's also going to be affecting how we design user interfaces and everything around it as well, because we have a share button, but what does a share button even mean if you don't care about sharing to other people?

54:46 April Zhu

Yeah, I think that the theory of third space is really interesting because that is really shifting more and more into the virtual third space, in my opinion. A lot of people today probably don't really have a third place to go to. Personally, I find myself in this situation. I don't have a physical third place to go to. So I'm very reliant on a virtual third place, on the online platform or things like that.

55:14 Kirupa

Exactly. You know, when I think of what was my favorite third place, I used to love bubble tea. I love bubble tea. And as I get older, of course, he realizes it's actually not the healthiest of things you can eat. But I used to go there all the time, and all the friends that I made, any friends that I made during work, I would take them with me to bubble tea as well. And so, to me, that was my third place.

And now that place is no longer around. It's been torn down. Now it's a big office complex, as all these things go. And so I'm like, okay, the office complex probably provides a lot of value, and it provides jobs and all of these things. But multiply that by hundreds of thousands of times, every single person where their third place was, let's say, a park bench or a park that they used to play with that has now turned into a shopping mall and so on.

I love progress, I love development, I love all these things, but is there a way to balance that with the way of preserving some of these spontaneous encounters that you can have, which you can only have in a third-level space? Not in something you can do day to day, where the spontaneity is going to be missing. You have encounters, but it won't be spontaneous.

56:25 April Zhu

I mean, I think it's so important, though. We should try to bring it back. And it matters a lot for our happiness. So maybe, ideally, when AI take over our job, when we don't have to work, maybe we can all go out and meet each other more and talk to each other more. Because it doesn't matter anymore.

56:41 Kirupa

I hope so. Because right now what I'm seeing is, if AI can do, let's say you spend eight hours a day doing work, and the problem is, oh, with AI, you can get it done in four hours. But it doesn't mean that the remaining four hours, you're now spending time meeting people. You now have twice as much work to do, because you've done four hours, you can do more work as well.

And so it's not a case where we're benefiting from this technological progress. We're more productive, but our leisure time has probably decreased significantly since AI has been introduced.

57:15 April Zhu

Yeah, I really don't like the trend. Why am I doing more work? Just because I'm more productive? I hope eventually we don't have to work as much.

57:26 Kirupa

Yeah, I don't know. I hope so too, but it just seems like it's always a race. It's always a competitive nature of humans in some ways as well. And it's one of those things where if I'm not working these four hours, someone else is, and they're going to out-execute me and out-hustle and outdo me on all these activities. So I need to keep up. And so the only way I can keep up is by doing more work than they are, even using AI. It's a never-ending race.

57:55 April Zhu

Maybe it's a situation where it gets worse before it gets better. So it gets worse when we are more competitive with AI, and then eventually because it's so strong that we are all not needed anymore. And then we can finally be freed.

57:59 Kirupa

I hope so. Yeah, I mean, that's under the assumption that, just like a classic machine, it's mechanical, that it has a breaking point. So after a certain point, the parts of this fall apart and then the machine just stops working.

What I'm seeing is, going back to the earlier thing, we don't quite know when AI is going to stop becoming more capable and doing more things that we thought was going to be what we're doing. What we're seeing is, even the machine is going to show signs of breaking slightly, the AI comes in and fixes it.

58:36 April Zhu

Going at full speed. Yeah.

58:39 Kirupa

Yeah, it just keeps fixing it. There's never been a breaking point yet.

58:44 April Zhu

It's crazy. Yeah.

58:46 Kirupa

Yeah, it is a crazy time.

58:47 April Zhu

It's crazy. Maybe when they run out of R&D money, it will be slightly slower. I don't know.

58:54 Kirupa

I think through all of history, there's never been a stopping of R&D money, whether it was for the cultivation of...

58:54 April Zhu

But they are being invested in a lot of money. Oh, we are using super-subsidized AI models right now.

59:04 Kirupa

Exactly, right? They keep saying that eventually the AI model will not be as subsidized as they are right now. But my view is that the cheapest model right now is not very subsidized anyway, because it's not as compute-heavy as the top models are. And six months from now, our current top-of-the-line models will be the cheapest model that's going to be there.

If you had to tell me, you know what, all I can use for the rest of my life from here on out is, like, Gemini Carbon, or ChatGPT's Sol 6.1, or Opus 5.5, I think I'd be okay with that. I'm like, you know what, this is really powerful. I can do a lot of things with this. I'm okay with this being the only model I can use, and everything else gets around to it, gets the cost down.

My hunch is that there's a next version of the model being developed right now, and the next version, the next version after that, is going to be even more powerful than it was before. And so I think we're going to constantly be in this race.

1:00:01 April Zhu

Yeah, no one can see where it's gonna end, and it's really heading at full speed. Every month you will wake up with a new model that's so much more powerful. And then you have these AI companies competing with each other.

1:00:15 Kirupa

Exactly, exactly. So yeah, it's gonna be fun times. Do you ever go to, or have you ever been to, one of these electronics trade shows where they often have prototype devices or upcoming things that are being showcased?

1:00:29 April Zhu

Not in Singapore. As much as I'm aware, there aren't a lot of this going on. Maybe more in where you live?

1:00:41 Kirupa

I see mostly on YouTube. I actually haven't been to one myself.

1:00:43 April Zhu

Or on YouTube, okay.

1:00:44 Kirupa

I see a lot of them on YouTube, essentially. And it's always interesting if you look at, like, over the last three or four years, right? And it's always interesting to see what they are predicting will be the next big thing that's going to be big in the next 12 months, for example, and how many of those actually turn out to be right.

And it's always fascinating to see the trends of what actually ends up being successful and what just ends up being a proof of concept that never actually goes and sees the light of day beyond a certain limit.

And the trends are always very interesting because, going back to what we talked about earlier, the world really is getting back into more of the fundamental basics of, do I even need a computer? Do you even need to do mouse and keyboard and typing? What are more natural language ways of getting things done?

And there's so much of it going in that direction, and robotics as well. And so when I always think about, okay, how do I up-level my skills? I'm a product manager at a big tech company, at Microsoft, right now, and my interest and hobby has always been on design and animation and all these kind of things, which are not part of my day job, but it is what I spend a lot of my time thinking about and playing with.

How do I reskill myself to be relevant in this new world when the problem space I'm solving for is not a website? It's not an application that you interact with and draw and drag things out to. It's a combination of real and virtual. What does it even look like? And those are some of the things where I often find those kind of conference videos to be interesting, in terms of the highlights of where the world's going.

1:02:14 April Zhu

Please recommend me some. I think that sounds like a really interesting topic. I've not watched those kind of videos, but I would love to.

1:02:22 Kirupa

Yeah, no, actually, I'll put some links in the description after a recording here. But yeah, it's always fascinating seeing where people predict the future to be.

1:02:28 April Zhu

Yes. What are you talking about?

1:02:30 Kirupa

We always had this, right? Like, the World's Fair was a trend that many countries have had for many years, where every year they'd have the smartest minds and the most creative people predicting what the future is going to be like. There are a lot of comical examples as well, like wirelessly sending electricity back and forth, like what Nikola Tesla predicted many years ago, to the monorail, which is a very common thing that many cities now have.

But it was considered science fiction back in the day, a single track carrying people back and forth. Who would have thought of that as a thing? And so it's always fun to see where people's imaginations and reality intersect, and the timing of it as well. Because one of the things we often find, you probably run into it as well, it's not just good enough to have a good idea or a good plan. The timing needs to be right as well.

Sometimes a great idea at the wrong time is just as ineffective as a bad idea in a bad time. And getting that timing right is always something I always love history for. You can always see, like, we're almost too far ahead of where the future is going to be, and they didn't succeed.

1:03:36 April Zhu

Yeah, I think that would be interesting to see in the past what people have been predicting about the future, and the trend of what turns out to be true, what turns out to be not true. That would be quite interesting.

1:03:48 Kirupa

Yeah, yeah, I think it was probably decades of completely being wrong until AI came along, and then within two years or three years, it turned all of those predictions that were, like, too far imaginable now into a reality.

1:04:03 April Zhu

Did people imagine AI to be this powerful, like, 10 years ago?

1:04:09 Kirupa

You know, I think people imagined it purely from a science fiction point of view. Like, I have this robot that can basically do all these things and more, but the actual practical applications of it... We always had examples. Star Trek's a great example. They often spoke to the computer, and the computer spoke back to them in various ways.

There's a classic TV show called Knight Rider, which is like a car that could actually talk and talk back to you and be self-driving and do all the kind of tricks that we're able to do right now. So there were things that were sort of predicted in a certain way, but it was always one of those things where we know very clearly what the problem is right now. We just want it to be solved in a more elegant way.

But what we're seeing, I think the more interesting ideas of predicting the future are things that people never even thought were problems, but then we have a solution for them. Dynamic UI is a great example of this, right? Nobody thinks it's a problem per se, but once you see it, you'll never be able to unsee it.

1:05:05 April Zhu

I see it, yeah.

1:05:07 Kirupa

Exactly. So I think those are the kind of things that are within the realm of possibility, because with AI, you now have the ability to do so. And that's something I always struggle with as well. It's like, how do I decouple my inherent feeling of constraints? Because I know this is physically impossible or this is technologically impossible, I don't even entertain that as an option.

But if I take a step back, maybe it's not impossible. Maybe it is doable. And having a fresh pair of eyes, which in some ways AI actually is, has been a new way of unlocking some of these potentials.

1:05:44 April Zhu

Yeah, I think we're at a point of time where we are running out of imagination because things are really... Yeah, AI has developed in a way I think most people did not expect it to. And just imagine carrying this level of intelligence with the new robotics technology. I think all the sci-fi imagination we had in the past will become reality. Like Detroit: Become Human, the game. We're gonna have that level of, like, physical...

1:06:14 Kirupa

We're going to have that. I think that's totally going to happen.

1:06:20 April Zhu

My God, we're probably going to witness it in our lifetime, which is kind of crazy because I didn't—like, just a few years ago, I wouldn't have thought that it's going to happen in my lifetime.

1:06:31 Kirupa

Yeah, exactly right. And so then it's like, okay, what are the things I never thought were possible that are now going to be possible also within our lifetime?

1:06:39 April Zhu

Going to other planets, maybe? Yeah, building a moon base.

1:06:42 Kirupa

I think so. I think, if you can... Like, right now, we've always been told the speed of light is something you just cannot get faster than. Maybe that's true, but what if there's, like, the classic bending of space and time, where your speed of light is still constant, but the terrain you're traveling through is not a straight line anymore? It's actually a curve.

Or maybe even you fold the distances between long spaces, where instead of going a thousand light years, you fold it to a thin piece of paper, and now within, like, 10 minutes, you just cut across all of it. Still, this space has a speed of light, but the environment you're in is manipulative. It's malleable.

1:07:22 April Zhu

Yeah, I mean, AI can do quantum physics, right? So maybe it's gonna be... We're gonna find out.

1:07:29 Kirupa

Yeah. Once it gets beyond breaking all the cryptographic codes that we thought were never going to be breakable, and the entire world changing around that, yeah, I think it's going to be a lot that's going to be happening.

It's both good and bad, but I think it's up to us to figure out how do we make sure, again, going back to the happiness example, how do we make sure we use AI to maximize the overall happiness of society in a way that also balances out that my happiness doesn't cause you to become unhappy.

1:07:56 April Zhu

Yeah, I know. I think it's a huge challenge for whoever is running the world. It's a new world order, and the current system is probably going to fail before something's rebuilt on top of it.

1:08:08 Kirupa

I think so. I think so. Well, April, it was fantastic talking to you about all of the topics that we touched upon.

1:08:13 April Zhu

Thank you so much.

1:08:17 Kirupa

And yeah, looking forward to seeing all the more cool things you keep creating. I know you share a lot of interesting things that you're working on, and so I can't wait to see what you're going to be creating in the future.

1:08:28 April Zhu

Thank you. I hope you keep exploring, because it's just so fun for me. For me, a lot of the motivation just comes from the satisfaction of making things and finding it.

1:08:40 Kirupa

Fantastic. Right. Thanks, April.

1:08:41 April Zhu

Yeah, thank you so much. I think you got me thinking about a lot of things, I have to say. Please share with me the conferences.

1:08:49 Kirupa

Will do, yeah. I'll put a link in the description so that you and everyone who's watching it and listening to it can also have access to it.

1:08:55 April Zhu

Yes, thank you so much.