How AI is Changing Design Thinking, with Sumin Chou

00:02

Announcer

At the AI Readiness project, we believe that in this remarkable age, AI isn't the main character.

00:08

Sumin Chou

You are.

00:09

Announcer

While the tech is racing ahead, it's the humans who learn to harness AI Mindfully that will win. Each week, we meet remarkable people doing just that. Join Kyle Shannon, tech leader and AI instigator, and Anne Murphy, fundraiser and AI consultant, as they lead the conversation about staying grounded, growing smarter, and leading with what makes us human.

00:36

Kyle Shannon

Oh, yeah, look at that hair. What have you done? Normally, you're here with a baseball cap. You've got it all hidden. You're hiding it from us. You. You've gone.

00:48

Sumin Chou

You've got everyone.

00:50

Kyle Shannon

It. It is. It is stunning.

00:53

Anne Murphy

It is. It is a national treasure.

00:55

Kyle Shannon

It is. You are a national treasure. The hair just goes along for the ride.

01:00

Anne Murphy

There we go. There we go. This is what I very humbly tell my hairdresser and my colorist. I just grow it. That's all I do.

01:11

Kyle Shannon

Exactly. That's your. That's your job. And carry it. You do have to carry it.

01:15

Anne Murphy

I do have to carry it. That is true.

01:16

Sumin Chou

Carrying it.

01:17

Anne Murphy

And it is a burden. I mean, it is a heavy crown.

01:19

Kyle Shannon

To wear, I will admit, and this is probably a little embarrassing. I do not have a colorist, which is. Hence the situation.

01:32

Anne Murphy

Yeah, but you. You're. You're. You're. Salt and pepper is. It's a good. It's a good salt and pepper. Yeah.

01:38

Kyle Shannon

You know, I'm getting. It's. That's it. My wife always says it's just not fair that, you know, as men get older, they get. You know, she likes it.

01:45

Sumin Chou

Whatever.

01:45

Kyle Shannon

It's good.

01:47

Sumin Chou

All good.

01:49

Kyle Shannon

We're here. We're talking about. This is AI readiness. We're talking about fashion readiness or. Or.

01:54

Anne Murphy

Oh, yeah, I forgot.

01:57

Kyle Shannon

But AI readiness. So I know that you have been hot and heavy deep into the Claude code world. Can you. You've launched a new thing. So last time were here, you were watching a new thing. You're very excited about it.

02:08

Anne Murphy

Yes.

02:09

Kyle Shannon

Is it launched? Story of your life.

02:14

Anne Murphy

So. So I'm gonna. I'm gonna hit on some key points that I think are. Are unique and potentially interesting to the audience.

02:20

Kyle Shannon

Okay.

02:21

Anne Murphy

Number one, last week, we talked about moxie, our, like, totally disruptive nonprofit technology, and we're just waiting for our first contract to be signed, and it's gonna happen any day, and then we're gonna launch. Launch, and we're gonna have a big event, and everyone's invited, and we'll unveil and we will sign up. New customers. Okay, so that's happening in late. Potentially. Well, I'm not going to tell. Okay. And then what's happening now is, in addition to that, is that we've been building the world's first matriarchal agentic business. So.

03:03

Kyle Shannon

World's first matriarchal agentic business. Okay, got it.

03:06

Anne Murphy

Company.

03:07

Kyle Shannon

Company. I have the words. Okay.

03:10

Anne Murphy

So I don't know if it was last week or the week before when I told everybody about how I stepped on my laptop and I lost my whole agentic company. Well, I rebuilt it from the ashes or from wherever. We don't even know from wherever it is. Somewhere in God's green earth.

03:26

Kyle Shannon

You didn't. You didn't restore it. You've recreated from what you learned.

03:31

Anne Murphy

Yeah, I recreated it from my brain, but with working with other women. And so it was like, all right, how can we do this better? Like, how can we bring in all the things that Beth has taught us about compound engineering? That's how we make our compound engineering skill. It's how Beth taught us to do it, not how.

03:48

Kyle Shannon

That's cool.

03:49

Anne Murphy

Repo over there told us to do. Right. You know, I've got, what, a couple hundred hours of recordings of women talking about what they need in their tech and in their businesses, and so we're infusing everything with it, being a matriarch. Thank you, silver Fox. And so I'm writing a story about it along the way, because, of course, as. As you are aware, we're. We're simultaneously converting to cloud code in our company, which is also my house hold.

04:28

Kyle Shannon

So wait, so just to be clear, Moxie is a new business. Matriarchal agentic company on a chip is another business, and then you're converting your business to full AI control.

04:43

Anne Murphy

Full AI control, yes. Yeah.

04:46

Kyle Shannon

And you. And you're still married, and you're doing this with yourself.

04:52

Anne Murphy

This is the only reason why I say barely and a lot. A fair number of people. Fair number of people in the audience know him. So you. I mean, he's a lamb. He's a lamb. But Don. Don Wackerley does not like Claude code. Right now, we are in the. We are in the bad phase. Oh, yes.

05:10

Kyle Shannon

Is it.

05:10

Anne Murphy

Oh, yeah.

05:11

Kyle Shannon

Is it misbehaving for him, It.

05:13

Anne Murphy

Well, it's doing all the dumb things that it does. But unlike. Unlike some of us, he. Don has never seen, like that, like, tantalizing, you know, thing that a Claude code or an open claw or whatever can do. He's like. It's just not like, nothing Good. Is really happening.

05:32

Kyle Shannon

So, wait, so he never had his Kevin McAllister moment with Claude code? He's just configuring it.

05:38

Anne Murphy

He's just trying to make it work, and it's not working.

05:41

Kyle Shannon

That. That's.

05:43

Anne Murphy

That's the deep, dark place.

05:48

Kyle Shannon

It's a deep, dark place.

05:49

Anne Murphy

One of the chapters of my book is called let's just Burn the freaking House Down Instead of learning Claude code. That's a chapter. Because he. And he did not say freaking.

06:01

Kyle Shannon

He was fully expressed.

06:04

Anne Murphy

He was fully. The fully expressed version. Who would have thought the Claude code would be the hardest part of being married to your business partner?

06:15

Sumin Chou

Wow.

06:17

Kyle Shannon

Wow.

06:18

Anne Murphy

Freaking code.

06:21

Sumin Chou

Darn.

06:22

Anne Murphy

Has the dark matter faster. Darn house. Yeah, exactly. So that's me. How about you? Or thoughts on my. On my circumstances?

06:34

Kyle Shannon

Well, thoughts on your circumstances. I. I mean, more power to you running a business with your spouse. I've tried to do that before. We. We have an agreement now. We don't do that anymore. But, but I. But I. It is a laudable enterprise, and in your case, Triple Enterprises. That's really cool. I mean, the only advice would be, I think for both of you. The thing that we already know, which is probably not going to make him happy, is that anything that he's struggling with right now will probably be irrelevant in three months anyway. There will be some other thing. But as I always say, what you learn in the journey means that whenever the new thing comes, that's easier, you know exactly what to do with it. So I don't think this is wasted effort. And the madness right now is madness.

07:31

Anne Murphy

It's maddening.

07:32

Kyle Shannon

Yeah, it's. It's really maddening. I mean, the. The problem with everything right now is that the. The. The core frontier models are still just not quite there.

07:44

Anne Murphy

Yep.

07:45

Kyle Shannon

They're not quite good at writing. They're not quite good at marketing. They're. They're pretty good at coding. But even there, they're confused at best. These agentic things like Claude code and openclaw and Hermes and all of those sort of things are nearly awesome. Like, everything is nearly awesome. And so I think that there's an expectation that they're awesome. And therein lies the rub, that whatever your gap of expectation of awesomeness is to where they really are. Yeah, that is so frustrating. And so I think it's going to get better. I mean, it is rumored that this week or next we'll see new versions of Opus, which we may or may not. But. But a chat GPT 5.5, which is apparently Supposed to be quite good, but we'll see.

08:34

Kyle Shannon

Is it going to change things or is it just going to be like another ratchet up in. Nearly awesome.

08:39

Anne Murphy

Nearly awesome.

08:40

Kyle Shannon

Yeah.

08:41

Anne Murphy

Well, listen to this little, this little Happy Horse shit that I tried to sell everybody, which, that. The great thing about Claude Code is that you just start planning really far in advance now. It's like, it's just so cool how the technology has supported us in planning so far in advance because you have to plan forever in advance because nothing freaking works. I'm like acting like it's a feature, not a bug. So. Great. Because we have this deliverable on next Monday. Start now. Start now. We've got, we've got six days.

09:19

Kyle Shannon

Well, I'll tell you. So let me share something with you that is something that has come up in my conversations over the past two weeks, and that is that I'm starting to see people that are using AI a lot level up to what I'm considering like 2.0 level behavior. There's like AI 1.0 behavior and AI 2.0 behavior. AI 2.0 behavior starts to look like this. Oh, AI is actually really not good at X. So I'm just going to do X on my own. And so, like, I'm starting to see people get much more sophisticated about, okay, I will absolutely use AI here. I may or may not use it here, and I will definitely not use it here.

10:08

Kyle Shannon

That idea of chain of craft and pulling a lot of the chain of craft out of the digital realm and either going back to full analog or just going to like, I'm just going to write this in Microsoft Word or whatever it is. Right. Google Docs and not use any AI because it's just, it gives you, it does. It gives you a false sense of completion.

10:30

Anne Murphy

Yep, yep.

10:32

Kyle Shannon

And you may look on the surface like you're 90% done, and then you realize through six hours of work that you were only 40% done. And that's maddening. So you might as well just, instead of doing six hours of undoing, just take the four hours it would have taken you to do it for real anyway. So that I'm starting to see. And I think that's very healthy. Right. I think that's, that, you know, I guess what it looks like is people are reclaiming their agency. They're not treating AI as this genius that's going to do all the stuff for them. They're saying, okay, it's a component in a mix.

11:11

Anne Murphy

Yep.

11:12

Kyle Shannon

You know.

11:13

Anne Murphy

Yep. Hey, you want to. Along these lines. So the, like, the human AI interaction. So I did a test because I'm. I'm putting all of my, like, fundraising content into a little, you know, area a professor or something like that. Or maybe it's just a resource for me, maybe it's for my clients. I don't know. But I wanted to test it, and so I took all of my files, I anonymized my files for a client, everything we had done for a whole big campaign, and I used some of my other templates from other things that I've done throughout my life, mushed it together, ran it through my fundraising brain, and created a plan that I could compare to the plan that was my deliverable for the engagement. I mean, obviously, this story writes itself. My deliverable was amazing for that campaign. I loved it.

12:18

Anne Murphy

It worked. We were happy and everything. Yet objectively, in terms of, like, excellence, what AI and I made together was objectively much better. Really better. I. In fact, I am sending two to two people to review and tell me, if you receive this from me, what would you think?

12:42

Kyle Shannon

And better in what way? Like, what leveled it up? And I also want to hear about your process about. Because I assume it wasn't just stick it in Claude code and squirt out a thing. Right. It was probably lots and lots of iterations. So I. I just want to hear about that process. But why is it better first?

13:01

Anne Murphy

Why is it so. A couple of things that make it better is that when if I'm working on a report, let's say I'm working on a. Like a. A feasibility. A feasibility study. And I can say here are five or six narratives that came up again and again. Very, very helpful. But you know what's a little bit more helpful is to have, like, database storytelling for them to click to and see a head start on their next grant application. That's just better. It's better. It's not me saying, I'm your consultant. Here you go. You need to tell stories about social justice and open space, intersection with racism. Go. It's like, yeah, you need to tell this story like this. And here's three examples, or here's three things that you can go with. That's. For me, that's a biggie.

14:06

Kyle Shannon

Oh, so it's like. So it's like AI. AI is in the mix, augmenting, bringing to. That's really good. Because what that allows you to do is scale you. Right?

14:17

Anne Murphy

Scale me.

14:17

Kyle Shannon

Because before writing, it had to be that Voice, right? They'd say, well, what does that actually mean? You'd be like, well, let me get back to you. Let me write some things up. Let me go find a story like that. Now it's just like, there.

14:29

Anne Murphy

It's there. And then if they discover through that process that what they really want to do is more storytelling. Now we've had a little discovery moment in our relationship. They're like, oh, you guys do stories the way that we want to do them. And I'm like, okay, cool. We can do. I can do. I could tell. Write a lot of good stories with. With everything that I know.

14:49

Kyle Shannon

That's when you introduce them to storyvine.

14:51

Anne Murphy

That's when I talk to them about storyvine. I'm like, you know what?

14:54

Kyle Shannon

You got to build that into your tool that the answer to every story question is storyvine with your affiliate link.

15:09

Anne Murphy

So that's, you know, an example of what makes it better.

15:14

Kyle Shannon

And what was the process like? Like, how long did it take you? And, you know, where. Where did I do the heavy lifting? Where did you do the heavy lifting?

15:26

Anne Murphy

Gathering everything up and. And indexing. I. That's not possible for me. Doesn't matter. You could, you know, can. No can do as well as we say for you.

15:36

Kyle Shannon

Yeah, yeah.

15:38

Anne Murphy

So having Claude Code or Manus or Perplexity or Comet or whatever, all of my things, I created all these research agents for them to just, like, go into all the nooks and crannies of my world and bring things back together and make them make sense as, like, sum is bigger than the whole of the parts thing and then pull some stuff from outside. So I created a really great database of, like, things I believe, things that I know, things that I've conveyed to people before, like, totally approved everything. I kind of liken it to, like, imagine were all members of a weird crow nation, and we're just crows going around picking up the little sparkly things and bringing them back to our context. So we want our context to have the best stuff. They don't want any junk.

16:30

Anne Murphy

So that's a big, huge piece of it. And upon that, I can build an interface, right. That allows access to all of that and not to anything else that would interfere with my way of doing an a feasibility study.

16:49

Kyle Shannon

That's really cool. Have you. Have you put a bunch of that stuff into Notebook LM and played with it?

16:55

Anne Murphy

It's very much like that. Yeah, exactly. Yeah.

16:58

Kyle Shannon

It's similar to that.

17:00

Anne Murphy

Notebook LM made me smart again because I can see things instead of.

17:04

Kyle Shannon

Yeah, yeah.

17:06

Anne Murphy

Hi, Gareth.

17:08

Kyle Shannon

Gareth, what's happening, dude?

17:10

Anne Murphy

I owe Gareth an email or a DM or maybe both.

17:14

Kyle Shannon

Wow.

17:15

Anne Murphy

Just so everybody knows, all right, you.

17:17

Kyle Shannon

Were officially on notice.

17:18

Sumin Chou

You're.

17:19

Kyle Shannon

You're officially taking responsibility. This is what. Part of my life's work is figuring out how to get back to people in a timely manner, so.

17:28

Anne Murphy

Oh, you know, I have something for that.

17:30

Kyle Shannon

Okay. Yes, I'll take anything.

17:36

Anne Murphy

So I have a skill. His name is Romy.

17:39

Kyle Shannon

Okay.

17:39

Anne Murphy

Like Rome. Rome. Rome, romance. Kind of like Romy is. His one and only job is to see how behind schedule we are with the people who we love the most. Our collaborators, our friends. He. He goes through everything. And every morning lets me know, hey, here's where you're not getting back to somebody. And it would. It kind of sucks that you're doing that.

18:05

Kyle Shannon

That would be. Yeah, that. I'll take that. I'll take one of those.

18:07

Anne Murphy

I'll give you a Romy.

18:09

Kyle Shannon

Yeah, you just tell me where to put him and what to do.

18:12

Anne Murphy

I would love someone else to tell. Remind me of that. You know what I mean?

18:15

Kyle Shannon

Yeah, no. Oh, I. I totally know what you mean. In a deep and visceral way. That's really good. That's really good.

18:26

Anne Murphy

That's a lot about me. Kyle, please tell all of us how you are.

18:32

Kyle Shannon

I'm doing a lot of speaking. So I spoke Monday in Boulder on radical self expression. And basically. Basically the idea that. That our job now is just to have ideas. And if you get curious with AI and you understand all of what's possible, you can have an idea and you can express it in way more ways than you can, you know, you could have historically. And that people seem to enjoy that. And then right after this conversation today, I am on a panel with the. For themed Entertainment Association. So basically theme park designers. And I'm gonna get to show them Sydney, my musical and. And talk about how I created that in this kind of radical self expression kind of way. And then next week, I'm in D.C. for an AI salon. Fly in.

19:24

Kyle Shannon

I'm in Boston for an innovation summit talking about the great repurpose. And then the following week, I'm at Social Media Marketing World talking about AI readiness as a. Embracing AI readiness as a basically career extender. So I am busy.

19:44

Anne Murphy

Boy, these people are lucky. They get to hear from you and see you and meet you. Awesome.

19:50

Kyle Shannon

Good job. Some of the Kyle magic. I'm trying to get myself to state Treasure. You know, maybe at some point, but I got. I want to get Sue Min up here. I'm really excited. He's. He's one of my favorite people on the planet. Let me just tell you about the AI Salon. If you've not joined the AI salon, please, please go to the salon. Check us out. The salon has been around since the week ChatGPT came out and is really a place for us to be humans first and talk about AI as an amplifier of our ideas rather than, you know, the replacer of our ideas. And I, I deeply believe in that. The more I use it, the more I think about it that way, the more it is within the salon. We have the AI Salon mastermind.

20:47

Kyle Shannon

And we just kicked off a series that Andy Scarantino is teaching about separating your identity from your work. As all of our work is about to shift, all of our identities are about to be borked. And so there's just all sorts of really amazing stuff and amazing people there. So, so go check out the AI Salon.

21:08

Anne Murphy

Really, really good people. I'm so happy to see you see what you all are offering now. This is good stuff.

21:15

Kyle Shannon

Yeah.

21:18

Anne Murphy

Oh boy. Sheely day. I One of the things that we do is we have an in person conference and we have been reviewing, a group of us have been reviewing all of the speaker applications this year and absolutely incredible. The breadth. The breadth and the like quirkiness as well and the just the totally different points of view on AI and the future. Just I feel I can read like 10 or 20 before I like, my brain is just like so stimulated because of the. Just. Yeah, we just hang around with some really freaking smart people, Kyle. Very, very smart, very creative people who can package their ideas up in a way that makes it irresistible. You're like, I think we're down to like 75 people for the, for a couple of spots that are left.

22:16

Anne Murphy

I mean we started with 200 and I just, I want to take everyone.

22:21

Kyle Shannon

Yeah.

22:22

Anne Murphy

Yeah. So we're an in person conference and we are, we gather. Also we have a free gathering every Saturday from 10am to noon Pacific time. It's social Saturdays. We have one or two experts come in and teach us from their zone of genius. And then we have a membership program and we have our cohorts where we certify AI educators. We have an incubator for AI consultants and we're working on launching a certified AI officer program soon.

22:55

Kyle Shannon

Beautiful. I love it.

22:56

Anne Murphy

Yep, super excited.

22:58

Kyle Shannon

So yeah, we'll check out the beads AI next we're going to bring up Su Min Cho. He is someone that I know from Many years ago. I knew Su Min when he was just a little brother of someone that worked@agency.com and he has since gone on and he is with Schema Design. He's one of my favorite people in the world. So I am really excited to bring Sue Min up. Sumin, Hello.

23:24

Sumin Chou

Nice to see you.

23:27

Anne Murphy

That's high praise.

23:30

Kyle Shannon

Cool dude, man. He's a cool dude. So why don't you introduce yourself, tell us who you are, what you're up to, and then let's dig in. Let's open up the conversation.

23:41

Sumin Chou

I was thinking back to how long ago I've known you, Kyle. I think it was embarrassed to say. The late 90s. Mid to late 90s.

23:53

Kyle Shannon

97, 98. Probably 98. But somewhere. Somewhere back there. Yeah.

23:59

Sumin Chou

And then. Yeah, starting at Time Inc. I think you and Chan had these parties at Time Inc. With Vibe online. And then also working with you@agency.com was great times. So. But yeah, and then fishing together in the Bahamas.

24:19

Kyle Shannon

Oh, yeah, that was pretty amazing.

24:22

Anne Murphy

That sounds fun. I'm gonna go fishing in the Bahamas right now.

24:25

Sumin Chou

Yeah, me too.

24:26

Anne Murphy

Well.

24:26

Kyle Shannon

Well, wait, so I got it. So, funny story about fishing in the Bahamas. So, yeah, I. I don't know if you know this about me, but I'm a bit of a trash talker. And so. So we're getting ready to go out fishing in the Bahamas, and I'm. And I'm like, making fun of every. Everyone about how seasick they're all gonna get. And then. So we get out there on the. On the. We're trolling and we've got following seas. And I don't know if, you know, following seas, but they make you really sick. And so I got seasick. And so I secretly. No one was around me, so I secretly just, you know, right over the side. And then I thought I got away with it. And the captain of the boat goes, yeah, it looks like someone lost it.

25:12

Kyle Shannon

And I was, like, looking around, like, who? And meantime, I admitted it after the fact. But that was pretty awesome.

25:22

Anne Murphy

That is awesome.

25:24

Kyle Shannon

Yeah. So tell us. Tell us. Tell us about, you know, your background, what you're up to.

25:31

Sumin Chou

Yeah. So starting way back in college, I was studying studio art and art history. Actually, right after college, I was teaching at the Brooklyn Museum, teaching art and. And giving studio tours, planning to go to graduate art school. And then I just sort of fell into computers and design specifically. But one of the things that I sort of credit to my interest in hypermedia and the Internet was I was teaching. I would Take little kids around and we'd have a blast. We'd go on these treasure hunts where and these are like five and six year old kids. And I'd say, you know, we're going to go on a visual treasure hunt to find all the triangles we can in the museum. And so we'd run around the Egyptian Gallery and find triangles there and go to the contemporary gallery upstairs.

26:32

Sumin Chou

And it created these kind of linkages in concepts that were so resonant with the kids and also the parents that were coming along and they actually advocated for adult tours as well. But to me it was so powerful that you could basically break through these walls where, you know, why is the Egyptian Gallery separated from the African Gallery? You could kind of break those traditional barriers down and create linkages. And that was just for me, you know, the web was still just starting, but you know, the power of a hyperlink to connect one page or a concept from one idea to another was incredibly powerful.

27:24

Kyle Shannon

And so, so simple and so powerful. Right?

27:28

Sumin Chou

Yeah. And, and so that got me really into interactive media. And you know, we have very simple tools. And you know, Kyle, I'm sure you remember these experiences were very basic and fundamentally it was about just when were working at Time Inc. Was about scraping the Time Inc. Content and basically putting it up in very simple web pages. I would actually argue that kind of continued for a long time, probably longer than it should have. And I think really when ChatGPT came out, it really kind of made this moment where we had to really rethink digital experiences. I think it's been exciting, but also really kind of questioning everything that what does it mean to be a designer? What does a website mean? What is it for? You know, all these fundamental existential questions have surfaced.

28:41

Sumin Chou

So I feel like I'm deep in the weeds of trying to figure that out.

28:46

Kyle Shannon

Can you talk a little bit about schema design and sort of what you were doing traditionally, you know, two years ago, three years ago, in the olden timey days.

29:00

Sumin Chou

So from timing to. And so I was at Pathfinder, which was one of the first corporate websites online, and then to agency.com and you know, working with Kyle and working on British Airways and some really interesting client accounts. And then I was also working at later several digital agencies, then became design director at the New York Times, which I learned so much. Then I, after the times formed my own little design studio called Concentric. And during the pandemic things were like collapsing and I got an RFP and it was both data Visualization and sort of traditional content website design. And I reached out to Christian Schmidt who had founded Schema. He and I had worked together. He was my design intern at the New York Times and I knew he was the big things amazing. He was at IDO and then moved to Microsoft.

30:19

Sumin Chou

And, and so we talked about, we worked on this partnership on this project and we just decided that the future really was going to be these content and data rich experiences. And so I joined Schema I guess now close to five years ago and we've been doing really fascinating projects that are, you know, really data driven experiences. And you know, I think with AI so much is even more so now all about the data. And once you have that data, how do you express it? And that's sort of where we are today is trying to still define how we can create interfaces that really help users access large data sets.

31:15

Anne Murphy

Is that what like when you say a data, a rich data driven experience is the experience, what is the experience that you're trying to design?

31:28

Sumin Chou

Yeah, I can run a large gamut. We do large conferences, data visualizations, for example for Bloomberg, the NES conference for several years. And so these are kind of the backdrops for major conferences and we've done that for some financial clients as well. Also just on the back end side, we've done, we did a project for the New York Times as well on their advertising platform as well and some other things. So it's all about just sort of being able to kind of manage data and sort of controlling how you see it to kind of get deeper insights as well.

32:18

Anne Murphy

Cool, cool.

32:20

Kyle Shannon

Super cool. You know, and Ann earlier was talking about the idea of, you know, taking this work she had done before and you know, being able to first organize it in a way that she couldn't normally. Right. Or just, you know, with standard tools and then coming up with a new improved version of this. Like what's been your experience? Is AI making things easier? Is it making new things possible? Is it you get to 80% and think you've solved it and then you realize you're not even at the start line anymore, that it's totally screwed everything up. Like where. Talk, talk about your experience of like where are you on the AI? Love it, hate it.

33:02

Sumin Chou

Yeah, I mean I think all of the above that you mentioned, you know, we're designing or building this experience for a client and they were seeing figma static screens of what's intended to be a dynamic parallax experience. Figma for all its prototyping and even Figma make. It just doesn't translate how the experience would work in full motion as a coded prototype. I just decided, I was like, oh, what if I can code at least like how we could set up a framework for the Parallax experience, how the.

33:49

Kyle Shannon

Animation might work, for example.

33:50

Sumin Chou

Exactly. It created a incredible widget tool where it said, oh, each of the layers. So it showed the Parallax showing different layers with sliders. So I could perfectly slow down or increase the speed layers. The client could understand how Parallax Motion works where you're having different layers of content or images go by at different rates of speed to get that effect. But to fine tune it as well, to get that magical experience, that kind of cinematic quality is, you know, they had the ability to kind of see how that all could work together and we could fine tune it. So that was just a moment where it's like, oh, there's this like little piece in the workflow that, you know, AI totally solved for us. And I just extrapolate that to, you know, what, how can we apply this to all these different areas?

35:04

Sumin Chou

So that's really exciting. And yeah, I don't know where it's going to go, but I'm definitely in, you know, in that pathway right now.

35:15

Kyle Shannon

Yeah. So in that case, it lived as a effectively a prototyping demo for something you were building. Did you end up incorporating that into the site in any way or did it just live as the standalone thing?

35:28

Sumin Chou

And yeah, I mean to get it to production code is not as big a leap from that. We can fine tune all the timing sequences the client can think about. Okay, we're going to have fundamentally like four or five different layers. We can think about the content flow of assigning some images, maybe that are the background layer and then foreground images that are featured that affects upstream their content delivery. To think about how that's going to go into the template. Well then we can take all of the settings that we fine tune and build that into the actual production code as well. Just from that little sequence. Think about, oh, how do we expand that to deploy it? So, you know, that was just incredible because otherwise traditionally it would take so long to do that.

36:29

Sumin Chou

And just the ability to shortcut that workflow is great.

36:38

Anne Murphy

What are some of the trends you're paying attention to or some of the. I don't know how you are about tools. Kyle and I have kind of gone back and forth over the last few years about whether we care about the tools, whether the tools are in charge of us or whether it's a combination of the two. So can you talk a little bit about trends or tools you're especially enthused about from your point of view?

37:06

Sumin Chou

Yeah, I've been working and we have, on our schema website, we have a separate lab section. So we do these little prototypes experiments that we've done some Apple Vision Pro experiments that we posted, but I'm still working on a knowledge graph where, you know, using vector search, we can sort of find concepts or ideas in a data set that are basically of like compelling interest. And then what I want to do is figure out an interface where you can select a couple of those ideas or nodes and synthesize understanding around those. And so that's something that I'm working on and hoping to share at some point.

38:01

Kyle Shannon

But that's very cool, I think. Yeah. Anything that can help people understand how this stuff works.

38:08

Sumin Chou

Exactly. I mean, I think that's where I'm really excited is less like the workflow efficiency to me. But what if we can create tools that give people a deeper meaning or insight? And so I don't know if that's a trend, but that's something I'm interested in.

38:29

Kyle Shannon

There's something that I mentioned it a bit before that people are sort of choosing when to use AI, when not to use AI. A good buddy of mine, Kevin Clark, created what he called slow AI, which was he basically did a prompt so that whenever the AI got to a decision point where it had five or more choices, he asked it to display those choices. So he was introduced and then he would choose which of the five and then it would get to another decision point and show him another five. And so he was introducing friction into the system. Yeah, And I think that there's, I mean, you come from a traditional art background. The AI is largely a friction free environment, which is part of what makes it so addictive and exciting and fun.

39:19

Kyle Shannon

And part of the genius of humans is the moments of friction. Right. And it's like, I'm curious, like, where's your relationship with. Because you've always been digital. Right. But you also have this traditional background and you really, you appreciate being in the museum and walking from one room to another. Can you just talk about your relationship with the speed of AI?

39:44

Sumin Chou

That's a really good question. I think that's been, and I think were talking about it before this idea of a journey that is so fast because I've learned so much and. But it also just brings up lots of questions and I Think this is a fundamental question of what is everyone's relationship with AI? And, you know, I think there are people that are really averse to using AI and that's okay, too. I'm sort of coming in with it with a little bit of skepticism. There are some people that are just like, gung ho and they're. They're trying to figure out also, like, what can I do to get myself out of the way so that I'm not causing friction.

40:31

Kyle Shannon

Right, right.

40:33

Sumin Chou

Interesting path too.

40:35

Kyle Shannon

Yeah.

40:37

Sumin Chou

So. But I think everyone's own experience and. And use is fascinating and. Yeah, I mean, it could be just like a personal AI relationship. I don't know.

40:51

Kyle Shannon

Yeah, yeah. It's just a different thing, you know, when you do it. When you. When you choose to add friction in a frictionless environment, it's actually a really fascinating choice.

41:02

Anne Murphy

One of our. One of our friends, Beth Lyons, talks a lot about how the work in front of us is not helping AI understand the context that a lot of people are talking about and, like, how to navigate all that better. It's every bit that you can get it to be smarter about how to work with us as individuals. Like, that's how we're actually going to have our moats. For one thing. I can really relate to this idea of what Kevin's saying with the, you know, slow decision making and have put a couple of different friction points for my stuff as well.

41:45

Anne Murphy

Like, it h. It has to pass these three things because of the thing that you and I have talked about so often, Kyle, with, like, with people like us who have lots and lots of ideas, sometimes we don't have a ton of respect for our own ideas because we know there's going to be another one right around the corner. On the other hand, we also know that there are some gems in there.

42:09

Kyle Shannon

So some of them are good, some of them are worth keeping our worth.

42:14

Anne Murphy

The friction point of is this something aligned with who I want to be when I grow up? Is this something that helps me, whatever. This, that, and the other? Or is it just some freaking rabbit hole that isn't going to help me? And I'm going to wake up in three days, as you've seen Kyle, and you know, only be able to wear a baseball hat for like four days while I fall asleep on the. On the desk so that I love the friction. The friction piece of it. I think it's healthy and I think we're really gonna notice what it's like to be out in the world when we haven't had a lot of friction, you know.

42:53

Kyle Shannon

No, I think so too. I think, I think there's also just the amygdala, like that speed of, you know, sort of endorphin pumping, like you just go, going. It's, it's, you know, it's frying a lot of brains. In fact, I think there are some open claw people on Twitter that are changing their sleep schedules so they can manage their agents 24 7, they're sleeping in four hour stints so they can wake up and manage their agents. It's like, that ain't sustainable. Subin, I want to ask you about, as someone with a traditional art background, I know you're a fan of digital, I know you're embracing AI. And I'm curious, you know, do you have any boundaries around, you know, generative AI for art? Do you, like, what are your thoughts on it?

43:41

Kyle Shannon

I know there's, you know, ethical quandaries and quagmires and black holes and gray areas. I'm just curious, what's your philosophy on the new tools? You know, where those boundaries are?

43:56

Sumin Chou

Yeah, so that was one of the first questions I tried for my, to answer for myself. And so when ChatGPT first came out, I started, I created my own GPT, which was a drawing tool. And I was basically prompting it to create these what I called wireframe sketches. And aesthetically I was just seeing so much sort of AI imagery that to me just looked very, you know, just not in keeping with my own aesthetic. So that was like one of the things. And I posted a bunch on my LinkedIn profile where I really wanted the output of these drawings to look like they were natural hand drawn by a human. And so I actually said, you know, it should, the line shouldn't be necessarily straight, it should be printed on natural linen paper and so on. And the output was actually really surprising.

45:10

Sumin Chou

And so I'm, you know, I was really interested in like, how can I kind of push AI to kind of creep into more humanistic Russians? And that's something I'm still advocating and doing. I was just at a friend's. My friend Gary Zamczyk is a professor at Columbia University at the Engineering School. And all of his students were working on these prototype concepts using FIGMA make. And one of the things that, and I know these were just like early concepts and prototypes, but one of the observations was that because it was primarily done through prompting and through FIGMA make, the interfaces still felt like they were searching for that sort of human piece where it didn't. It felt like it was designed by an AI agent, basically was communicating to another, not a human.

46:19

Sumin Chou

A lot of the experiences kind of lack that context or just that kind of feeling or expressiveness that would, I think, engage with another human. And I think that's also a challenge in what we're designing today is how do we still, you know, design, even though we're using all these AI tools, How do we, you know, fine tune it to kind of be engaging for humans? I mean, you know, designing for AI agents is a whole other sort of area. But I think how do we, you know, and that's why focusing on the brand identity, the aesthetics, that kind of traditional side is still really important to do in a way that's meaningful for humans, I think.

47:14

Anne Murphy

Okay, I, I have a perfect example of that. Our. On our website. The she Leads AI website is beautiful. It has like really bold design and people really resonate with that to it. Like, it does not. It boggles my mind how many people have said to me, the reason why I got involved is because I. Of the design of the aesthetics, the brand. Well, here's the thing. I love it too. But we have nothing. No, there's no SEO, there's no aeo, there's not nothing. No, no agent, no robot, no Google, no. Nothing is ever going to find us on the Internet. And so it keeps coming back to. It keeps coming back to, well, we should just, you know, rebuild the whole website. And I'm like, yeah, but here's the thing, right?

48:05

Anne Murphy

In this case, like, let's just say zero, you know, let's say we've got zero reach via SEO or whatever that came from our website, right. Enough people showed up there, the right people showed up there. Right. And then those people then came into our programs already feeling connected. So it's so easy to fall into the hype. And I love nothing more than to be pitched like a solution to a problem, like website schema. No, thank you. Happy to have somebody else do it. But that doesn't really solve the. That's only one piece of it. We have to remember, right. People are. Humans are humans. We need to see things.

48:50

Sumin Chou

Yeah. So I'm really interested in those types of experiences or UIs that augment or help humans, like reach new insights that they never could have done before. And I think that's going to change expectations for people going to websites. And I think the old websites, which I would argue haven't really changed since our early days at Time Inc. Kyle.

49:20

Kyle Shannon

Well, they've gotten more boring and more predictable.

49:23

Sumin Chou

Yeah, but now it's really like a challenge to rethink. If, if I can do so much with Claude code, why can't a website do that too? And, you know, and so I should expect to be able to, you know, ask questions and get insights that are really meaningful and not just like reading, you know, an article that took me 10 clicks to get to find the right. That's still a problem. And so I think we're going to start really seeing sort of pre AI websites versus post. And the difference is going to be really stark, if it's not already.

50:11

Kyle Shannon

I'm curious. So Schema's an agency model, right? You're doing work for hire for brands and things like that. Is there a tension or is there a pull to create tools? Because I feel like you're on the front end of understanding how we might interact with these more dynamic tools. Probably better than most people on the planet. And it feels to me like a seductive thought might be, hey, do we build tools to do this? Where's that tension? Are you happy being a service business? Do you know, have aspirations for the tool? Not that one's better than the other. I'm just curious. It feels like you're set up to be able to do that. I'm curious if there's a design there.

50:54

Sumin Chou

Yeah, I mean, I think our whole model needs to change also. I, I think there's very serious AI transformation going on where we, there's a value that we can provide which is kind of unearthing these unique UX's and UIs and tools that really can transform a company, whether it's their internal processes or external communications, etc. So that's what I'm really excited about with our clients going forward.

51:31

Kyle Shannon

Beautiful. We're at that time. So we have a question that we ask all of our guests and this too. And I think this is fascinating coming from you. And that is what does AI readiness mean to you? And then follow it up with, what would you say to someone just getting started with all this AI nonsense?

51:51

Sumin Chou

I think AI readiness means for me at least, embracing AI being ready and open to AI, not being fearful. And Kyle, you talked about, to me about this a while ago, but you know, we're rethinking what AI means as a design agency, what it means for our capabilities and services and so being open to that and embracing that, but, you know, steering our Clients as well, to all the potential is, I think, a big part of that too.

52:29

Kyle Shannon

Yeah. Beautiful. And what would you say to someone just getting started?

52:33

Sumin Chou

Just dive in. We're. We're all at the beginning.

52:37

Kyle Shannon

Yeah, exactly like.

52:38

Sumin Chou

Oh, my gosh, back in those days.

52:43

Kyle Shannon

It really is. Well, you know, subin. I mean, like, you were there at the early days of the web. I was too. Like, compare where we are now to. To. To what was going on back then.

52:56

Sumin Chou

Yeah, I mean, it's just. It feels so compressed right now. But I. I tell you, there's like a real messiness right now. There's a gap which is like, I'm still working in Claude code and then. Yes. I'm sometimes like pasting Council Claude code into figma to export. Like, it just seems like there's. And maybe haven't like optimized my workflow, but it. It still, you know, there's like, this isn't you.

53:30

Kyle Shannon

Okay.

53:33

Anne Murphy

I have never.

53:34

Kyle Shannon

Going through. Right. This is like, this is what Donnie's dealing with.

53:38

Anne Murphy

Oh, my God. Yes. Yeah. I mean, I've. I have never mapped out so many processes, trying to figure out ways to like, is this the way we're. You know, and then you're like, no, I'll try it this way. No, I'll try it that way. No, I'll stand on my head. No, it's the day that ends in why it's definitely not going to work. I mean, it's just. Yeah, yeah, it's hard. I think it's hard for normal, everyday people.

53:59

Kyle Shannon

It is. It's crazy times right now. Well, well, listen, Suman, thank you for doing this. Thank you for being here.

54:06

Anne Murphy

Really great. Thank you.

54:07

Kyle Shannon

Love you.

54:08

Sumin Chou

Thanks for having me.

54:10

Kyle Shannon

Yeah, this is. This was awesome. We're. We're cutting out a little early. I got to jump over to another talk right now. So. So thanks for indulging that, everybody. And, and thank you so much, as always, so much.

54:22

Anne Murphy

Have a great. Have a great talk. Thank you. So good to meet you. Thanks, everybody, for joining us.

54:27

Kyle Shannon

Cheers, everybody.

Transcribed by https://fireflies.ai/